SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 [Whole Volume] Cultural Chronology of the Gulf of Chiriqui, Panama Olga Linares de Sapir Research Associate, Smithsonian Institution SMITHSONIAN INSTITUTION PRESS City of Washington 1968 A Publication of the SMITHSONIAN INSTITUTION United States National Museum LIBRARY O F CONGRESS CARD 68-60007 U N I T E D STATES GOVERNMENT P R I N T I N G OFFICE, WASHINGTON, 1968 For sale by the Superintendent of Documents, U.S. Government Pr in t ing Office Washington, D.C. 1.04,02. Price $4.50 Preface This field work was part of a much larger program lasting three years entitled " In te r ? relationships of New World Cultures" and organized by the Institute of Andean Research with financement from the National Science Foundation (Ekholm and Evans, 1962). Evidence of interconnection during the Formative period and at other points in time was sought in particular regions of the Intermediate area between Mesoamerica and South America. The Panamanian region surveyed is the Pacific coast from the Costa Rican border to the Colombian border. I t was covered in two consecutive seasons. The first, in February through May of 1961, included the western coast and principal offshore islands from Punta Burica to Punta Mariato in the southwestern corner of the Azuero Peninsula. The second, in 1962, extended the work to cover coastal Darien. This is a report of the work done in the first season. The participants of the project were Dr. Charles R. McGimsey I I I of the University of Arkansas, as principal investigator, and two assistants, besides me, who were also graduate students at the time: Mr. William Bishop, geologist, and Mr. Freeman Mobley, ethnologist, both from the University of Arkansas. In four months we visited over 60 sites and tested 12 of these. To each of the sites that we recorded we assigned a letter-number combination. The letters designated the district, and the numbers, the sites in the order found. The site of El Cangrejal was thus given the code number S L - 1 , with SL standing for the district of San Lorenzo in the province of Chiriqui, and number 1 standing for the first site found in this district. Island sites could not be assigned to districts, so they were simply given a number preceded by the letters IS which stand for "isla." Thus, Isla Villalba became IS-7, Isla Palenque became I S - 3 , and Islas Las Secas became IS-11 . Reported in this study are the four most important sites where the most concentrated work was done. These are the sites mentioned above, all of them located in the Chiriqui Gulf. Three are on islands, the fourth on the mainland. Distances between them vary. The closest to each other are Isla Villalba (IS-7) and El Cangrejal (SL-1), separated by 10 kilometers of water channels. The farthest away is Las Secas (IS-11); it is 37 kilometers from Isla Palenque (IS-3), 53 kilometers from Isla Villalba (IS-7), and 55 kilometers from El Cangrejal (SL-1). The three island sites are located on ridges in back of the beaches, often the highest spot available. The mainland site is on a high bank near the edge of an estuary. All four sites consist of middens formed by the deposition of abundant cultural materials: discarded pottery, stone tools, and the remains of marine and land fauna. These were choice loca? tions and, to judge from the refuse, the occupation of them was continuous. Listed in summary form, the main aims of this report are as follows: (1) To estab? lish a preliminary chronology for the Gulf of Chiriqui, based on local sequences of pottery; (2) to trace, in time and space, the cultural contacts of the inhabitants of these sites with PREFACE each other and with peoples on the mainland; (3) to reconstruct as much as possible the manner of living and the subsistence patterns of the occupants of these sites; (4) to sum? marize ethnohistorical accounts of the area in an effort to tie these in with the termmal part of the archeological sequence; (5) to place the Chiriqui phases in proper chronological relationship to other Panamanian sequences; and (6) to relate the chronology thus estab? lished to those of other lower Central American areas. Two approaches were used in our pottery classification. The first was a subdivision of the decorated sherds into types, differentiated from each other primarily by decorative treatment and secondarily by paste attributes. Their percentage frequencies were plotted by levels. Since the plain pottery does not always appear to show any distinctive differences in temper or paste, it was not divided into the full, conventional pottery type description, but instead was sorted into groups named A-J . Only one plain ware was distinctive enough to be described fully and was labeled Tarrago Bisquit Ware . Perhaps a larger sample from more sites in Chiriqui, or from sites with deeper stratified refuse, will eventu? ally permit the detailed refinement of the plain wares, but for this study they receive secondary treatment even though the percentages were always calculated and actually show some interesting trends in the site seriation charts. There are also some small samples of pottery types that fit into the "Classical" Chiriqui wares known for so many decades from the large museum collections; they are listed under their well-known published names. The pottery type description follows the style established in earlier Smithsonian publications on the archeology of South America, with each type subdivided into the major categories of paste, surface, form, decoration, and chronological position. T o these I have added one other category: geographical distribution and comparative materials. For the color identifications of paste or surfaces, words rather than the Munsell Color Chart have been used, for no more precise a code is needed to describe pottery with only simple paints, no polychrome, and variable firing. Hardness is stated in the Mohs scale. It is perhaps interesting to record how the rim profiles were drawn. T o facilitate and increase the accuracy of the rim profile drawings, a section from each rim was cut with a blunt-edged carbide geological power saw at right angles to the lip. Unless the r im was badly mutilated, I always cut a section on the right side of the r im exterior so that, when drawn, all r im profiles faced in the same direction. By cutting the r im it was easy to trace the correct profile because it lay flat on the paper. A second method of analysis used in this report, first separately, and then in conjunction with the type approach, is a modal analysis of all appendages (supports and handles) found in each level of the pits. I have chosen to use " m o d e " to mean " a small group of inseparable attributes" because this is useful in making geographical comparisons. If, for example, a tripod foot shares a cluster of salient attributes with a tripod foot in another area, one can be reasonably certain that some kind of historical relationship, ra ther than mere chance, accounts for the similarities. In this study, modes have been kept separate from types by using different, though not mutually exclusive, criteria for classification. The main criterion used in the type approach is surface decoration; in the modal approach, the main criterion for distinguishing modes is shape. These criteria could just as well be reversed in another area. The result is that modes and types often crosscut, so that the same mode may appear in more than one type. This may have interesting cultural and chronological implications. The potters, conforming to standardized stylistic patterns (types), were able to isolate features (modes) in this pattern and reproduce them in another context. Chronologically, a mode that crosscuts two types may sometimes indicate con? temporaneity as, for example, with Tripod Foot Mode e, which appears in two different but contemporaneous types. A separate modal approach, based on shape elements rather than on painted motifs, is particularly suited to an analysis of the Gulf of Chiriqui collections because this area belongs to a primarily plastic tradition. Paints and slips were used, to be sure, bu t vessels of the Gulf were not elaborately painted as they were in the Cocle-Azuero region. Ra ther , PREFACE it is the shape of handles, supports, and rims, and the manner in which these were decorated with plastic motifs, such as applique strips, incisions, and punctations, that make Chiriqui pottery so enormously varied. Using a modal analysis had advantages of a very practical nature. Handles, feet, and supports are so distinctive that they can be recognized easily in any context. Since the surface collections (Ranere, Appendix 2) were made without any regard to randomness, they contain a disproportionate number of such fragments. A modal approach serves to relate these to the excavated collections. The distribution of certain modes beyond the Chiriqui Gulf has also offered important insights into the influences that came there from Costa Rica and from the central provinces of Panama. In short, modes were used in this study as timemakers and as indices of geographical distribution. From them we could draw inferences concerning wider cultural contacts. I have used several criteria to interdigitate the site sequences and to mark off units of time into phases: (1) the appearance of new types in the stratigraphic columns; (2) marked contrasts in the percentage frequencies of pottery types; (3) the presence or absence of particular modes; (4) the association of types; and (5) the presence of trade wares from known areas. The use of these criteria was facilitated by the deep stratified sequence of Pit No. 3 at Site IS -3 . A very distinctive type of pottery was strongly represented in the bottom levels, different types were associated with the intermediate levels, and the most recent levels were marked by the appearance of several new types. In addition, modes were absent from some levels and present in others, and trade pottery from several phases of the Cocle-Azuero area came in at different levels and served to confirm my phase divisions. The other three sites, none of which had the complete sequence represented, were therefore interdigitated with reference to IS -3 . The assumption made was that all four sites were close enough geographically to show the same general chronological trends. Contrasts in the percentage frequencies of pottery types proved useful in coordinating pit levels within a site and in differentiating between units of levels in which certain types were heavily represented from units in which they were not. The pottery types are described by phase for easy reference, moving from the oldest to the most recent in the sequence. The modes are described separately. Stone tools and shell ornaments found at these sites were not used in establishing a tentative chronology because they did not show enough changes through time. Along with bones and mollusks, however, they are essential to our cultural reconstructions. A description of surface collections from 26 sites in the adjacent districts of San Felix and Remedios has been appended (Ranere, Appendix 2) to the body of the report, which describes our excavations, to enlarge the scope of this analysis. In this appendix Mr. Ranere has contributed ethnohistorical data and new insights into the archeological material, making his section an important part of this work. I would like to thank first Dr . Charles R. McGimsey I I I , principal investigator of the project, for inviting me to collaborate in the field work and for asking me to report and interpret the results. The success of the expedition was also due to Mr. William Bishop and Mr. Freeman Mobley, who were indefatigable in their search for sites and acute in their observations of the Panamanian countryside. The authorities of Panama, as well as Dr. Alejandro Mendez, Director of the Museo Nacional, and Dra. Reina Torres de Araiiz, Professor of Anthropology at the University of Panama, greatly supported the work by their kind cooperation. The entire survey was facilitated by cartographic aids provided by Ing. Amado Araiiz of the Subcomit6 del Darien and of the Pan American Highway Program. Ample laboratory space to analyze the collections was generously provided by Professor John Otis Brew, Director of the Peabody Museum of Archeology and Ethnology at Harvard University. A grant from the National Science Foundation and a Thaw Fellowship in the Peabody Museum permitted survival during the summer and fall of 1963, when the first, unrevised version of this monograph was being prepared as a doctoral dissertation (Linares, 1964). PREFACE Drs. George Cowgill, James Giffo-rd, and John Ladd were knowledgeable in their advice on how to handle particular problems in the analysis. Mr. Anthony J . Renere worked with me on the Chiriqui collections and expanded this study in his Appendix No. 2. Mrs. Alexandra W . d e Garcia-Brycfe advised me on the latest sources for Latin American Archeology. Miss Margaret Currier of the Peabody Museum Library was helpful in the initial research. Mr . Arsen Charles took the photographs on plate 14. From Professor Gordon R. Willey of Harvard University^ I have received endless encouragement through the years and haost of my knowledge of New World Archeology. My intellectual debt to,him is enormous. Expert advice was provided in 1964 by several investigators of the Museum of Com? parative Zoology, Harvard University. Dr. W. J . Clench, Curator of Mollusks, guided m y identifications of the shellfish material; Dr. E. E. Williams, Curator of Reptiles and Am? phibians, sorted out at a glance any reptiles represented on a mass of tiny" fragments. All mammal bones were identified by Mr . Charles Mack. Mr. William Bishop, geologist for the expedition, made the rock identifications while in the field, and the material identifications in the descriptions of stone artifacts are all his; his report on the geology of the coast, from which Mr . Ranere and I have quoted exten? sively, has been of additional value. Mrs. Charles R. McGimsey did the field cataloging of all materials collected and took the photographs of stone artifacts for plates 19 and 20. Dr. Clifford Evans and Dr. Betty J . Meggers of the Smithsonian Institution have spent countless hours over the last years teaching me to classify pottery, making suggestions about plans of research, and helping me in every phase of preparing this manuscript fot publ ica? tion. I cannot thank them enough for their generous help. Dr. Alexander Wetmore, Research Associate of the Smithsonian Institution, provided various maps from his vast collections that were the basis for the final maps produced in this report. To Mrs. Marcia P. Bakry, Scientific Illustrator of the Office of Anthropology, Museum of Natural History, I wish to express my appreciation for preparing the final copies of the maps and seriation charts. I have profited greatly from conversations with Dr. Claude F. Baudez of the Musee de r H o m m e , regarding his archeological investigations in nearby Costa Rica. Recently, Dr. Roberto de la Guardia of the Museo Chiricano in David, Panama, brought me up to date on finds in Chiriqui. In typical Latin American style, my work became a true family enterprise. Sr. Frank Linares Danz, my father, loaned a jeep to the expedition, and Sr. Guillermo Tribaldos, J r . , my grandfather, made his boat available to us. Sra. Olga Tribaldos de Linares, my mother, did the drawings of the artifacts, prepared an original version of the pottery type fre? quency charts, and typed this report twice, first as a dissertation, then as a manuscript for publication. Their interest in my work is immensely gratifying. I also thank my husband, J . David Sapir, for his patience and continuous encouragement. O . L. de S. Department of Anthropology University of Pennsylvania Philadelphia March 24, 1967 Contents INTRODUCTION 1 Geography of Chiriqui 1 . History of archeological work in Chiriqui 4 T H E SITES . 7 Isla Palenque (IS-3) . . 7 Location ; . . . . , . . , . 7 ' Excavations 7 T h e pottery type sequence 9 El Cangrejal (SL-1) 10 Location 10 Excavations 10 T h e pottery type sequence 11 Villalba (IS 5-7) 12 Location 12 Excavations 13 T h e pottery type sequence 14 Las Secas (IS-11) 14 Location 14 Excavations 15 T h e pottery type sequence 16 ANALYSIS OF CULTURAL REMAINS 17 Burica Phase 19 Isla Palenque Maroon Slipped 19 Unclassified plain wares 23 San Lorenzo Phase 23 Arayo Polished Line 23 Banco Red Line 26 Caco Red Slipped 26 Cangrejal Red Line 27 Castrellon Red Slipped 30 Centeno Red Banded 31 Horconcitos Red Banded 32 Linarte Zoned Red Line 32 Pan de Azucar Red Line 33 Zapote Red Banded 35 Unclassified plain wares 36 Chiriqui Phase 37 Cavada Applique and Red Banded 37 Tar rago Bisquit Ware 38 Villalba Red Streaked ' 41 CONTENTS Page Unclassified pottery types ^^ Alligator Ware '*"' Black-on-Red Negative '*'* Chocolate Incised '*'* Red on White ^^ Red Slipped Bisquit Ware ^^ Unclassified plain wares Appendage modes Supports '*? Tripod feet *^ Mode a ^^ Mode b '^^ Modec 48 Mode e 48 Modef 49 Mode g ^" Mode h 49 Mode i ' ^" Ringstands ^^ Mode a ^0 Mode b, c, d ^1 Pedestal bases 5^ Mode a 51 Modeb 51 Mode c 51 Moded 51 Mode e 51 Modef 51 Mode g 51 Modeh 53 Mode i 53 Modej 53 Modek 53 Model 53 Mode m 53 Strapped feet 53 Modes a, b, c 53 Handles 54 Double handles 54 Mode a 54 Mode b 55 Mode c 55 Strap handles 55 Mode a 55 Mode b 55 Mode c 55 Mode d 56 Round handles 56 Mode a 56 Mode b 56 Ladle handles 57 Horizontal handles 58 Miscellaneous handles 59 Miscellaneous applique motifs 59 Stone artifacts 60 Celts 60 Flakes 60 Manos 61 TABLES Page Metates 61 Net or line weights 61 Pebble polishers 61 Pecking stones 62 Pounding-anvil stones 62 Projectile points 62 Sandstone saws 62 Scrapers 63 Sharpening stone 63 Bone and shell ornaments 63 T H E SEQUENCE, CULTURAL INVENTORY, AND W A Y OF L I F E OF THE G U L F OF CHIRIQUI INHABITANTS 65 T h e sequence 66 Burica Phase 66 San Lorenzo Phase 66 Chiriqui Phase 66 T h e cultural inventory and way of life 68 Burica Phase 68 San Lorenzo Phase 68 Chiriqui Phase 72 EUROPEAN CONQUEST AND ASSIMILATION OF THE INDIANS OF CHIRIQUI 75 Early explorations of coastal Chiriqui 75 17th-century explorations of coastal Chiriqui 78 Main land Indians of Chiriqui 78 Inferences 80 CHRONOLOGICAL SUMMARY AND COMPARISON WITH O T H E R AREAS 83 Chronology of Chiriqui 83 Relationships with neighboring Panamanian areas 87 Chronological links with other areas 89 Final remarks 91 LITERATURE CITED 93 APPENDIX 1 (Tables 1-7) 97 APPENDIX 2 : Analysis of pottery surface collections from the Pacific districts of Punta Burica, San Felix, and Remedios in Chiriqui, Panama. By Anthony J . Ranere . . . 107 PLATES 121 Tables APPENDIX 1 1. Frequency of pottery types from Pit No. 3, Site Isla Palenque (IS-3) 97 2. Frequency of pottery types from Pit No. 1 and Pit No. 3, Site El Cangrejal (SL-1) . . 99 3. Frequency of pottery types from Pit No. 1 and Pit No. 2, Site Villalba (IS-7) . . . . 100 4. Frequency of pottery types from Pit No. 1, Site Las Secas (IS-11) 101 5. Frequency of miscellaneous sherds, necks, supports, handles, and rims at Sites I S - 3 , IS -7 , I S - 1 1 , and SL-1 102 6. Frequency of stone artifacts at Sites I S - 3 , IS-7 , I S - 1 1 , and SL-1 103 7. Animal bones from Sites I S - 3 , I S - 1 1 , and SL-1 105 APPENDIX 2 1. Distribution of pottery types in the district of Punta Burica 117 2. Distribution of pottery types in the districts of San Felix and Remedios 118 3. Distribution of modes (handles, supports) in the districts of Punta Burica, San Felix, and Remedios 118 4. Distribution of plastic decorative motifs in the districts of Punta Burica, San Felix, and Remedios 119 xii ILLUSTRATIONS 10 5. Number of sherds in each pottery type of the test excavation at Site BA-8 in the Punta Burica district Illustrations FIGURES 1. Map of Panama, showing area of research in the Gulf of Chiriqui in relation to selected sites in the Cocle-Azuero area 2. Map of the central part of the Gulf of Chiriqui, showing location of excavations and geographical features 3. Site map of Isla Palenque (IS-3), showing location of excavations and topographical features 4. Pottery type sequence of Pit No. 3, Site Isla Palenque (IS-3) 5. Site map of El Cangrejal (SL-1), showing location of excavations and topographical features ' ^ 6. Pottery type sequence of Pits No. 1 and No. 3, Site El Cangrejal (SL-1) 12 7. Map of Isla Villalba (Isla Muertos), indicating location of excavated Site IS-7 and location of surveyed Sites IS-5 and IS-6 13 8. Pottery type sequences of Pits 1 and 2, Site IS-7 on Isla Villalba (Isla M u e r t o s ) . . . . 14 9. Map showing Site IS-11 on Isla Cavada of the Las Secas complex of islands, with location of excavations and geographical features 15 10. Pottery type sequence of Pit No. 1, Site IS-11 on Isla Cavada in Las Secas Island complex 16 11. Rim profiles of Isla Palenque Maroon Slipped, thick variety, Burica Phase 18 12. Rim profiles and side views of Isla Palenque Maroon Slipped, thin variety, Burica Phase. 20 13. Burial urns of Isla Palenque Maroon Slipped, thick variety, from the sites of Las Tinajas and Dolega, reported by other investigators 22 14. Rim profiles of Arayo Polished Line, San Lorenzo Phase 24 15. Painted exterior designs and a few rim profiles of Arayo Polished Line, San Lorenzo Phase 25 16. Reconstructed vessels of Banco Red Line, San Lorenzo Phase 26 17. Rim profiles of Caco Red Slipped, San Lorenzo, and Chiriqui Phases 27 18. Rim profiles and exterior of Cangrejal Red Line, San Lorenzo Phase 29 19. Rim profiles and interiors of Castrellon Red Slipped, San Lorenzo Phase 30 20. Painted designs on the exterior surface and rim profiles of Centeno Red Banded, San Lorenzo Phase 32 21. Type sherds of Horconcitos Red Banded, San Lorenzo Phase 32 22. Body wall profiles and decoration on Linarte Zoned Red Line, San Lorenzo Phase. , 34 23. Type sherds of Pan de Azucar Red Line, San Lorenzo Phase 35 24. Rim profiles and designs on the interior of Zapote Red Banded, San Lorenzo Phase. . 3 6 25. Rim profiles and exteriors of Cavada Applique and Red Banded, Chiriqui Phase. . . 38 26. Rim profiles, side views, and adornos of Tarrago Bisquit Ware, Chiriqui Phase . . . . 39 27. Rim profiles, exterior and interior views, Villalba Red Streaked, Chiriqui Phase. . . 42 28. Sherd of Alligator Ware, Chiriqui Phase (especially diagnostic of "Classical Chiriqui Culture") 43 29. Sherd of Black-on-Red Negative, Chiriqui Phase (especially diagnosfic of "Classical Chiriqui Culture") 44 30. Tripod Feet Modes from all three Chiriqui Gulf Phases 47 31. Ringstand Modes of San Lorenzo and Chiriqui Phases 50 32. Pedestal Base Modes of the San Lorenzo and Chiriqui Phases 52 33. Strapped Feet Modes of the San Lorenzo and Chiriqui Phases 53 34. Double Handle Modes of the Chiriqui Phase 54 35. Strap Handle Modes of the San Lorenzo and the Chiriqui Phases 56 36. Round Handle Modes of the Chiriqui Phase 57 ILLUSTRATIONS Page 37. Ladle Handle Modes of the San Lorenzo Phase 58 38. Horizontal Handle Modes shared by the San Lorenzo and Chiriqui Phases 58 39. Miscellaneous Handle Modes from the San Lorenzo and Chiriqui Phases 58 40. Applique motifs of the San Lorenzo and Chiriqui Phases 59 41 . Shell pendants from Las Secas Site ( IS-11) , Chiriqui Phase 63 42. Seriated sequence of selected stratigraphic pits in the Gulf of Chiriqui area, showing ceramic trends and presence or absence or pottery types in each of the three phases . 66 43. Distribution of Appendage Modes in the San Lorenzo and Chiriqui Phases 67 44. Occurrence of stone artifacts in the San Lorenzo and Chiriqui Phases 69 45. Distribution of moUusk remains by site 70 46. M a p of the Gulf of Chiriqui area, showing selected modern town and villages, arche? ological sites on the mainland reported by other investigators, and surface sites in the Burica, San Felix, and Remedios districts 84 47. Chronological chart of regional archeological sequences in western and central Panama, with approximate time periods indicated 88 48. R i m profiles of Balsa Polished, groups 1-3, Burica Peninsula I l l 49. R i m profiles of pottery type Quebrada Baules Brushed, groups 1 and 2, Burica Peninsula I l l 50. Three-Toed-Tr ipod Foot belonging to Scarified Ware of La Goncepcion complex from Site BA-9 in the Burica Peninsula 112 51. Applique figurine from Site BA-9, Burica Peninsula, resembling figurine from Aguas Buenas complex in Costa Rica 113 52. Polychrome trade pottery at Site BA-8 114 53. Double Handle Mode from Site S F - 1 , San Felix district; a Chiriqui Phase marker . . 114 54. Decorative Modes from the districts of San Felix and Remedios 115 55. Appendage Modes in sites of the Burica Peninsula, associated with Isla Palenque Maroon Slipped, Burica Phase 116 PLATES 1. Typical estuaries south of David, showing water level at low tide. 2. Typical mangrove vegetation on western shores of the Estero de Horconcitos. 3. Views of islands in the Gulf of Chiriqui. 4. Views of fields with vegetation cut and burned for maize planting in the San Felix district. 5. R i m sherds of Isla Palenque Maroon Slipped, Burica Phase. 6. Isla Palenque Maroon Slipped, Burica Phase. 7. Type sherds diagnostic of the San Lorenzo Phase. 8. Vessels of Banco Red Line, San Lorenzo Phase. 9. Cangrejal Red Line vessel and rims, San Lorenzo Phase. 10. Type sherds diagnostic of the San Lorenzo Phase. 11. Type sherds diagnostic of the San Lorenzo and Chiriqui Phases. 12. Type sherds of Linarte Zoned Red Line, San Lorenzo Phase. 13. Type sherds diagnostic of the Chiriqui Phase. 14. T rade pottery from other Panamanian regions, and sherds of Classical Chiriqui pottery. 15. Tr ipod Feet Modes of the San Lorenzo and Chiriqui Phases. 16. Tr ipod Feet Modes of local and trade origin, Chiriqui Phase. 17. Strapped Feet and Pedestal Base Modes and Ringstands, diagnostic of the San Lorenzo and Chiriqui Phases. 18. Handle Modes of the San Lorenzo and Chiriqui Phases. 19. Stone artifacts of the San Lorenzo and Chiriqui Phases. 20. Stone artifacts of the San Lorenzo and Chiriqui Phases. Cultural Chronology of the Gulf of Chiriqui, Panama Introduction G E O G R A P H Y OF CHIRIQUI Within the small crescent-shaped Isthmus of Pan? ama, which is approximately 75,500 square kilometers, there are varied climatic and vegetational zones. Despite this variety, the deep jungles and tropical savanna, the lowlands and the highlands of Panama can all nonetheless be considered variations of an essentially tropical environment. The province of Chiriqui covers 8,960 square kilo? meters in the southwestern comer of Panama (fig. 1). In contrast to the lowland tropical forests of the neighbor? ing northern province of Bocas del Toro, it offers two, rather than three, of the prevalent Panamanian climatic zones: highlands with appreciable plateaus and valleys and broad coastal plains with savanna vegetation. The highlands are formed by the continu? ation of the Talamanca Range from Costa Rica into the western part of the isthmus, where it is called the Continental Divide. This mountainous backbone extends across Chiriqui and Veraguas, stopping at a point northeast of the town of Penonome in the province of Code . In the westernmost section it is capped by deposits from Miocene and Pleistocene volcanoes. The now inactive Volcan Baru in Chiriqui is 3,470 meters high, the highest spot on the isthmus; the rest of the Continental Divide averages 1,000 to 2,000 meters in height. Mountain flanks and plains are composed predominantly of Oligocene, Miocene, and Pleistocene marine sediments plus recent alluvium (Bishop, 1961). The mountains of the Continental Divide determine to a large extent the amount of rainfall and conse? quently the type of vegetation found on either side of the western section of Panama. Warm, rain-filled clouds formed over the Caribbean hit against the Continental Divide and drop their waters along the 284-757 O 68?2 Atlantic coast. There are places in the province of Bocas del Toro where the mean average rainfall a year is 3,500 millimeters. The Pacific coast of Chiriqui is drier than the Atlantic coast; this is true of all of the isthmus as well. Along Chiriqui, through the southern part of the provinces of Veraguas and Code, runs a coastal belt of semideciduous forests and savanna, varying between 5 and 40 kilometers in width. The rainfall here av? erages 1,500 to 2,500 millimeters; it is concentrated in the marked rainy season extending from May to December. The combination of sufficient rainfall and temperatures in the eighties turns the coastal plains of Chiriqui into Panama's richest agricultural lands. In addition, there are large rivers and innumerable streams to keep the land well watered even during the dry season. From descriptions by the chroniclers of the 16th century, one can be reasonably certain that the rich coastal savannas of Chiriqui existed in aboriginal times. The body of water south of the coastal plains is called the Gulf of Chiriqui (fig. 2). In the central part of this Gulf, some 16 kilometers south of the city of David, there is a complex embayrnent of estuaries stretching in an east-west direction for 40 to 50 kilometers. This embayment is formed by the drainage into the sea of several large rivers that flow down from the Cordillera Central (Bishop, 1961). Most important among these are, from east to west, the Rio Madronal, the Rio Chorcha, the Rios Gualaca and Esti, the Rio Chiriqui, the Rio Cochea, and the Rio David. Minor rivers flowing directly from the skirts of the Volcan Baru into the Gulf are the Rio Piedra, the Rio Chico, and the Rio Platanal. 1 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 Punia Mariato FIGURE 1.?Map of Panama, showing area of research in the Gulf of Chiriqui in relation to selected sites in the Cocle-Azuero area. Re? search area is shown in hachure. Great estuaries are formed at the mouth of these rivers; through them flow sluggish, semibrackish waters that are constandy depositing fine silt and clay in the bordering swamps (pi. la). Today, these mosquito-ridden swamps are uninhabited by man, but the calm waters that border them are rich in fish, and the adjacent mud flats teem with mollusks and shellfish. They may have provided fishing and gather? ing grounds for people who lived elsewhere (pi. \b). The outer fringes of the estuaries are protected from the action of tides by numerous islands and sandbars. When the tide is low, great expanses of mud and sand are exposed as tidal flats and beaches. Only temporary encampments of fishermen dot these beaches today. Judging from the absence of archeological sites, there were no large prehispanic occupations in this area either. One of the largest estuaries of the Chiriqui Gulf is the Estero de Horconcitos, formed by the rivers Chorcha, Madronal, and Corrales. On the western shores of its mouth there are only swamps and man? groves (pi. 2), whereas high savanna lands and some hills, the largest of which is Cerro Pan de Azucar, render the eastern shores habitable by man. The site of El Cangrejal (SL-1) was found in these savannas (fig. 2). The Estero de Horconcitos drains into the protected Bahia de Muertos, flanked on the west by the islands of Sevilla and Mono, on the north by the low savannas of the mainland coast, on the east by the higher savanna lands near the town of Horconcitos, and on the south by the large island, Isla Brava. O n the western corner of this bay is Villalba (more commonly known as Isla Muertos), where Sites I S - 5 , 6, and 7 were found (fig. 2). Two different channels can be taken from the Bahia to reach the open sea and oflFshore islands. Boca Brava runs between the easternmost t ip of Sevilla Island and the westernmost tip of Isla Brava. Heavy surf and treacherous sandbars make this a dangerous route to navigate in any type of craft. The channel that curves around the eastern part of Isla Brava and passes south of Isla de Cedro is a far more navigable route. O n the mainland facing it is the village of Boca Chica, connected with the larger town of Hor ? concitos by a narrow dirt road. Once the open sea is reached by way of either of the above channels, the waters become choppy and the winds strong. Solitary islands and groups of islands dot the sea at varying distances from the mainland. Isla Palenque, where we found one of the most im? portant sites ( IS-3) , is just off the southernmost ex? tremity of Isla Brava. Farther to sea is a complex of WHOLE VOLUME INTRODUCTION ?CERRO = ^>../""" PUNTA B a r r i l e s .v..>-^"/< 'J^o, ... S ^""^-'^ ? = Modern towns A = Sites excavated ? - Site reported by other invest igator ^. / C i. ^?*"^\IA,5 jn^-r^ 2.00? f "???"/, 2 0 0 0 n;\.., - Palenque i^San Panda Jose Islands GULF OF C H I R I Q U 8 ? 0 0 ' Las Secas iS - l h Jfe^S' \ Cavada | 3 0 ' 115 8 2 ? i 0 0 FIGURE 2.?Map of the central part of the Gulf of Chiriqui, showing location of excavations and geographical features. SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 islands called San Jose. About 10 kilometers south of Isla Brava is found the very large and heavily wooded island called Parida, surrounded by small islands and tiny rock promontories. The composition of these islands has been described in detail by the geologist of the expedition. Bishop (1961, pp. 7 and 8 ) : Rocks composing the islands southeast of the estuary and the mainland east of the estuary are highly deformed, and they appear to be older than the strata of the estuary and the marine plain. They consist of shale, siltstone, sandstone and graywacke. Deformation has resulted in severe fracturing, faulting, and recementation of the rocks. Minor recrystallization has occurred in some instances. The strata have an average strike of N 70? W, with a dip to the southwest ranging from vertical to horizontal. On the islands of San Jose and Parida there are excellent exposures of pillow lava and columnar jointed lavas. Pillow lavas result when the lava is extruded into the ocean from submarine fissures. Between the "pillows" of lava there are deposits of chalcedonic quartz and green chart and other minerals. These minerals result from the chemical reaction between the sea water and the low temperature components of the lava. Columnar jointing results from the rapid cooling of lava when it occurs as a shallow sill. At one location in Parida island the pillow structure could be observed to grade into columnar jointed andesite. Much of the strata of these islands were sedimentary. The close association of the pillow and columnar jointed lava with the sedimentary rocks would suggest that the lava was intruded into the sedimentary strata during or shortly after their deposition. The vulcanism which gave rise to these lavas was probably associated with the severe deformation which produced the present physiography. If the deformation was Recent in age it would explain why this area of Panama possesses a submergent appearing physiographic character whereas the marine plain to the west and north definitely establishes a recent emergent history. The shore line of these islands consists of narrow boulders, gravel, or sand beaches and vertical sea cliffs. Far out to sea, about 40 kilometers from the island of Parida and 24 kilometers from the mainland, is the island complex of Las Secas, formed by two large and three small islands (fig. 2). Site IS-11 was found on Isla Cavada, one of the two large Secas Islands. T o reach the site one must land on a narrow beach that cuts across the island and climb for about 100 meters on the steep sides of the nearest hill to the west, appropriately called Cerro Las Huacas. At the time of this survey, the U.S. Geodetic Service had cut a faint trail through the deep vegetation to its lookout tower on the top. Long distances separate Site IS-11 from the other sites in the Gulf. The mainland Site SL-1 is roughly 55 kilometers away, while Isla Palenque (Site IS-3) is approximately 37 kilometers to the northwest. T o reach Isla Villalba (Sites IS 5-7), one must navigate around the bend of Isla Brava through the straits of Boca Chica for at least 53 kilometers. HISTORY OF ARCHEOLOGICAL WORK IN CHIRIQUI In the year 1838, a native of the town of Santiago de Alanje, together with a companion, found a large graveyard somewhere in the Cordillera de Chiriqui (Lothrop, 1919). Excited by this find he persuaded the famous General Morazan to organize a party and return to exploit the graveyard, but they were never able to relocate it. This is the first documented instance of what later became a lucrative sport among Pana? manians and foreigners alike, "la huaqueria ." No more was heard of large Indian cemeteries until August of 1859, when the town of Bugavita had a gold rush of its own. In an effort to clear a garden plot, a man pulled a tree from the ground and found adhering to its roots spectacular gold ornaments, more of which lay in the ground below. The news of the treasure spread fast, and within a week the natives of Bugavita had dug more than 225 pounds (or 102 kilos) of golden objects (op. cit. p. 35). Only a few of these objects survived in museums; the bulk were melted down and sold as raw gold. The looting of graves continued through the years. Collectors became more eclectic in their tastes, incorporating into their collections not only gold images but also pottery vessels and stone artifacts. Finally thousands of Chiriqui pots found their way into museums. One of the largest collections of Chiriqui antiquities in existence was made by an indefatigable explorer, J . A. McNeil, in the 19th-century style. H e personally supervised the opening of thousands of graves; the contents were sent to the U.S. National Museum, where they were classified later by Will iam H. Holmes (1888). Holmes was the first person to publish an exhaustive classification of Chiriqui pottery. H e divided the McNeil collection into two plain wares and nine painted wares. Today these wares are called " types" : groups of vessesls sharing clusters of associated features. Holmes said of the criteria he used to separate wares (Holmes, 1888, p. 84): The characters upon which the classification is based are some? what heterogeneous and include material, color, shape, finish, ornamentation, method of manufacture, and evidences of use. With few exceptions, the criteria used by Holmes to separate wares are perfectly sound, though limited. WHOLE VOLUME INTRODUCTION They are purely descriptive criteria, devoid of chron? ological information. Holmes recognized the limi? tations of his classification. There was little he could do to extract chronological information from the McNeil collection because "Unfortunately our ob? servations in the field are not sufficiently accurate to enable us to utilize associations or methods of oc? currence in the graves as a means of classification" (op. cit. p . 84). There is a sort of pseudochronology in Holmes' thoughts about the evolution of art forms. He was an exponent of the notion that art underwent a necessary process of simplification, from representational forms that copy nature and are hence ideographic, to art forms that are geometric and nonimitative because they are produced by mechanical agencies. He did not, however, let this belief interfere with the ex? cellence of his detailed observations of Chiriqui antiquities. A second work in classification, this time a monu? mental volume, was written by George Grant Mac- Curdy in 1911. It is based on the large Chiriqui collections of Yale University and on sizable private collections, such as those of George G. Heye and Minor C. Keith of New York. MacCurdy (1911, p. 47) respected Holmes' classi? fication, but recommended two changes of nomen? clature: changing Holmes' name Terra Cotta or Bisquit Ware to Armadillo Ware, and Holmes' Tripod Ware to Fish Ware. More fundamental is MacCurdy 's addition of three new wares: the Salmon- Colored and the Handled Ware (added as subdivisions of the Unpainted Group) and the Chocolate Incised Ware (added to the Decorated Group of wares). MacCurdy noticed that the Scarified and Chocolate Incised Wares differed from the rest of the wares, but did not attribute this difference to time or trade. Since his time, the Scarified Group has proved to be a Formative period pottery complex preceding by many hundreds of years the other Chiriqui Wares. The Chocolate Incised Ware is either a Costa Rican trade ware found in Chiriqui or a local imitation of a neigh? bor's product. The Polychrome Ware of both Holmes and MacCurdy is definitely trade from the Cocle- Azuero region. A third publication on Chiriqui ceramics, this time a short but very valuable article, was written by Cornelius Osgood in 1935. Osgood took a second look at the Yale collections and, while passing through Panama, inspected large private collections in Boquete, Chiriqui. Of the Holmes-MacCurdy combined total of 14 wares, Osgood retained only 4 wares that accounted for at least 95 percent of all the Chiriqui pottery in the collections he studied. These are Armadillo-Terra Cotta Ware, Fish-Tripod-Handled Ware, Lost-Color Ware, and Alligator Ware. The rest he explained as trade items or as problematic pieces. He traced the Polychrome Ware to C o d e ; Red Line Ware, White Line Ware, and Chocolate Incised Ware to Costa Rica. The origins of the Serpent-Black Incised Ware, the Maroon Ware, and the Scarified Ware, he left as problematic. With admirable concise? ness, Osgood listed the salient characteristics of the four Chiriqui Wares that he retained and plotted their known distribution. The majority of them were found in Bugavita, Divala, and in an area 30 miles west-northwest of David. Wolfgang Haberland of the Hamburg Museum has split some of Osgood's groupings and included others, emerging again with 14 wares belonging to what he calls the "Classic Chiriqui" complex of A.D. 1200 to 1500 (Haberland, 1958). Following is a list of Haber- land's major revisions: a. He split the Lost Color Ware into four groups: Negative Ware, Black on Cream Negative, Black on Red Negative, Red on Red Negative. b. He reinstated the Red Line and Chocolate Incised Wares, plus the Black-Incised or Serpent Ware (which he renamed Punched Band Ware) , as Classic Chiriqui pottery types. Osgood had discarded them as being of Costa Rican origin. c. He eliminated the Scarified Ware from the list because it is older than the Classical Chiriqui pottery types. d. He added four new wares to the list: Red on Orange, Redbrown Applique, Black Ware, and U n ? painted Ware. The value of all the distinctions made by Haberland cannot be ascertained from the present study. Two of the phases I establish for the Gulf of Chiriqui precede in time his Classic Chiriqui complex. The most recent phase in the Gulf is Classic Chiriqui, but it is repre? sented by only two of the popular types associated with this period on the mainland: Bisquit (or Arma? dillo) and Fish-Handled-Tripod Ware (renamed Villalba Red Streaked by me, called Red Painted Ware by Haberland). The absence of other "Classic Chiriqui" types in our most recent phase may be due to the fact that Negative, Red on Orange, Choco? late Incised, and Red-Brown Applique are typical mostly of the Highlands of Chiriqui (Haberland, 1958, p. 345). Splitting, regrouping, and adding new wares to the Chiriqui complex reflects the obvious fact that little is known of the chronological or geographical sig? nificance of the Chiriqui ceramics. This is a situation which may be remedied in part by the present study. The Sites Four archeological sites were investigated in the Gulf of Chiriqui. The location of each site and its physical appearance are described, and excavation procedures at each pit are reported as fully as possible. The pottery type sequence yielded by these excava? tions is summarized following each site description, although pottery type descriptions constitute a sep? arate chapter. Cultural remains other than pottery are included in Appendix 1, tables 6 and 7. ISLA PALENQUE (IS-3) Location Palenque is a small island, 2.4 by 1.6 kilometers, located at 82? 15' west longitude and 8? 10' north latitude, immediately south of Isla Brava. These two islands are separated by a shelf less than a kilometer in length and only one fathom deep. Isla Palenque is surrounded by waters of relatively shallow depth (1)^ to 2% fathoms). It is not, however, in a protected location as are the islands in the Bahia de Muertos. Like the islands farther out to sea, it receives the full impact of tides and winds on its southern side. Although neighboring island complexes may be far away, they are easily accessible from Palenque. Isla Villalba is 20 kilometers away and LasSecas is 37 kilometers out to sea. To reach the mainland site of El Cangrejal (SL-1), one must navigate for 22 kilometers. In aboriginal times there must have been continuous contact between the inhabitants of all of these islands. Canoes, and perhaps even rudimentary sailboats, probably traveled back and forth much in the fashion that the scattered fishermen of today visit each other in crafts essentially unchanged since prehispanic days. Site IS-3 is located near the center of the island on a hill 50 meters high (fig. 3). The archeological site is recognizable by four closely spaced artificial mound'" built up in part by the deposition of one to two meters of cultural debris (pottery, bone, stone artifacts, and shell) on a naturally raised ridge one meter high. The ridge is about 100 meters long by 20-25 meters wide and runs northeast-southwest. Another ridge of the same height runs off the west end of the first ridge for about 50 meters. There are also visible middens on this second ridge. The area between the ridges is strewn with scattered potsherds. Judging from excavations at Pit No. 1, located 40 meters to the north of the ridged area, the depth of the deposit is a meager 20 cm. in the flat area surrounding the mounds. The mounds are covered with grass, while the sur? rounding area is heavily forested. We were informed that there is a freshwater supply from a sizable stream on the island, although we did not determine its distance from the site. Excavations Four pits were excavated at Site IS-3 (fig. 3) but only the largest. Pit No. 3, has been included in this report because it is the only one with considerable depth and cultural refuse. This pit, placed on the farthest mound at the northeast corner of the ridged 7 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 Pi t -3 / Old lake bed ? P i t -4 I I I 0 15 20 M FIGURE 3.?Site map of Isla Palenque (IS-3), showing location of excavations and topographical features. area, measured two meters north-south by one meter east-west. The arbitrary levels of 10 cm. in thickness were measured from the east wall of the pit, which was one level higher than the west wall. No natural stratigraphy was visible at the time of the excavations. The following description has been reconstructed in retrospect from field observations on the physical appearance of each level as it was being excavated, which permit subdivision of the deposit into sections. These sections form natural units in terms of the consistency, the color, and the contents of the soil. In addition, these units often correspond to sharp breaks in the amounts of cultural materials present in the levels, though not necessarily to the phase divisions as drawn in this report. Description follows the order of excavation, that is, from top to bottom. LEVELS 0-30 cm.?Soil was dark brown, soft, porous humus, and with a slight silt content. In the uppermost level there was a large quant i ty of frag? mented rocks, two or three fists in diameter, plus many roots. At 30 cm. the soil turned slightly more yellow, but kept its soft consistency. In the two top levels (0-20 cm.) the amount of cultural material was impressive: a gunny sack of potsherds per level, plus several small bags of stone artifacts. Some bone fragments, bu t no shell, were also present. A fractured vessel was found in Level 10-20 cm. I t was laid, face down, 20 cm. from the east wall of the pit and 45 cm. from the south wall. In Level 20-30 cm. the amount of pottery decreased sharply to about one-half that of the upper levels. This sudden decrease may represent a temporary abandonment of the site by some of its inhabitants. WHOLE VOLUME THE SITES The three top levels correlate with the third and last phase of the pottery sequence at I S - 3 . The 30 cm. depth marks, therefore, a cultural change in the deposit. LEVELS 30-90 cm.?At 30 cm. the soil turned slightly lighter and more compact in texture due to a higher clay content. This trend continued to a depth of 90 cm. where the soil turned definitely hard and compact. Fractured stones were numerous, although less abundant than in the top levels. At 90 cm. a type of laminated shale rock, called locally "laja muerta ," first appeared. The amount of cultural material found in the Level 3 0 ^ 0 cm. was almost twice that of Level 20 - 30 cm., confirming the observation that the latter level marked a decline in intensity of occupation. Two nearly complete vessels and one tall pedestal base were found in Level 40-50 cm. One vessel (p. 26) was leaning against the east wall of the pit at a point 36 cm. from the south wall. Next to the west wall and 40 cm. from the south wall lay a fractured but nearly complete shallow tripod (p. 29). Adjacent to it was a tall pedestal base (p. 52). From the next level down we extracted several fractured shallow bowls, one inside the other, 76 cm. from the south wall, very close to the east wall. The presence of nearly complete vessels in three levels of this pit does not necessarily imply that the stratigraphy of Pit No. 3 was disturbed. I t clearly was not, judging from the pottery type sequence (see dis? cussion in the following pages). Furthermore, a pre? liminary sorting of the sherds from Pit No. 2 (not in? cluded in this report) showed the same sequence of pottery types. The unit of levels between 30-90 cm. differs from the preceding unit by having soils that are slightly more compacted due to higher clay contents. Like? wise, it differs from those in the unit below by a relatively looser texture and by the absence of "laja muer ta" (shale). This unit corresponds only in part to the cultural division which, on the basis of pottery type distribution, has been drawn at the 130 cm. mark. LEVELS 90-160 cm.?The soil of Level 90-100 cm. had a burned yellow color due to the large amount of "laja muer ta" (shale). Its consistency was hard and rocky, becoming so compact at Level 110-120 cm. that a small, sharpened shovel had to be used to break the ground. Patches of nearly white soil could be observed throughout these levels, resulting from decomposition of limestone rock. Level 140-150 cm. contained very hard, almost pure igneous rock, weathered buff to tan in color. Preserved in it were exfoliated fractures typical of the exposed bedrock of these islands. Levels 150-170 cm. consisted of deeply weathered bedrock of either igneous or metamorphic origin and were practically devoid of cultural material. The amount of cultural material declined from 1,083 sherds in Level 90-100 cm. to 460 sherds in the next lower level. There was an increase again in Level 110-120 cm. to 960 sherds, followed by a second decline in the next level down. The next two levels contained between 100 and 200 potsherds; only a handful of sherds was found in Level 150-160. The unit of levels 90?160 cm. does not correspond in its entirety to the earliest phase at the site. Only the levels below 120-130 were included in this phase. The Pottery Type Sequence FIGURE 4; APPENDIX 1, TABLE 1 Site IS-3 (Isla Palenque) is a stratified site with a sequence of three phases distinguished from one another by the presence or absence of certain pottery types. The diagnostic types of the earliest phase, the Burica Phase, are Isla Palenque Maroon Slipped which occurs in frequencies of 30 to 40 percent and Plain Ware C which occurs in similar percentages. Levels 130-150 cm. correspond to the Burica Phase. The middle phase, San Lorenzo, is defined by large percentages of pottery types decorated with red lines, red bands, or a red slip: Arayo Polished Line, Banco Red Line, Caco Red Slipped, Cangrejal Red Line, Castrellon Red Slipped, Centeno Red Banded, Horconcitos Red Banded, Pan de Azucar Red Line, and Zapote Red Banded. Four of these types are restricted mainly to this phase; the others also occur in the next phase. The plain wares do not show any significant correlation. Levels 30?130 cm. corre? spond to the San Lorenzo Phase. The most recent of the phases, the Chiriqui Phase, is marked by the appearance of three new pottery types: Tarrago Bisquit Ware, Villalba Red Streaked, and Cavada Appliqu^ and Red Banded. Levels 0-30 cm. correspond to the Chiriqui Phase. 10 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 EL CANGREJAL (SL-1) Location The Estero de Horconcitos bends east at 8? 20' north latitude to receive the waters of the Rios Madronal and Corrales and to carry them south to the Bahia de Muertos. The archeological site (SL-1) is on the mainland coast by the eastern edge of the Estero de Horconcitos, only some 100 meters from the edge of the water (fig. 2). Its exact location is 82? 13' west longitude and 8? 19' north latitude. T h e inhabitants of the site could have sailed directly south, hugging the eastern shores of the calm Bahia de Muertos to reach the eastern tip of Isla Brava, roughly 12 kilometers away. By sailing another 10 kilometers they could have skirted the tip of Isla Brava and reached Isla Palenque. Instead of sailing all the way, they could have walked south, reached the area where the town of Boca Chica stands today, and from there on sailed to Palenque Island. They could have also easily sailed southwest to reach the northeast tip of Isla Muertos, 10 kilometers away. The site (fig. 5) consists of four roughly ovoidal mounds in a half-moon arrangement. In part they are formed by natural undulations in the savanna; in par t they have been built up by about one meter of cultural refuse. Two of them were sterile of cultural material, but potsherds and other artifacts were found on the surface of the flat, unmounded areas, so that the total site area is more than 200 meters square. The site itself is between a grove of "corozo" palms to the north and the dropoff line to the Estero de Horconcitos on the west. The mound tested is only 16 meters from the edge of the bank that drops off gradually to meet the calm waters of the estero. A brook called Quebrada del Caco runs about 130 meters to the east of the site. About 90 meters to the northwest is a waterhole called locally "El Zapote ." Both of these could have kept the people of SL-1 well supplied with fresh water. Excavations Two small pits were excavated in Mound A, the nearest mound to the bank. The surfaces of Mounds B and C, were found to be sterile. Pit No. 2, on the flat area to the northeast of Mound A, yielded a small number of potsherds. Pits 1 and 3, included in this report, were dug in artificial levels of 10 cm. No natural stratigraphy was observable. Test Pit 1, measuring 1 by 1 meter, was placed on the northern half of the mound. LEVELS 0-60 cm.?The soil was brown in color and slighdy compact, tending to break into very small clods. Well-rounded and fractured small stones occurred throughout, but were especially a b u n d a n t in the top three levels. The amount of cultural material yielded by these levels was uneven. Only 215 potsherds were found in the first level, while the second level (10-20 cm.) had the largest number of potsherds in the entire pit?1,093 sherds. The other four levels in this unit produced between 200 and 500 potsherds each. Fragments of mammal bones and a small number of shells were found throughout. In Level 30^40 cm. there was a concentration of bones, mainly of fish, in a circular area 20 cm. in diameter in approximately the center of the pit. These bones were associated with pieces of carbon and fired earth, suggesting that this had been a small cooking hear th . In the southwest quarter of the pit, a t Level 50-60 cm., a concentration of small univalve shells of an edible variety was encountered. There were a few rocks at the margins of this shell concentration, but the association between rocks and shells may be accidental. The upper 60 cm. of the deposit corresponds to the last, or more recent, of the two subphases postulated by this study for Site S L - 1 . LEVELS 60-90 cm.?At 60 cm. a sharp change occurred from the brown, slightiy compact soil of the upper levels, to a yellowish, hard , and compact soil that tended to break into very large clots. Com? pactness increased until, in the bot tom level, it was a red-brown almost impenetrable clay. The number of potsherds declined from 342 in the preceding level to 110-158 in Levels 60-80 cm. T h e bottom level had only 22 sherds. T h e rare shells found in these levels were often so disintegrated that they appeared as mere white lenses on the soil. This unit of levels corresponds to the earliest of the two subphases established at S L - 1 . Test Pit 3 was placed 3 meters to the south of Pit No . 1 in an effort to increase the sample of the later subphase. I t was begun with the dimensions of two meters east-west by one meter north-south, but heavy downpour forced us to reduce the area to one-half at the third level down. Reducing the dimensions facilitated covering the excavated area with a small tarpaulin and also accelerated the work. T h e sample size in both pits, considered together, is large enough to establish a sequence of deposition for 0 - iO 1 0 - 2 0 2 0 - 3 0 3 0 - 4 0 4 0 - 5 0 5 0 - 6 0 6 0 - 7 0 7 0 - 8 0 80- 90 9 0 - 1 0 0 100-110 liO-120 I 2 0 M 3 0 130-140 140-150 D D D UNCLASSIFIED PLAIN TYPES- I > I ? I I 0 10 20 30 40 5 0 % Scale A D Q 0 D ? ? 3 a D H J I I 0 5 10 15 20 2 5 % Scale B FIGURE 4.?Pottery type sequence of Pit No. 3, Site Isla Palenque (IS-3). D ? I 1 a I ? \ Q 0 a Scale A < - D D D EZD UJ O LU ill CL CL CO o CO 5 d k i Q ] dn m L: nn D D ? D D D D LiJ 0 Q D ? D D Q LU iD CC Q- y^ 04J ocf) < 1 1 Q LU Q 2 < GO LU h- Q 0 UJ a. a: < ??^ Q ^ n 10 R ^ 0 0 a 1 ti D _l 0 LU X CO 0 _I > 0 < Q. cr < Q LU Q < 0 m z UJ Q 1- UJ z cr LU 0 I 0 Q a 0 I ? n n a I 0 D D D D D D Q Q D D D n D CA R 3 UJ NJ Z < ? ^ - I LU Q Q LU Z (T < a_ 2 0 _i _j Ul (r H- cn < 0 Q Ul CL CL _J CO Q Ul cr (O 0 }? 0 ?z. 0 0 cr 0 X Q UJ 0 Z < CD Q LU cr LU -J Z < "i -0 -" UJ cr Q C3 LU z cr < 0 D 0 0 Ul z 0 N LU 1- cr < z UJ z _l Q UJ QC _J en LJ O o O UJ QC < CD LU < O CD < cr cr CD < 3 O CO Q UJ cr < CD 1 < -J Q LU V < U) \x. \- CO UJ 3 0 -1 CL Q- < < Q < > < 0 Q UJ Q Z < CD Q Ul cr (C W H O L E VOLUME THE SITES 11 o ' Ujl CO' UJI K.' / < i / O ' Or/ Q/ \ \ \ \ \ _, ? /^ / y ? / y y / Mound A /?? P i t - I ? ?"''"? \. P i t - 3 / ^ .. ^ 1 / - ? ? ? ? , ????'??? . . . . ( ' . . . , .? ? ^ 2 ^ ^ -Mound^jU /. . \ D / ^ J!^ - ....t. A... 1 5 0 m . to sou thern d rop-o f f t o sea ^ ? ~~-^ - ^ P i t - 2 ' ^ ^- 1. Previous ^ excava t i on / Mound \ NQ 2 B ;? Mound; C 1 1 0 ,x Previous ^ ^ excavation No. 1 N ? 1 8 I6M FIGURE 5.?Site map of El Cangrejal (SL-1), showing location of excavations and topographical features. Mound A, apparently the most heavily concentrated refuse in the area. LEVELS 0-50 cm.?The soil in these levels was dark brown in color and slightly compact like that in Pit No. 1. Abundant roots and some stones made the digging difficult. In contrast with Pit 1 where the upper 10 cm. yielded only a couple of hundred potsherds, the top level of Pit No. 3 had 1,690 sherds, the highest number of potsherds in the pit. Level 10-20 cm. had 1,665 sherds. At Level 20-30 cm. the southern half was left unexcavated; this caused a sharp reduction to 364 sherds in this level. Between 200 and 500 sherds per level were yielded by the other two levels in this unit. A small number of edible shells occurred throughout with no concentration in any one level. The unit of levels corresponds to the latest of the subphases postulated for Site S L - 1 . LEVELS 50-90 cm.?The change from a darker clay to a lighter, more compact clay occurred in this pit 10 cm. above the level where the same change took place in Pit No. 1. Otherwise conditions were the same. The number of sherds in Level 50-60 cm. increased by 200 from the level above. In Levels 60-80 cm. it dropped to ca. 140 sherds; the bottom level had only 41 sherds. The scattering of bone and shell found in the up per levels continued to the bottom. Carbon specks were collected from these levels for carbon sampling. This unit of levels between 50 and 90 cm. corre? sponds to the earliest of the two subphases established for Site S L - 1 . The Pottery Type Sequence FIGURE 6; APPENDIX 1, TABLE 2 El Cangrejal (SL-1) is a single phase site, belonging entirely to the San Lorenzo Phase. Changes in the physical aspect of the soil and the presence or absence of types permits separation of the two pits at SL-1 into subphases: Early and Late San Lorenzo. Two pottery types, Castrellon Red Slipped and Linarte Zoned Red Line, are present in the Late Subphase, but absent from the Early Subphase. The Early San Lorenzo Subphase includes the two 12 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 bottom levels of Pit No. 1: 60-70 cm. and 70-80 cm. (Level 80-90 cm. has not been plotted because it had only 21 sherds.) In Pit No. 3 the Early San Lorenzo Subphase includes the three bottom levels: 50-60 cm. to 70-80 cm. The Late San Lorenzo Subphase includes the upper six levels of Pit No. 1 and the upper five levels of Pit No. 3. The plain sherds of Site SL-1 are difficult to sort into types without a very minute analysis of the paste under a binocular microscope. As with the decorated types, the paste seems undifferentiated. VILLALBA (IS 5-7) Location Isla Villalba, also called Isla Muertos, is a long, narrow island, about 4 kilometers northeast-southwest by only 0.6 kilometers at its widest point. Its exact location is 82? 18' west longitude and 8? 16' north latitude (fig. 2). The island is in the Bahia de Muertos, a bay whose calm waters reach a depth of 2 to 4 fathoms. The contour of the surrounding islands and mainland to the north, east, and west is broken by arms of the sea locally known as esteros. The largest is the Estero de Horconcitos with its two branches that extend far inland to receive the waters of rivers that drain into the bay from the highlands. The mosquito-ridden swamplands along these esteros are not suited for habitation; the esteros are navigable waterways and are good for fishing (pi. 2). The very large Isla Brava safeguards Villalba from the action of the tides. There was probably contact between these islands in prehispanic days. The people of Villalba also navigated around the eastern tip of Isla Brava to reach islands such as Palenque (20 kilometers away) and Las Secas (53 kilometers away). They could not have gone around the western end of Isla Brava as easily because the waters of the channel of Boca Brava are extremely turbulent and dangerous. A series of connecting ridges 120 meters above sea level run as backbones along the center of Villalba (pi. 3 a, b). The sites (fig. 7) are on these ridges. Site IS-5 is on a ridge at the northern end of the island, slightly south of a freshwater stream running east. The site is approximately 50 x 50 meters in a cleared area where the present inhabitants keep pigs and cows. Many sherds are exposed on the beach, where they must have been carried by the action of rain water on the eroding hillside. Only surface collections were made. Site IS-6 is on another ridge just north of the center of the island?about 1.2 kilometers southwest of Site IS-5 . Scattered potsherds were encountered all along the flanks of the ridge, with no concentration of sherds at any one spot. Although mentioned in passing by writers like Holmes, Site IS -6 was first described by Haber land (1960 a, p. 10): The main feature of the site is obviously a low mound, with walls built of boulders, still some 30 cm. above the ground. Two niches were observed, one of which was occupied by the remains of a pillar, . At the time of our visit, six basalt columns tha t originally served as bases for carved figures lay on top of the mound. Judging from their position, they were once arranged in a circle. The figures that were carved on top of these columns have disappeared. T h e former owner of Villalba removed two fairly complete statues to his home. These have also been described by Haberland, who calls them "ornamented stone pillars" because the stone figures at the top are dwarfed by the shafts of the columns. One of the statues has a column measuring about 110 cm.; on top of it is a figure of an armadillo only 15.5 cm. high. Haber land (1960 a, p. 14) describes it as follows: All four legs, the carapace and the head are nicely and typically modelled leaving no doubt as to its identity. The whole figure, which is 9.0 cm. broad and 26.0 cm. long, is set off from the pillar by a small platform at both broad sides, while the head and tail surmount the corners of the base. The second statue, of a female, measures 58 cm. in height, 20.5 cm. in shoulder width, 23.0 cm. across the knees, and 13.5 cm. in depth across the small of the back (Haberland, 1960 a, p . 14): The top of the head is horizontal and the face nearly triangular. The nose is prominendy modelled, with a wide lower part. The eyes appear nearly closed, the slits were made very low in the half-rounded bulges which are supposed to be the lids. No mouth is indicated. The body is somewhat angular and rough, with prominent buttocks if seen from the side. The upper arms are pressed along the sides of the body, while the short lower arms lie horizontally and meet one another in the middle of the belly. Hands are not especially indicated. If there are legs, they are only faintly discernible and would indicate a squatting position. By comparison with the other figures it is more possible that this is the remnant of a girdle and that the figure was standing. Two small breasts indicate the female sex. I - Q- 0 - 10 CM 10-20 2 0 - 3 0 3 0 - 4 0 4 0 - 5 0 5 0 - 6 0 6 0 - 7 0 7 0 - 8 0 4 5% ro 0 - 1 0 CM 10-20 2 0 - 3 0 3 0 - 4 0 4 0 - 5 0 5 0 - 6 0 6 0 - 7 0 7 0 - 8 0 o UJ CO z O) ? o Q- z D D ? D CH D ? D D D 1 J a ? 1 1 1 1 ? d ] 1 1 Q UJ Z O UJ N 2 N A R T E R E D L 1 A P O T E N B A N D E D R E D o UJ L L ON SL IP P A S T R E R E D o Q 0) UJ M CI TO B A N D lO RC O R E D -L. CZl D [ZZI CZl ? IZH cr < o N < UJ O < a. UJ z o UJ a: hi z o UJ X en < Q- < z: o u < o Q UJ Q . a. I J Q Ul ? LiJ ^ ? -3 _J UJ a: Q O UJ z a: < o o UJ o z < O CD z UJ Q h- UJ z oc UJ o FIGURE 6.?Pottery type sequence of Pits No. 1 and No. 3, Site El Cangrejal (SL-1). WHOLE VOLUME THE SITES 13 I s l a ^ Mono Vil lalba (Isla Muertos) Isla Manglar Muertos FIGURE 7.?Site map of Isla Villalba (Isla Muertos), indicating location of excavated Site IS-7 and location of surveyed Sites IS-5 and IS-6. Only surface collections of sherds were made at Site IS-6 . One bag was collected from the potrero at the bottom of the ridge, another from the hillside, and a third from the area of the basalt columns. We found no concentration of refuse. Instead, there were depressions in the ground in the shape of graves. Site IS-7 is 90 meters north of Site IS-6 on the same ridge. It covers a small area, 10 to 15 meters in diameter, where shell and potsherds appear abun? dantly on the surface. Excavat ions Two test pits were dug 10 meters apart on the Villalba ridge at the site of IS-7 in arbitrary levels of 10 cm. A summary of the most important physical and cultural changes in these pits has been made by grouping levels into descriptive units. These will be presented from top to bottom in the order of excavation rather than in the order of deposition. Test Pit 1 measured 2 meters north-south by 1 meter east-west. It was placed a meter away from a large hole made previously by someone unknown; it showed the largest concentration of sherds in any one spot at the site. LEVELS 0-50 cm.?The soil was very dark, loose porous humus and contained many roots, especially in the upper 20 cm. Fragments of unworked stone, larger and more rounded toward the top, angular and smaller toward the bottom, were plentiful. Shells were most abundant at the top level; slightly fewer were found in each of the levels below. Sherds were very numerous between 30 and 50 cm. The pottery types in these levels belong to the Chiriqui Phase. LEVELS 50-70 cm.?The soil became progressively harder in consistency and lighter in color. Fragments of unworked stone became scarcer at 40 cm. and disappeared altogether at 70 cm. No shell was found below 50 cm. The number of sherds decreased sharply from 688 in Level 40-50 cm. to 62 in Level 50-60 cm. The earth below 70 cm. was hard in consistency because of the large amount of "laja muer ta" (shale) found in it. The "laja" also gives it a bright yellow tint. Only a handful of sherds was found between 60-70 cm.; below 70 cm. the soil was sterile. A small hole dug 20 cms. deeper in the center of the pit confirmed that this was the end of the refuse deposit. Level 50-60 cm. was placed in the San Lorenzo Phase. Test Pit 2, located at the side of the raised area, was 1.5 meters higher than test Pit 1, which was in a depression. Dimensions were 2 meters north-south by 1 meter east-west. Because it was on the side of the mound, the northwest corner was 34 cm. higher than the southwest corner. Excavation followed the contour of the land. The total depth of this pit, 140 cm., was almost twice that of Pit 1. LEVELS 0-80 cm.?Soil was dark, loose, and porous and contained many unworked stones. Approximately half a kilo of shells was found in the top three levels, increased to a kilo in the 3 0 ^ 0 cm. level, and de? creased again by one-half in the next four levels. Potsherds were very abundant in Levels 30-40 cm. and 40-50 cm. (1,561 and 864 sherds respectively). The remaining levels in this pit produced only a few hundred sherds per level. Stone artifacts occurred in the three levels between 30 and 60 cm. This unit of levels corresponds to the most recent of the two Chiriqui Subphases established for IS -7 . LEVELS 80-100 cm.?In Level 80-90 cm. the soil began to get slighdy lighter and became hard, with angular shale fragments, at 100 cm. The number of sherds decreased from 195 in Level 80-90 cm. to 121 in Level 90-100 cm. There were no stone artifacts in this segment of the pit. LEVELS 100-120 cm.?Below one meter the soil became light in color and fairly soft in texture. The 14 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 number of sherds suddenly increased to 411 sherds in Level 100-110 cm. and 194 sherds in the level below. Stone artifacts, which had disappeared at 60 cm., re? appeared in Level 100-110 cm. No shells were re? covered in this unit. This and the previous unit of levels correspond to the middle subphase of the occupation at Site IS-7 . These subphases, however, have only local importance. In terms of the entire Chiriqui Gulf sequence, both are subsumed under the last or most recent of the Gulf Phases, the Chiriqui Phase. L E V E L 120-130 cm.?The soil was light in color, like that in sterile levels, but its texture was soft and porous. There were 180 sherds in this level and only two in the level below; for this reason the bottom level does not appear in any of the pottery type frequency charts. Level 120-130 cm. corresponds to the oldest of the Chiriqui Gulf Phases, the Burica Phase. The Pottery Type Sequence FIGURE 8; APPENDIX 1, TABLE 3 Villalba (IS-7) is a two-phase site. The bottom level (120-130 cm.) of Pit No. 2 has been included in the Burica Phase, the earliest of the Gulf Phases. This level contains a large percentage (33.33 percent) of Isla Palenque Maroon Slipped. T h e rest of the levels in Pit No. 3, and all of Pit No. 1 have been placed in the last or most recent of the Gulf Phases on the basis of large percentages of Tarrago Bisquit W a r e and Villalba Red Streaked. T h e middle or San Lorenzo Phase is not represented at Site IS-7 . T h e Chiriqui Phase in turn has been separated into two subphases on the basis of observed changes in the physical appearance of the soil, correlated with the presence and absence or the percentage fluctuations of the pottery types. The older of the two subphases includes Levels 0-50 cm. of Pit No. 1 and Levels 90-120 cm. of Pit No. 2. The more recent of the two subphases at IS -7 in? cludes all the levels of Pit No. 2, from the surface to 90 cm. These subphases are only of local importance. They indicate that the occupation of Site IS-7 began earlier and lasted longer in the locale of Pit No. 2. As far as the popularity of pottery types is concerned, Tarrago Bisquit Ware and Villalba Red Streaked were more popular in the later than in the earlier of the two subphases. LAS SECAS ( I S - l l ) Location Las Secas is a complex of five offshore islands located some 30 kilometers (about 5 marine leagues) south of the point on the mainland where the Rio San J u a n empties into the sea (fig. 2). Their exact position is 8? north latitude and 82? 2' west longitude. In contrast to other islands of the Chiriqui Gulf, Las Secas are far away from the mainland, in open sea. All five islands are within an area 8 kilometers in diameter. Their jagged contours produce narrow shores strewn with boulders or sea cliffs rising vertically to join rugged hills 100 or more meters in height. A few small beaches are found hidden away in protected coves. The composition of these and the geology of Cavada itself have been described by Bishop (1961, p. 22): The beaches are composed of calcium carbonate sand which is derived from shell and coral fragments supplied to the beach from the adjacent shallow waters which abound with marine fish. The islands are composed of sedimentary rocks which appear to have been derived from poorly weathered volcanic material. The most abundant lithologic types seem to be rocks that grade from graywacke to subgraywacke. Las Secas are surrounded by shallow waters, so clear in places that they are almost t ransparent . Underlying them are coral reefs teeming with fish. Stretches of soft muddy banks also abound with mollusks of all kinds. It is easy to dive for them from the shore or from any boat anchored nearby. The aboriginal inhabitants of these islands had a rich marine fauna to exploit. Site IS-11 was found on Isla Cavada, the largest and northernmost of the islands (fig. 9). The site occupies the Cerro Las Huacas. I t is easily recognized by the abundan t shell and sherd material strewn over an area at least 200 meters in diameter. Near the northern edge of this area, before the slope drops sharply to the sea, there are three shell middens aligned in a northwest-southeast direction. T h e two outer middens are at least 20 meters in d iameter ; the central one is much smaller. T h e middens are on a naturally raised ridge about 2 meters high. T h e concentration of cultural refuse. 0 - 10 CM 10- 20 20- 30 30 -40 40 -50 50-60 0 10 20 30 40 50% CM I - 0 - 1 0 CM 10- 20 2 0 - 3 0 3 0 - 4 0 4 0 - 50 5 0 - 6 0 6 0 - 7 0 7 0 - 8 0 80 -90 90-100 100- 110 110-120 120-130 I I 1 I I I D D D ? D D D D ? n CZl ? ? n czi CZl D J CZl en ? n a n n en n n n D D D D ? D D n n 1 1 [ZZl ? o Ui 0 ) 2 cn ? z 3 tiJ QC < 0 5 o r OQ < o UJ < Ul en _J UJ _ i tt: Ul Q . UJ Q . ^ z a: < < 0) 0 UJ a z < tD UJ h - Q 0 UJ Q. (C < N 0 CO UJ 0 Q H Z o2 Z O Q 0 UJ a: cr 0 X 0 UJ z 0 N UJ 1 - tE < Z UJ ^ _l 0 UJ cr UJ z L 2 UJ I - UJ o z < OQ O UJ UJ o o UJ a. Q . I J en Q O UJ < o UJ z - 1 0 UJ X CO O - i > o < fl? oe < < 0 < > < 0 0 0 Ixj z o < z < UJ CD 3 O Q _jUJ Q L O : Q. < UJ H X ^ Z 0 0 K UJ * QC cn ? FIGURE 8.?Pottery type sequences of Pits 1 and 2, Site IS-7, on Isla Villalba (Isla Muertos). WHOLE VOLUME THE SITES 15 Pit -2 ???... ? N S \ \ V r About 500m. ? - ~ ^ _ - . to Sea ...??????-, Old { ^ ?? excavation .\^ <^. \M^i - ? - ~ -Drainage ,Old excavation Local s teepening on side / " " > of hill \ . \ Pit-I D / 9^, 9 Oyster' midden Large hil l VCrest of h i l l I \ \ ^ I ^x ^ \ \ ^ Local \ \ \ ^ "^ steepening"" N \ 1 ------ --'-^ / / N / / I ? ? ? I ? ? ? I 0 10 2 0 M FIGURE 9.?Map showing Site IS-11 on Isla Cavada of the Las Secas complex of islands, with location of excavations and geographical features. much of it shell, increases the height of the midden to about 3 meters. The two outer middens are hidden by thorn trees and heavy secondary growth. On the westernmost mound, part of a carved basalt column 60 cm. in height was found, with only human feet remaining carved on the top. The rest of the figure had been broken off. One hundred meters down the hill, on both ends of the beach, there are two freshwater streams. Water must have been carried up the steep sides of the hill to the inhabited area of the site. Excavations Two pits were dug on the crests of the shell middens. Only the largest, Pit No. 1, is included in this study. Test Pit 1 was placed on the crest of the large southernmost midden where the deposit seemed to be deepest (fig. 9). The pit measured 2 meters north- south by 1 meter east-west. As was the case with the other tests, the pit was excavated in 10 cm. thick arbitrary levels. The enormous concentration of shell in the midden would have made it impossible to recognize natural strata if they had been present. For purposes of description, the levels have been grouped into two sections which are naturally and culturally distinct. LEVELS 0-60 cm.?The soil consisted of a dark, soft, sandy clay, which was loose and porous, rich in humus, but contained little silt. Numerous roots con? tributed to the looseness of the soil. Small fragmented angular rocks of a fist to a half-fist in size were abun- 16 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 dant in the two top levels and decreased to almost none in Level 3 0 ^ 0 cm. The amount of cultural material in these levels was enormous. Most of it was shell and potsherds with a small number of bones and some stone artifacts. The concentration of shell was mostly on the eastern walls of the pit, which faces the slope. Evidently the inhabit? ants of the site threw their refuse over the slope. One gunny sack of shells per level was extracted from the first 40 cm. This quantity was doubled in Level 4 0 - 50 cm. I t decreased to half this volume in the next level down. The volume of potsherds was considerably smaller than the amount of shell per level. The top level produced 1,019 sherds; the other five levels, down to 60 cm., had only from 300 to 500 sherds each. The levels on this unit correspond to the second or more recent of the two subphases established for IS -11 . LEVELS 60-100 cm.?Starting in Level 60-70 cm. the soil turned compact and fairly hard in consistency in the northern half of the pit. Its color also changed to a darkish red and yellow. The southern half pre? served the looser, more porous soil. This division in the pit between the northern and the southern half is preserved to level 90-100 cm., where compact red clay extended over the pit area. The number of potsherds in Level 60-70 cm. was nearly half that of the preceding level. This reduction in the quanti ty of cultural material also held for the shell. All shell, bone, pottery, and stone artifacts were found in the southern half of the pit. Few sherds and shells came from Level 90-100 cm., and the next level down was sterile. Levels 60-100 cm. correspond to the oldest of the two subphases established for Site IS-11. T h e Pottery Type Sequence FIGURE 10; APPENDIX 1, TABLE 4 Las Secas (IS-11) is a one-phase site belonging to the most recent of the Gulf Phases, the Chir iqui Phase. The diagnostic pottery types, Tar rago Bisquit Ware, Villalba Red-Streaked, and La Cavada Applique and Red Banded occur in large percentages. On the basis of observed changes in the physical appearance of the soil and the incidence of pottery types, the occupation at IS-11 has been divided into two subphases. The recent subphase includes the top six levels of Pit No. 1, while the earlier one includes the four bottom levels. Pan de Azucar Red Line, Cavada Appl ique and Red Banded, and Plain Ware A are types restricted to the older subphase. Arayo Polished Line is most abundant at this time. In the more recent of the two subphases, a n u m b e r of plain types appear : E, F, G, Beige Plain, and Polished Black Plain. None of these have been described in detail in this study. Decorated types restricted to the more recent subphase are Red Slipped Bisquit Ware, Red on White , Chocolate Incised, and Black on Red Negative. These types were not found in the other Gulf of Chir iqui sites. 0 - (0 1 0 - 2 0 2 0 - 3 0 3 0 - 4 0 40-50 50-60 60-70 70-80 80-90 90-100 II ? A B UNCLASSIFIED PLAIN TYPES 1 ? 1 ' I ? I ? I ? ' 0 10 20 30 40 50% Scale A D ? ED D 1 1 ? D CZl 1 1 1 1 1 1 1 r 1 1 cn ? 0 CD i 1 1 1 1 i 0 5 10 H5 20 25% Scale B I 1 I I LU cr < CO o CO Q CD UJ a: > S z < - 1 CL IG E UJ CD ca le Q 1 ti T LI SI o CL B o 00 CZ3 ? CZl 0 0 D D ? Q UJ cr Q LU < LU _ l CO < CD D rzi D d LU O _ l Q_ CL < d UJ Q < QD < Q < > o c6 Q UJ cr D dZI UJ o CL < Ci UJ o < CD Q LU cr D D CD CZ Q UJ o N UJ H ct: < UJ Q LU Q: D D CZl ? UC AR N UJ < 2 : _ j LU Q Q LU -z. cr < CL IS HE D PO L 0 >- < cr < IN E _i Q Q UJ UJ cr a. a. 0 -1 0 CO < 0 cr 0 1- UJ LI GA W AR _ i < CO N CI TO 0 0 cr 0 X UJ BA N Q UJ cr LA TE SE D 0 0 0 z 0 - X 0 LU H 1 W H I ^1 0 Ch UJ cr Q UJ CK O N R EG A TI VE < z. _i GQ FIGURE 10.?Pottery type sequence of Pit No. 1, Site IS-11 on Isla Cavada in Las Secas Island complex. Analysis of Cultural Remains T h e Chiriqui Gulf sites yielded a large quantity of cultural material in the form of potsherds, stone implements, and shell ornaments. The fact that no objects of gold were found is understandable since our collections came from refuse mounds and not from graves. Absent from our inventory are also pottery artifacts such as figurines, needlecases, musical instruments, and rattles. These occur in Chiriqui graves on the mainland and have been described under the Negative or Alligator Wares (MacCurdy, 1911). Since neither of these wares occurs in our excavations in significant quantity, it is not surprising that the pottery objects associated with them are unrepresented. Artifacts are described under four major categories: pottery types, appendage modes (handles and sup? ports), stone artifacts, and objects of shell. Other materials that may be called cultural in the sense of having been used by man, though not produced by him, such as shells and animal bones, have been summarized in tables at the end of this study. They are more valuable for cultural interpretations than for chronology. The description of the pottery types is alphabetical within each phase, beginning with the oldest phase, passing to the intermediate phase, and ending with the most recent. Distribution by phase has been graphically presented (fig. 42). The Burica Phase has the following pottery types: Isla Palenque Maroon Slipped and Unclassified Plain Ware C. The San Lorenzo Phase is typified by the following pottery: Arayo Polished Line, Banco Red Line, Caco Red Slipped, Cangrejal Red Line, Castrellon Red Slipped, Centeno Red Banded, Horconcitos Red Banded, Linarte Zoned Red Line, Pan de Azucar Red Line, Zapote Red Banded, and Unclassified Plain Wares D, H, I, J . Cavada Applique and Red Banded, Tarrago Bisquit Ware, and Villalba Red Streaked are restricted to the Chiriqui Phase. Additional pottery types are "Classical Chiriqui": Alligator Ware and Black-on- Red Negative. Probable Costa Rican imports: Choco? late Incised, Red on White. Local variant of Tarrago Bisquit Ware: Red Slipped Bisquit Ware. Unclassified Plain Wares: A, B, E, G, Beige Plain, and Polished Black Plain. 284-757 0-68?3 17 18 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 3 FIGURE 11.?Rim profiles of Isla Palenque Maroon Slipped, thick variety, Burica Phase. All interiors are oriented to the left. Arrows indicate slipped surfaces. WHOLE VOLUME CULTURAL REMAINS 19 BURICA PHASE Isla Pa lenque M a r o o n Sl ipped PASTE: Method of manufacture: Probably coiling. Trace of coil lines completely obliterated on the surface, but cross sections show faint depressions that may be coil junctions. Temper: Abundant , well-rounded quar tz sand; particles are pinpoint size. Temper particles are lighter in color so that they stand out in gray areas. They are distributed homogeneously throughout the paste. Texture: Compact and fine grained. T h e abundant temper grains give a sandpapery feel to the paste. Breaks along fairly straight lines, but not along coil junctions. Color: Incompletely oxidized firing has produced a buff to light orange line of uniform width along both sur? faces, leaving a dark gray core. T h e oxidation band is the same in the interior as in the exterior. It ranges from a little less than 1 millimeter to 3 millimeters in width, correlated generally with the thickness of the body wall. Occasionally, the oxidation penetrates com? pletely through the cross section, producing an even buff color throughout. SURFACE: Color: Light buff to light brown, with occasional areas that are black to gray as a result of fire clouds; however, these are uncommon. O n a single vessel, color is rela? tively uniform. Treatment: Smoothed to produce a relatively even sur? face, but not polished, remaining slightly abrasive to the touch and leaving scattered small imperfections. Some j a r necks show in the exterior broad, parallel striations (about 3 mm. wide), indicating scraping with a shell tool. Hardness: Very soft, 2-2.5. FORM: Body wall thickness: There are two distinct groups: thick, ranging from 10-14 mm. and thin, ranging from 3-5 m m . Group 1 belongs in the thick variety; groups 2 and 3 in the thin variety. Rim and Lip: Three major groupings. Group 1 (fig. 11 (2-r; pi. 5 a-h, d-e): Everted rims be? longing to large globular jars (probably burial urns) with long necks. Rims are thickened externally to form a slanted, enlarged lip that may be anywhere from 1-5 cm. wide. Thickening was done by the addition of one or more coils to the outside of the lip. Traces of coil marks remain on the underside of the lip as undulations separated by shallow grooves or lines made by the unknown tool used in the smoothing process (fig. 11 m-r; pi. 5b). Contour of enlarged lips either angular, with a flat top and slightly convex under side (fig. 11 h, i), or rounded at the top and at the bend of the lip (fig. 1 1 / , ^ ) . Lip is at a 45?-90 ' ' angle to the neck. R i m diameters range from 32-50 cm. Oroup 2 (fig. 12 a-h, k-m; pi. 5 c, / , i): Direct r ims (that is, rims that are continuous with the vessel wall) belonging to bowls. Great variety in terms of angle of curvature (incurved in closed bowls, outcurved in open bowls) and in the way the lip is treated. 1. Unmodified, incurved rims with lip and vessel wall the same thickness: as thin as 5 mm. (fig. 12/) or as thick as 10 mm. (fig. 12a). Diameters are 4 cm. and 15 cm. respectively. 2. Sharply incurved, meeting the vessel shoulder at an angle. T h e lip tapers toward a point (fig. 12 ^, ^, m; pi. 5 c, / , 0- Ririi diameters average 12 cm. 3. Opened, with a gradual outcurve and internally thickened lips that may be twice as wide as the vessel wall (fig. \2d-f). R im diameters range from 18-26 cm. T h e thickening may be on the exterior of the lips (fig. 12^ , h). Group 3 (fig. 12 n-v; pi. 5 h, I): Rims belonging to bowls with curved, short necks; the lips are curved from the vessel wall. 1. Thin, perpendicular rims with a gentle S-curva? ture. Angle of juncture with the wall of the bowl is marked. Lip slightly thickened (fig. 12 i, j ; pi. 5 h, I). R i m diameters average 20 cm. 2. Angular, with a sharp Z-curvature above the shoulder of the vessel. Lip may be long, flat at the top, and sharply everted at a 45? angle (fig. 12 n-p) or it may be only gradually outcurved (fig. 12q). R i m diameters range 15-20 cm. 3. Gradually S-curved, with either exteriorly thickened lips (fig. 12 r, s) or unmodified lips (fig. 12^. R i m diameters average 15 cm. 4. Everted, very thin rims belonging to small j a r s with short necks. The lip is thin and pointed (fig. 12M) or thick and wide (fig. 12y). These are less popular than the other rims. R i m diameters average 12 cm. Group 4 (fig. 12w; pi. 5k): Sharply incurved rims belonging to bowls with constricted mouths. Marked thickening of the lips by the addition of coil. R i m diameters average 8 cm. Base and Supports: 1. Concave unmodified base continuous with the side walls of large globular jars. Thickness: 2 cm. 2. Small, thin-walled ringstand 7 cm. wide and 5 mm. high (Ringstand Mode a) associated with outcurved, angular, opened bowls (fig. 12/). Ringstands may also occur with incurving or S-shaped bowls, but no whole vessels have been found to confirm this association. 20 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 f I I I I I I 0 I 2 3 4 5 CM FIGURE 12.?Rim profiles and side views of Isla Palenque Maroon Slipped, thin variety, Burica Phase. All interiors are oriented to the left. Arrows indicate slipped surfaces. Complete vessel in Museum of Natural History, Smithsonian Institution (Catalog No. Ar? cheology?108300). WHOLE VOLUME CULTURAL REMAINS 21 3. T h e tall, hollow tripod foot with a modeled anthropomorphic figure a t the top. Mode a l may belong in this group (see description of Tripod Foot, M o d e a l , and page 47). Reconstruction of major vessel forms: 1. Large, globular-bodied jars with either long, narrow necks or short, wide necks; sharply everted, thick? ened rims and rounded bases (fig. 11a). These were used in all probabiHty as burial urns (fig. 13). 2. Small jars with thin-walled globular bodies, short, thin or thick necks, and everted rims (fig. 12 u, v). 3. Deep bowls with gradually concave sides terminating in incurving rims with lips that are either unmodified or thickened in either the exterior or in the interior. Bases probably concave. 4. Deep bowls on ringstands with short, fairly straight rims (fig. 12x.; pi. Ga). These were probably used as fids for burial urns of vessel form No. 1 (fig. 13). 5. Shallow opened bowls with a sharp angle at t h e shoulder and either incurving or outcurving rims. 6. Angular bowls with short S-shaped necks and either strongly everted lips or more gradually outcurving rims. 7. Bowls with incurving sides, constricted mouth, and exteriorly thickened lips. DECORATION : Technique: A dark, maroon (brick color) slip applied evenly over the vessel surface; thinness of the slip has caused it to wear off" in spots. Polishing of the slip has left prominent tracks running parallel to the rim. Striations produce a slightly variegated appearance. T h e polishing produces a relatively hard surface, 3.5. Motif: The slip is applied to cover most of the visible surface and frequently extends into less visible areas such as the underside of the shoulder. When applied to the interior of the large j a r necks, the slip extends over the top of the rim, but the exterior of the neck is left unslipped (fig. 11 a-r). Most deep and shallow opened bowls are slipped in the interior also (fig. \2a-i, k, m, n-q, s, t). Closed-mouth bowls are left unslipped inside. Additional decoration: M a n y of the bowls have, besides slip, ornamentat ion of one or more of the following kinds: 1. Shallow parallel diagonal incisions about 3 mm. wide in zones 6-12 mm., placed on the exterior above the shoulder and below the lip (fig. 12;'; pi. 6 a-c). 2. Slightly raised parallel diagonal ridges of the same dimensions and placement as described above (fig. 120. 3. Applique curvilinear ridges and pellets, 5-12 mm. in diameter, just above the shoulder line and below the lip. Some look like an arm motif (fig. \2 m, o,w\ pi. 5 c, f, k). CHRONOLOGICAL POSITION: This is the dominant pottery type in the Burica Phase, the oldest phase in the Gulf of Chiriqui. I t is present in very large percentages in the bottom two levels of Pit No. 3 at IS -3 and in the bottom level of Pit No. 2 at Site IS -7 . Although the detailed description of the plain wares has been left out of this study, it may be of interest to note that Plain Ware G diff"ers from pottery type Isla Palenque Maroon SUpped in only one respect; it lacks a red slip. Otherwise they are identical. I suspect that they are in fact the same type, only the thin maroon slip has worn off type G. GEOGRAPHICAL DISTRFBUTION AND COMPARATIVE MATERIAL: Isla Palenque Maroon Slipped is the predominant and often exclusive pottery type in most of the surface collec? tions from hilltop sites along the shoreline of the Burica Peninsula (Ranere, Appendix 2, pp. 107-119). T h e Smithsonian Institution ( U S N M Archeology Cat. No. 108300) has a small collection of rims, body sherds, a complete bowl, probably a lid (fig. I2i; pi. 6a), all of which belong to this type. McNeil m a d e this collection in 1859 at a site called El Banco. There a r e at least two El Bancos in Chiriqui, one north of Da vid in the highland region near the town of Boquete a n d another south of Puerto Armuelles on the Burica Pen insula. T h e McNeil collection may have come fro m either place. In recent years the Isla Palenque M a r o o n Slipped type has been found in several cemeteries of Chiriqui: in Dolega and La Bagala (12 kilometers nor th of David on the way to Boquete), in Bugavita (near Goncepcion), in Querevalo and Finca Tinajas (west of David and south of Concepcion), and possibly in Cerro Punta (on the skirts of the Volcan Barii). T w o of these sites, Dolega and Tinajas, have been described in the literature (Miranda, Perez and de la Guardia , 19 66; de la Guardia, 1966). The others were mentioned by Dr. Roberto de la Guardia (personal communication). I have also seen most of his material firsthand. There is very little doubt in my mind that it belongs to my Isla Palenque Maroon Slipped type and should be in the Burica Phase. The following descrip? tion is based largely on the published evidence men? tioned above. (Figure 13 illustrates some of this material.) The sites of Dolega and Tinajas are at least 70 kilo? meters apart, yet the material they contain is very similar. They are cemeteries with secondary urn burials in large urns varying in height from 30 to over 80 cms. The burials are usually marked by river boulders, and the urns are fairly close to the surface. One of the Tinajas urns had two dismembered skeletons inside. Besides large urns, there were smaller ones with shapes and rims exactly like those of my Palenque Maroon Slipped type, thick variety. Several of the urns were covered by inverted bowls or lids identical to those from my Isla Palenque Maroon Slipped, thin variety (fig. 12z)- Other ceramic objects found at Dolega and Tinajas also occur in our Chiriqui Gulf sites. One of these was tripod feet in the shape of women with legs joined and broad hips (de la Guard ia , 1966, figs. 9 -12) ; our Mode b3 bears some resemblance to these. A pottery gourd shaped into a face found at Tinajas is unique. There was also a tall pedestal base with tr iangular cutout designs identical to one from Code , illustrated by Lothrop and assigned by him to his Early period (Lothrop, 1942, p. 167, fig. 337c). Stone objects ranged from a tall (66 cm.) base for a statue, of which only the feet remained, to many metates of diff"erent sizes and leg shapes. 22 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 FIGURE 13.?Burial urns of Isla Palenque Maroon Slipped, thick variety, from the sites of Las Tinajas and Dolega reported by other investigators. Redrawn from photographs. WHOLE VOLUME C U L T U R A L R E M A I N S 23 I t is now clear from the work by de la Guardia and his students tha t the Isla Palenque Maroon Slipped type is more widely distributed in Chiriqui than formerly thought (Linares, 1964). I t also seems that cemeteries with urn burials existed in the islands, and that we missed them by concentrat ing on the habitat ion sites. An urn with bones inside has been reported recently from Isla Palenque (de la Guardia , personal communica? tion) . Outside of Chiriqui the closest connections of Isla Palenque Maroon Slipped are with the oldest pottery in the Diquis Delta of southwestern Costa Rica : the Brown Ware and the Fugitive Red Wares of Lothrop (1963, pp . 45 to 75). Only some of the types subsumed under each of these headings were found in the bottom levels of the Diquis deep stratigraphic deposits. These types were Large Brown Bowls (including Short and Tall Tripods and a Thick Lip Type) , Large Brown J a r s with parallel-hne incisions, and thin undecorated sherds belonging to a Fugitive Red Ware . The latter were not described in detail except to state that they were slipped in red on one side and that no rims, bases, legs, lugs, or handles were present (op. cit. p . 72). The Short and Tall Tr ipods of the Brown Ware (op. cit., p . 49, fig. 22 and p. 66, fig. 43) are exactly like some recently found in the Burica Phase cemetery of Tinajas (de la Guardia , 1966, p . 11, figs. 9-14). The rims of the gigantic Brown J a r s (Lothrop, 1963, p . 52, fig. 25, a,b) duplicate those of our Isla Palenque Maroon Slipped, thick variety. T h e large Parallel-Incised Ja r s , found definitely stratified in the lower levels of Site A, F a r m 4 (op, cit. p . 55), show definite resemblances to funeral urns (fig. 13) found in the Burica Phase cemeteries of the coast. Lothrop (op. cit. p . 57) speculates whether the Diquis J a r s were for storage or for burial. U n c l a s s i f i e d P l a i n W a r e s Pla in W a r e C a t Si te I S - 3 , P i t N o . 3 , is t h e on ly p la in w a r e in t h e B u r i c a P h a s e . I n all a t t r i b u t e s of pas te a n d surface it is iden t i ca l to Is la P a l e n q u e M a r o o n S l ipped , on ly it lacks t h e m a r o o n - c o l o r e d slip of its d e c o r a t e d c o u n t e r p a r t . T h e slip of this p o t t e r y t y p e is t h i n a n d m a y h a v e s imply w o r n off. SAN LORENZO PHASE Arayo Polished Line PASTE: Method of manufacture: Probably coiling. Vigorous polishing of both surfaces has obliterated all coil junctions. Temper: A "d i r ty" clay with some natural impurities: decomposed granitic sand, feldspar, quartz, hematite. T h e particles, clearly visible in cross section, are homogeneously distributed. Large (1 mm.) particles of rounded hemati te stand out in cross section. Texture: Fairly compact ; breaks along straight lines. Color: An even buff to pale gray depending on the degree of oxidation. Firing is variable. In some sherds it penetrated completely; other sherds have a thin (4 mm.) gray core or, in a few, a core that occupies 75 percent of the cross section. SURFACE: Color: Buff to reddish brown on the exterior. Interiors may be pale gray and different in color from the buff exteriors. Fire clouds are rare. Treatment: On globular jars both surfaces have been polished with a pebble, leaving striations at least 2 mm. in width. This is the only unslipped type where interiors are as smooth to the touch as exteriors. On short-necked jars the interior of the body is unpolished; the inside of the neck may also be polished. Hardness: Fairly hard, 4. F O R M : Body wall thickness: Consistently between 0.7-1.0 cm., wi th the majority 8 mm. Rims and Lips: There are only a few rims which can be assigned with certainty to this type. At least four groups are distinguishable: 1. Long (5.5-6.0 cm.), unmodified rims gradually everted from the vessel shoulder (fig. 14 a-g). The i r angle of outflare is generally 45 degrees. Lips vary slightly in shape. Mouth diameters are often difficult to estimate because many of the rims are lopsided, as they have been broken off at the point where the lip continues into a strap handle. R i m diameters are geneially between 30-38 cm. 2. Short (3.7 cm.) rims, sharply everted, with the m a i n flatness at the top and the lip rounded (fig. 14 h-i). Mouth diameter is difficult to estimate because of the strap handles, but averages 8.5 cm. 3. Inslanted direct rims belonging to deep bowls with fairly straight upper sides. T h e lip is folded on the exterior to form a very narrow (1-2 mm.) ledge (fig. 15/). 4. Straight rims belonging to tall-necked jars ; the rim is continuous with the neck. T h e lip is unmodified and in thickness is equal to the vessel wall (fig. 15^). Base and Supports: Rounded bases 1-2 mm. thicker than the vessel walls; no tripod feet or supports of any kind are associated with this type. Appendages: Strap handles are popular, judging from the junctions on the sherds where the straps were once attached and from published drawings of vessels of this type. They may occur singly on vessels with a 24 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 4 CM r/^^ FIGURE 14.?Rim profiles of Arayo Polished Line, San Lorenzo Phase. All interiors are oriented to the left. Arrows indicate slipped surfaces. modeled neck or paired on vessels with no neck. (Holmes, 1888, fig. 160) (See Strap Handle Mode a, variations a l - a 6 , page 56.) Reconstructed vessel forms: All vessels in this type are globular j a r s ; they vary, however, in the treatment of the neck. 1. An estimated 75 percent of the vessels are globular jars without necks and with rims that flare out directly from the shoulder (Holmes, 1888, fig. 160). 2. At least one vessel is an effigy j a r ; the short neck (2.2 cm.) is decorated with applique pellets to convey the idea of a h u m a n face (fig. I5h). T h e body of the vessel is globular. 3. Several vessels have tall necks (more than 8 cm. high) and globular bodies, the exact shape of which is difficult to reconstruct (fig. 15^). DECORATION : Technique: The most diagnostic criteria of this type are narrow (2.5 mm.) and wide (1.2 cm.) polished red lines or bands on the upper half of the vessel. T h e decorated area never covers the entire exterior surface. T h e lines were painted with an applicator (a brush?) unevenly steeped in paint so that width is greater at the top, becoming thinner as the lines are drawn out. Sections of the lines where the paint was thinnest have eroded. Motif: Sets of parallel lines meet or cross each other at right angles in the following ar rangements : 1. T h i n (2 mm. wide) parallel lines, 3-5 m m . apart , are arranged in sets of four to seven lines tha t meet without crossing (fig. I5b) or actually overlap (fig. 15a). T h e outside lines m a y have pendan t ovoidal blobs as additional decoration. A wide (1.1 cm.) band encircles the neck. 2. A horizontal band 2 mm. wide or a set of thin, horizontal lines have pendan t triangles, 2.0 by 1.5 cm. wide (fig. 15c, d; pi. 7k, m). 3. Sets of thin (3 mm.) parallel lines meet each other at right angles and are separated by wide bands on the outside (fig. 15 e-h; pi. 7 h-j, I, n, o). This is the most common motif in this type. Additional decoration: Effigy vessels have the neck decorated with applique coffee-bean eyes and appl ique strips in the form of a nose and two arms (fig. 15^). CHRONOLOGICAL POSITION: Arayo Polished Line is pop? ular mostly dur ing the San Lorenzo Phase at Site I S - 3 and occurs in large percentages in the same phase at WHOLE VOLUME CULTURAL REMAINS 25 I I I I I I 0 I 2 3 4 5 CM FIGURE 15 ?Painted exterior designs and a few rim profUes of Arayo Polished Line, San Lorenzo Phase. All rim interiors are oriented to the left. Arrow on rim indicates red-slipped surfaces. Dark areas mdicate red bands. 26 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 Site S L - 1 . Very small percentages also occur in the Chiriqui Phase at Sites IS-7 and IS-11. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: Vessels of this type were put by Holmes (1888) and Mac? Curdy (1911) in the " R e d Line ' ' group, which also includes vessels of Zapote Red Banded, Cangrejal Red Line, Linarte Zoned Red Line, Centeno Red Banded, and Banco Red Line. A sherd of Arayo Polished Line (fig. 15a) can be identified as identical to par t of a complete vessel illus? trated in Holmes' report (Holmes, 1888, fig. 160). H e gives no provenience. T h e sherd in figure I5h is similar to par t of a vessel in MacCurdy ' s report (1911, plate X X V c). A sherd of Arayo Polished Line found in the Diquis Delta of Costa Rica (Lothrop, 1963, fig. 69a) is decorated with motifs very similar to those on our figure 15^. Banco Red Line PASTE: Comparable to Linarte Zoned Red Line. See tha t type description for details. SURFACE: Color: Buff, darkened mostly in the lower half by fire clouds. Interior color is uncertain since this type is represented by complete vessels with constricted necks. Treatment: T h e exterior has a low polish that compacts the surface, but does not give it a smooth, shiny texture as does the polishing in Arayo Polished Line. Hardness: Between 2.5 and 3.0. F O R M : Body wall thickness: From 6-8 mm., with the range of variation encompassed by an individual vessel. Rim and Lip: Everted rims with unmodified lips belonging to jars . These are of the same general shape as the rims for Arayo Polished Line and Centeno Red Banded. For a detailed description of these rims, see the type description of Arayo Polished Line. Bases and Supports: A very slight flattening of an otherwise rounded base lends some stability to the vessel. T h e base is also thickened somewhat. Appendages: Larger jars have strap handles that extend from a point on the vessel wall just above the shoulder to the opposite side of the rim. T h e straps are wider at the top where they join the lip than where they join the vessel wall (Strap Handles, variations a6 -a7) . Small vessels have no appendages (fig. 166; pi. 8b). Vessel shapes: Globular jars with slightly angular to rounded bases (fig. 16a, b; pi. 8a, b), everted rims, and, in the case of the larger vessels, strap handles joining body and lip. DECORATION : Technique: See description of Linarte Zoned Red Line. Motif: The decoration is restricted, as in all the other j a r types that are decorated with thin lines, to the section of the vessel between the neck and the shoulder. This section is set off from the rest of the vessel by a horizontal line encircling the shoulder and either a FIGURE 16.?Reconstructed vessels of Banco Red Line, San Lorenzo Phase. Dark areas indicate red bands. continuous band around the neck or discontinuous neck bands. T w o motifs a re : 1. Intersecfing sets of four parallel thin lines wi th outer lines decorated by a row of pendan t dots or blobs (the motif is almost identical to Arayo Polished L ine ; compare fig. 15a with fig. 16a). 2. A triangular motif made up of parallel th in lines connected by transversal lines. T h e area in the center is decorated with solid red splotches of paint (fig. \6b). CHRONOLOGICAL POSITION: Restricted to the San Lorenzo Phase at Site I S - 3 . GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: Banco Red Line belongs in the same Red Line W a r e of Holmes (1888) and M a c C u r d y (1911) as the types Arayo Polished Line, Cangrejal Red Line, Centeno R e d Banded, Horconcitos Red Banded, Pan de Azucar Red Line, Linarte Zoned Red Line, and Zapote Red Banded. A vessel of the exact shape as figure I6b, but with slightly varying decoration, appears in MacCurdy ' s report (1911, fig. 160). No provenience is given. Caco Red Slipped PASTE: Method of manufacture: Probably coiling; traces ot coil lines were completely obliterated from the surface, but some interiors show depressions that may be coil junctions. Temper: A "d i r ty" clay, probably from a stream, con? taining waterworn particles of hard clay, decomposed granitic sand, feldspar, quar tz , and hemati te . T h e particles of quar tz and hemati te , ranging from 0.2-0.5 mm., stand out readily in cross sections and are clearly visible on eroded surfaces. T h e y are unevenly distributed throughout the paste. Texture: Gri t ty; temper particles are occasionally clustered. Breaks along irregular, angular cleavages. Large air pockets (some 6 mm.) are distr ibuted through? out the paste of the thicker sherds. W H O L E VOLUME CULTURAL REMAINS 27 Color: Thicker sherds show incompletely oxidized firing that produces a light buff band, usually on the exterior, varying in width from 0.5-^.0, leaving the remaining 75 percent of the cross section dark gray. In thinner sherds oxidation is complete, producing an even buff color throughout. SURFACE: Color: Buff, light orange to light gray depending on the degree of oxidation. O n any one vessel, color is fairly even because there are few fire clouds. Treatment: Exterior is compacted by polishing with a pebble tha t leaves wide striation marks. Interior is typically smoothed but not polished. Hardness: Fairly hard, 4. FORM: Body wall thickness: T w o distinct groups: thick (1.2-1.5 cm.) and thin (6-8 mm.) . Rim and Lip: 1. T h e thick variety of sherds has large rims with varying angles of eversion and lip treatment. They all belong to large globular jars (fig. 17 a-c). R im diameters vary between 28-33 cm. 2. T h e thin variety of sherds belongs in two very different r im groups: (a) Everted rims belonging to globular jars. These may be short with rounded lip (fig. 1 Id) or with a Hp that is fairly flat on the outside (fig. 17^). T h e lip may also be thickened exteriorly (fig. 17^) or tapered to a fine point (fig. 17^). Mou th diameters are, respectively, 16, 20, 16, and 18 cm. (b) Incurving rims belonging to fairly deep bowls. All have unmodified lips. They may be con? stricted at the shoulder (fig. 17^) or not con? stricted (fig. \lh-i). A variant is a small angular r im with a lip that is slightly thickened on the exterior (fig. 17/). M o u t h diameters are, respec? tively, 9, 16, 18, and 7 cm. Base and Supports: 1. Large jars have rounded bases that are slightly thickened (0.5 cm.) . 2. Small bowls also have rounded bases, but these are usually unthickened. 3. Deep bowls may rest on pedestal bases, but this association cannot be confirmed without further work. DECORATION Technique: A leathery red-brown slip applied on a pre? viously polished surface. After application, the slip itself is polished, leaving horizontal streaks parallel to the rims. Motif: 1. T h e slip is applied only to the exterior surface. (a) In one group the slip covers the entire outer surface. (b) In the second group the slip is applied in zones. There is no clearly demarcated line where the slipped area ends; instead, there is a shading off of the red zone. ? ((^ i f 0 I 2 3 4 CM FIGURE 17.?Rim profiles of Caco Red Slipped, San Lorenzo, and Chiriqui Phases. All interiors are oriented to the left. Arrows indicate slipped surfaces. 2. The interiors of the thick rims and some of the exteriors are slipped. 3. The thin rims show wide variation in the slipping. A few are slipped only on the interior, most are slipped only on the exterior, and a few are slipped on both surfaces. CHRONOLOGICAL POSITION: Caco Red Slipped occurs in all three phases of the Chiriqui Gulf, but it is most popular in the San Lorenzo Phase (the middle phase) a t Sites IS-3 and S L - 1 . I t occurs in very small percentages at Sites IS-7 and IS -11 . GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: This type has not, to my knowledge, been described in the literature. It is absent from the reports of Holmes (1888) and MacCurdy (1911). Cangrejal R e d Line PASTE : Method of manufacture: Coiling. Coil marks remain as faint undulations separated by depressions (coil junc? tions) on the interior surface of the bowls. T h e exterior has been polished, obliterating all coil marks. Temper: Like Zapote Red Banded and Castrellon Red Slipped, a fairly " p u r e " clay mixed with little, if any, tempering materials. Natural inclusions are mostly 28 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 abundan t rounded hemati te nodules 0.5-3.0 mm. in diameter. Very few feldspar particles appear as white specks in cross section. Texture: In most sherds the paste is so compacted that there are no air pockets visible in cross section. Where present, air pockets are very small and shallow. Breaks are clean and fairly straight. T h e fracture point is usually at the shoulder where the r im incurves markedly. Color: A light buff to dark gray, depending on degree of oxidation. Some sherds show complete oxidation so that they are a light buff throughout the cross section. In others, incomplete oxidation has produced a dark gray core 1-5 mm. wide. Th in oxidized bands (2 mm.) extend along the surface on either side of the core. SURFACE : Color: Ranges from bright orange to light buff or cream. Extensive fire clouds cover the areas below the shoulder of the vessel, occasionally extending up into the dec? orated portions of the rim. Bowls were probably laid upright in the firing process. Treatment: Polishing has produced surfaces varying from smooth to shiny. O n some sherds the interior surface has been polished more than the exterior. Where vigorous polishing has occurred, shiny striations appear parallel to the rim. Hardness: Hard , 4. F O R M : Body wall thickness: T w o distinct groups, one thick (8-10 mm.) and one thin (4?7 mm.) . In the thin group, individual vessels fluctuate in thickness from 4 mm. below the shoulder line to 7-8 mm. at the shoulder and the rim. The wall of the thick variety is often, though not always, thinner at the shoulder. Rim and Lip: All are direct rims belonging to deep bowls with incurving walls and constricted mouths. There is marked variation in the length of the rims, but only slight differences in lip treatment. Variations include: 1. Long (8.4 cm.) incurving, direct rims with un? modified rounded lips of the same general thickness as the body wall above the shoulder (fig. 18a). R i m diameter: 26 cm. 2. Long (6.4 cm.) incurving direct rims thickened gradually below the lip to 4 mm. thicker than the body wall just above or below the shoulder. Lip is slightly tapered (fig. 18^). R i m diameter is 24 cm. 3. Long (5.0 cm.) incurving rims with either thickened lips (fig. 186) or lips that are markedly tapered (fig. I8t) . Some rims of the same general length (4.7 cm.) are completely unmodified so at no point are they wider than the vessel wall. R i m diameters average 18-20 cm. 4. Short ( 3 . 0 ^ . 0 cm.) and fairly thin (7 mm.) rims 1-2 mm. thicker than the wall below the shoulder line. Their angle of incurve may be sharp (fig. 18^) or gradual (fig. 18^). Lips are rounded. R im diameters average 15-18 cm. Base and Supports: Vessels have a concave base resting on strapped feet that are placed approximately 3 centimeters below the shoulder line (fig. 18 A; pi. 9 a ) . See complete description under S t rapped Feet Modes . Reconstructed vessel shapes: All vessels of this type are variations of deep or shallow bowls wi th incurving side walls, constricted mouths , and tr ipod feet in shape of flat straps. T h e main groups a re : 1. Large, very deep bowls, approximate ly 14?16 cm. deep from the lip to the interior, and a calculated diameter at the shoulder of at least 30 cm. Rims are direct, and the mou th d iameter is a round 26 cm. T h e y rest on s t rapped feet. 2. Bowls that are slightly smaller than F o r m I, with a depth of 10 cm., a shoulder d iamete r of 20-22 cm., direct rim, and mou th d iameter a round 18-20 cm. Supports consist of s t rapped feet. 3. Shallow bowls, only 5-6 cm. deep, wi th walls that are thinnest below the shoulder. T h e y rest on strapped feet and have r im diameters tha t range from 15 to 18 cm. DECORATION : Technique: Th in , red lines of a bright rich red or leathery brown are painted on the polished, unslipped exterior section of the bowl between the lip and the shoulder. T h e lip is painted with a red band wi th uneven borders that extends for 4 mm. onto the bowl interior. Fre? quently, drips of paint show on the interior surface. T h e red paint was applied either thickly to cover the under surface or thinly so tha t the under lying surface shows through. T h e individual thin lines that decorate the exterior vary in width between 1 and 4 m m , due to unequal pressure of the paint ing tool. T h e y are wider near the lip band where the appl icator was more steeped in paint, becoming narrower as the line was drawn out. Where two lines intersect, they frequently over? shoot. T h e lines are fairly straight, bu t they are not evenly spaced. Motif: T h e section between the lip and shoulder is marked off by a red band on the lip and a thin (2-3 mm.) band circling the vessel at the shoulder line. Sets of thin parallel lines d r a w n diagonally to the rim are arranged in a zigzag pa t te rn a round the vessel. T h e spaces between the thin lines are in the shape of triangles with apexes facing al ternately up or down. T w o predominant motifs are discernible: 1. Several lines, 4.5 cm. long and 1-4 m m . wide, are ar ranged in sets tha t jo in al ternately at the rim or at the shoulder to define a tr ianguloid area. With in one vessel, the sets of lines m a y either meet or be as much as 2 cm. apar t . T h e lines defining the triangles have pendan t shorter lines, 6-9 mm. long, ju t t ing into the empty t r iangular space (fig. 1 8 ^ ; pi. 9 6, c). Wi th in one vessel, these decorative short lines may j u t at different angles from each other. 2. T h e number of parallel, d iagonal lines varies between 6 and 10 in a set, and they m a y be any? where from 5 to 9 centimeters long, depending on the size of the vessel. T h e space left between the sets of fines is filled entirely wi th paint , creat ing a W H O L E VOLUME CULTURAL REMAINS 29 0 I 2 3 4 CM FIGURE 18.?Rim profiles and exteriors of Cangrejal Red Line, San Lorenzo Phase. Hachure indicates red-painted areas. All rim interiors are oriented to the left, h, Reconstructed vessel shape. mofif of solid red triangles, alternately facing down from the lip or up from the shoulder (fig. 18 a, b; pi. 9 d,e). Additional decoration: A few of the shallower bowls have applied in the center of the triangular space between the sets of lines one or two small round nubbins 0.7-1.0 cm. in diameter (fig. 18.^, h; pi. 9a). CHRONOLOGICAL POSITION: Cangrejal Red Line occurs in the San Lorenzo Phase and in the Chiriqui Phase at Site I S - 3 . I t is very popular in the San Lorenzo Phase at Sites SL-1 and I S - 3 , but is largely absent from the other two sites. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: A bowl of this exact type, with decorative motif No. 2, has been described by MacCurdy (1911, p. 93). He includes it within his Red Line Ware and gives its provenience as El Banco. Unfortunately, there are several El Bancos in Chiriqui Province. It seems curious that the Cangrejal Red Line type is absent from our excavation at Villalba (IS-7) where the other San Lorenzo Phase types are present. Haberland (1960a) found a whole vessel of this type in his excava? tions at Villalba. 30 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 J I 0 I 2 CM FIGURE 19.?Rim profiles and interiors of Castrellon Red Slipped, San Lorenzo Phase. Hachure and arrows indicate red-slipped areas. All rim interiors are oriente d to the left. Castrel lon Red Slipped PASTE: Method of manufacture: Probably coiling; all traces of coils have been obliterated in the polishing process. Temper: Little, if any, tempering material has been added to the clay. T h e most abundan t particles are rounded hematite nodules varying in size from small (a pinpoint) to fairly large (3 m m . ) ; occasional feldspar particles show in cross section as white specks. Both arc natural inclusions in this type of clay. Texture: Compacted because of the lack of temper particles added to the clay. Small air pockets are visible in cross section. Breaks along fairly straight lines; the broken edge is sometimes jagged in ap? pearance. Color: Typically fairly uniform red-buff with a wide (6 nmi.) gray core, so light in color as to be almost imperceptible. Cores are located in either the center or along the interior of the cross section. In spite of these cores, the oxidation is fairly complete. A small minority is less completely oxidized, leaving a wide band (6-8 nmi.) of light gray at either the center or the interior, contrasting with the thin (1-2 nim.) red- buff band along the oxidized edge (or edges). SURFACE: Color: T h e majority is bright orange; some are an at tenuated buff, and few are lighter buff in color. Extensive fire clouds produce dark gray areas on the exterior and frequendy extend over the lip to form an irregular band along the interior of the r im ; occasionally, the fire clouds also extend down into the inside of the bowls. Treatment: Both surfaces have been polished. Striation marks horizontal to the r im are clearly visible, especially on the unslipped exterior surface. Polishing has pro? duced an even, smooth, compacted surface. Hardness: Fairly hard (4). F O R M : Body wall thickness: Fairly uniform; 7-10 mm. Rim and Lip: All sherds of this type are open bowls varying slightly in r im and lip t rea tment . T h e main variations are : 1. Slighdy incurved rims tha t become gradually thicker toward the top and have a fairly rounded lip or are interiorly thickened at the t ip (fig. 19 a, ^) . R i m diameters are 20-24 cm. respectively. 2. Same as above, only the rims are completely un? modified and the lip is rounded (fig. 19 / , ^ ) . R i m diameters are 18 and 24 cm. respectively. 3. Straight, outflaring rims with lips tha t are flat at the top ; in some the lip forms an angular shelf, 1.0-1.2 cm. wide, a round the r im (fig. \^c-e). R i m diameters are 24 and 28 cm. Base and Support: No complete vessels were found; neither have they been reported in the l i terature. T h e associa? tion of the following support modes with Castrellon Red Slipped is therefore tentat ive, based on a similar distribution and on shared at t r ibutes of paste and decoration. Detailed description of the modes is given under Ringstands and Pedestal Bases (pp. 50-53). 1. Ringstands that are plain or slipped on the exterior and slipped brightly in red in the interior (Ring- stand, Modes b, c) . 2. Low pedestal bases with an unslipped, incised ex? terior and a bright red slip in the interior (Pedestal Base, Mode a) . Reconstructed vessel forms: All are deep open bowls of two shapes, differing only slightly. 1. Concave side walls tha t are fairly thin (7 mm.) and insloping rims tha t are either left unmodified or have the lip thickened to twice tha t of the body wall. R i m diameters range from 18 to 25 cm. These bowls probably rest on r ingstands or on pedestal bases. 2. Straight-sided flaring bowls with outcurved rims and flat lips. Walls of vessels are 0.7-1.0 cm. thick, in? creasing toward the r im. T h e y may be associated with pedestal bases and ringstands. DECORATION : Technique: T h e majority has a bright rich red slip ap? plied thickly on a previously polished surface; in rims of group I, the slip is a dark red-brown color due to extensive fire clouding. In sonre specimens the slip has worn off (pi. lOy-w). W H O L E VOLUME CULTURAL REMAINS 31 After application, the slip has been polished with a pebble, leaving striation marks parallel to the rim. The polishing produces a hard surface. Motif: T w o groups depending on the area covered by the slip: 1. I n more than 75 percent of the sherds the slip is applied to the vessel interior, covering the entire lip. Except on markedly bevelled lips (fig. 19 c-e), the slip extends over the lip, forming a narrow (I mm.) , irregular band on the exterior of the bowl. 2. A small number , all with r im shapes of the second group, are slipped on both surfaces (fig. 1 9 / , ^ ) . In these the slip is less heavily applied; it has a streaked appearance , with the unslipped surface showing through in spots. CHRONOLOGICAL POSITION: Castrellon Red Slipped is re? stricted to the San Lorenzo Phase. It occurs in small percentages at Isla Palenque (IS-3) in the middle levels. At the Cangrejal site (SL-1) it occurs in the top five levels, with its max imum popularity in level 40-50 cm. It is absent from Site I S - 7 . GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: No vessels of this type have been reported by either Holmes (1888) or MacCurdy (1911) because this pottery type precedes in time the wares that they describe in their reports. Flaring bowls are common in the C o d e polychrome styles at Sitio Conte (Lothrop, 1942, Part I I , p . 9) and may have inspired the Castrellon Red Slipped bowl form. Centeno R e d Banded PASTE: Method of manufacture: Coiling. Coil junctions have been obHterated from the exterior surface, but on the interior they are prominent as wide, folded depressions parallel to the r im. Temper: An almost " p u r e " clay with a fair amount of very small rounded nodules of hematite inclusions, natural in this type of clay. Texture: Compacted because of the lack of temper particles added to the clay. No air pockets are visible. Breaks cleanly. Color: A completely oxidized cream color of the same even shade throughout. T h e surfaces are the same color as the cross section. SURFACE: Color: Even cream-color zones, broken by fire clouds, are visible in restricted areas. Treatment: T h e exterior has been polished when the clay was wet so as to float fine clay particles, giving the surface the appearance of having been slipped. No a t tempt has been made to polish the interiors or even to obliterate coil junctions. Hardness: Fairly hard, 3.5-4. F O R M : Body thickness: Variat ion from 0.7-1.0 cm. may occur on the same vessel. Rim and Lip: Everted, with lips that are gradually thickened and flat at the top. In a single vessel the height of the r im may vary, depending on whether or not the cross section is taken at the point where the lip extends into the strap handle. Base and Supports: A slightly thickened rounded base; no known supports of any kind. Appendages: Strap Handle Mode c (pp. 55-56). Reconstructed vessel forms: Effigy jars with a tall neck (3 cm.) and rounded thickened bases (fig. 20a). R ims are outflared and unmodified. One or two strap handles extend from the lip to meet the vessel wall above the shoulder and below the neck. DECORATION : Technique: Red-painted bands on the polished unslipped exterior surface. Bands are two widths: wide ones range from 0.7-1.0 cm. and narrow ones from 2-3 mm. Individual bands usually fluctuate in width to en? compass the range of variation because of unequal pressure of the applicator. Some of the narrower lines show greater width at the beginning, where the appli? cator was more steeped in paint, becoming narrower as the band was drawn out. Thickness varies from complete coverage of the underlying surface to a thin coat of paler color as a result of incomplete coverage. Broad bands were applied first, then narrow lines added, with intersections overlapping the broad ones and occasionally extending beyond them. Overshot intersections are typical of the narrow lines. Lines are generally straight, but not always evenly parallel. Motif: Wide bands divide the decorated area, which is between the base of the neck and the shoulder, into rectilinear zones containing straight, narrow lines (fig. 20; a-c; pi. 7 a-f). Occasionally, wide bands meet to form a more extensive solid red area. T h e typical arrangement is 3 parallel lines, 2-4 mm. apart, following the contour of the zone and leaving the center blank. T h e inner thin lines may have pendant ovoid blobs 5-9 mm. in length and 10-13 cm. apart, with spacing relatively consistent on a single sherd (fig. 206; pi. Ic, / ) . A red band also occupies the interior of the r im adjacent to the lip (fig. 20a). The strap handle is painted with red on the top and along the edges (fig. 20.). Additional decoration: The neck may have applique designs of a circle with a small hole in the center and a strip divided by incisions (fig. 20a). Although other parts of the neck are missing, the details indicate that this was probably an effigy vessel. CHRONOLOGICAL POSITION : Centeno Red Banded occurs in small percentages in the San Lorenzo Phase at Site I S - 3 . I t also occurs sporadically in the same Phase at Sites SL-1 and IS -7 . Traces of the type are found in the Chiriqui Phase at Site I S - l l . GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: No Centeno Red Banded has been described in the published literature, but it belongs in the general " R e d Line W a r e " category of Holmes (1888) and MacCurdy (1911). 32 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 CM FIGURE 20.?Painted designs on the exterior surface and rim profiles of Centeno Red Banded, San Lorenzo Phase. Rim interiors are oriented to the left. Arrows on rims indicate red-slipped surface. Areas filled in black represent red-painted bands. Horconc i to s Red Banded PASTE AND SURFACE: Like Linarte Zoned Red Line; see that type description for details. F O R M : Body wall thickness: 8-10 mm. Rim and Lip: Sharply everted, with a thick rim and a flat lip (fig. 21a; pi. \\e). Base and Supports: Rounded , slightly thickened base. Appendages: None. Reconstructed vessel form: A large globular j a r with a rounded, slightly thickened base, a constricted mouth, and a sharply everted r im with a wide (1.3 cm.) flat lip. DECORATION : Technique: Red bands 1 cm. in width are painted roughly parallel to each other (fig. 2\b; pi. 11a, d). Individual bands fluctuate in width between 8 and 9 mm., depending on the amount of paint applied. Motif: The bands are painted on the shoulder below the rim. The interior of the lip may be painted with a wide (1.5 cm.) band (fig. 21 a). Additional decoration: T h e most characteristic decoration in this type is shallow incisions, averaging I cm. in length and 2-3 mm., in width, made on the rim interior below the painted band (fig. 21 a) . Wide painted bands may also occur on the exterior (fig. 2\b). T h e exterior of the rim may bear a red dab . CHRONOLOGICAL POSITION : This type spans both the San Lorenzo and the Chiriqui Phases, but is most popular in the levels of Site IS -3 in the San Lorenzo Phase. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: A vessel of Horconcitos Red Banded has been included by MacCurdy in the Red Line Ware (1911, Plate X X V , fig. d) . Its provenience is Divala, roughly halfway between David and Puerto Armuelles. Undoubtedly , this type belongs in the same Red Line Ware category of Holmes (1888) and M a c C u r d y (1911) as most of the other types in the San Lorenzo Phase. Linarte Zoned Red Line PASTE: Method of manufacture: Coiling. Coil junct ions have been largely obliterated from the exterior surface and usually from the interiors. In a few sherds, the interiors show traces of coil junctions in the form of faint depressions running parallel to the riins. Temper: A fairly " p u r e " clay containing small, rounded nodules of hemati te , natura l in this type of clay. In 4 CM FIGURE 21.?Type sherds of Horconcitos Red Banded, San Lorenzo Phase. Rim interiors are oriented to the left. Arrow on rim indicates red-slipped surface. Areas filled with black represent red-painted bands, a, Interior and profile of same rim sherd, b, Exterior surface of body sherd. WHOLE VOLUME CULTURAL REMAINS 33 addition, the clay contains a very small quanti ty of decomposed granitic sand. Texture: Ranges from very compact in sherds with " p u r e " clay to slighdy porous in those containing small amounts of decomposed granitic sand. No air pockets are visible in cross section. T h e breaks vary from clean and fairly straight to jagged edged. Color: Orange-buff to very light gray. A few of the sherds have been completely oxidized. In most cases, incom? pletely oxidized firing has left a band of buff color varying in width from 1 to 4 mm. along the outer sur? face and a gray band along the interior of the cross section. SURFACE : Color: Exteriors are a rich yellow to orange except in areas grayed by fire clouds. In a single vessel the color is relatively uniform. T h e interiors are either light buff or gray, that varies from very light to almost black. Treatment: Exteriors are always polished, but the degree varies from very intensive polishing done with a pebble when the clay was wet, so as to float the fine particles and give the appearance of having been slipped, to a less intense polishing that creates a less shiny appear? ance. T h e interiors were brushed, leaving faint striations running in varying directions. Hardness: Fairly soft, 2.0-2.5. FORM: Body wall thickness: Ranges from 5 to 9 mm. on a single vessel. Rim and Lip: Vessels of Linarte Zoned Red Line share the same general kinds of rims as the following three types: Arayo Polished Line, Centeno Red Banded, and Banco Red Line. These are everted rims with unmodified lips and diameters that vary depending on where the strap handles are attached. For a detailed discussion of these rims, see Arayo Polished Line. Base and Supports: Rounded bases slightly thickened so as to add some measure of stability to the vessel. Additional jeatures: All vessels of this type have a "s tepped" body wall (fig. 22c, f-i); that is, they have a shelf that extends 3-5 cm. below the neck, breaking the contour of the upper half of the vessel wall into two sections. Appendages: T h e lip tapers at one side, or at opposite sides of the vessel, into strap handles that decrease in width as they join the vessel wall (Strap Handle Modes a l - a 6 ) . For a complete description of these, see p. 55. Reconstructed vessel forms: Globular jars with rounded slightly thickened bases, a stepped vessel wall, and outflaring rims that extend into strap handles. DECORATION : Technique: Narrow (2 mm.) and wide (8 mm.) bands painted in red on a previously polished surface. Individ? ual bands vary in width from 1.5 mm. to 3.0 mm. Like all the bands in related types, they are wider at the top where the applicator was more steeped in paint than at the bottom where the paint thins out. Parallel lines are not equidistant; some are as close as 2 mm., while others are 4 mm. apart . Motif (fig. 22a-i; pi. \2a-h): A red band is painted around the lip. The upper half of the vessel is set off from the lower half by a band around the base of the neck and one or two wide bands that encircle the vessel at the shoulder line (as in Centeno Red Banded). The upper half is divided again into two horizontal sections by a band under the shelf or depression on the vessel wall. Wide, vertical bands separate the patterns of four to six thin (2-3 mm.) parallel fines that have been arranged so as to meet at right angles or to crosscut each other. In contrast to some of the decoration on Arayo Polished Line or in Centeno Red Banded, the thin lines of this type seldom follow parallel banded areas. There are two main motifs: 1. Narrow lines, in sets of four or five, meet each other at right angles, alternating first near the neck band, then at the shoulder band (fig. 22a-c). The sets of narrow lines may also meet along a horizontal plane at the center of the vessel (fig. 22^-^); an additional feature in some is a small, curving " t a i l " (fig. 22a; pi. \2d). The thin outer line in a set often has a row of dots or pendant small blobs (fig. 22a-b; pi. \2c,e). 2. Sections of narrow lines intersect rather than just meet. Sets of only three parallel lines may crosscut (fig. 22d), or the number of narrow lines in a set may be five or more (fig. 22/; pi. 12a). Rather than being arranged into bands, several lines (often more than 10 in number) may intersect to fill tr iangular spaces (fig. 22/; pi. \2h). Additional decoration: The strap handles that accompany this type have applique incised strips, punctations, and coffee-bean motifs. No applique decorations appear on the vessel wall itself. CHRONOLOGICAL POSITION : Linarte Zoned Red Line oc? curs late in the San Lorenzo Phase and continues into the Chiriqui Phase. It reaches its maximum popularity in the upper levels of Pit No. 3 at I S - 3 ; it also occurs in the upper levels of both SL-1 pits. Sites IS-7 and I S - l l have traces of this type. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: Neither Holmes (1888) nor MacCurdy (1911) include drawings of vessels that are exactly like the sherds of Linarte Zoned Red Line, but one of the vessels that MacCurdy (1911, Plate X X V , fig. a) has in his Red Line Ware is similar to figure 22/. Undoubtedly, the Linarte type belongs in the Red Line Ware of Holmes (1888) and MacCurdy (1911). One sherd of this type has been recorded as a trade sherd in the Diquis area of Costa Rica (Lothrop, 1963, fig. 69,d). Its painted motif is exactly like that appearing in our figure 22(?. Pan de Azucar Red Line PASTE: Like Linarte Zoned Red Line; see that type description for details. 284-757 0-68?4 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 I I I I 1 I 0 I 2 3 4 5 CM FIGURE 22.?Body wall profiles a n d decorat ion on Linar te Zoned R e d Line, San Lorenzo Phase. D a r k areas ind ica te red thick a n d th in bands . W H O L E VOLUME CULTURAL REMAINS 35 0 I 2 CM FIGURE 23.?Type sherds of Pan de Azucar Red Line, San Lorenzo Phase. Rim interiors are oriented to left. Arrow on rim indicates red-slipped surface. Dark bands on exterior are red painted. SURFACE: Color: An even buff color; fire clouds rare or absent. Treatment: A low polish that leaves faint horizontal stri? ation marks. Hardness: Between 2.5 and 3. F O R M : Body wall thickness: Consistently thin walled, averaging 5 mm. Rim and Lip: Incurved direct rims belonging to shallow, open bowls. T h e rim is slightly thicker near the mouth than lower on the body wall (fig. 236). Bases and Supports: Rounded unthickened bases. No supports. Appendages: None. Reconstructed vessel forms: Rounded bowls with curving or slightly angular shoulders, constricted mouth, and un? modified base and rim. DECORATION : Technique: Th in red lines that vary in width individually from 1-2 r rm. decorate the exterior of the bowl be? tween shoulder and rim. Overshot intersections are characteristic. Motif: A thin line encircles the bowl at the shoulder and around the lip (fig. 23 a, b). Parallel lines that are not equidistant, but vary from a couple of millimeters to over two centimeters apart , are drawn between the lip and the shoulder band. They may cross over the band. CHRONOLOGICAL POSITION: Pan de Azucar Red Line spans the San Lorenzo and the Chiriqui Phases; its largest percentages are in the San Lorenzo Phase. I t is not the predominant or even a popular type in the San Lorenzo or the Chiriqui Phases. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: This type belongs to the general " R e d Line W a r e " of Holmes (1888) and MacCurdy (1911), but in neither of of their reports is there an exact duplicate of a vessel of this type. Zapo te Red Banded PASTE: Method of manufacture: Probably coiling, though all tiaces of coil junctions have been obliterated from both visible surfaces. Temper: A homogeneous " p u r e " clay containing very little feldspar, but large amounts of small rounded hematite particles up to 0.5 mm. diameter, which occur naturally in this type of clay. Texture: Compacted because of the lack of extraneous temper particles added to the clay. Small air pockets are occasionally visible in cross section. Sherds break along fairly straight lines. Color: Incomplete oxidized firing has produced a light, gray unoxidized band varying in thickness from 1-8 mm. on 75 percent of sherds. Band may be on interior, exterior, or center of the cross section. In about one- fourth, complete oxidation has produced an even buff color throughout. SURFACE: Color: About half of the sherds are an even light buff or orange-red color on both surfaces. The other half are light to dark gray on both surfaces or, more com? monly, only on the exterior. Dark fire clouds occur on most sherds around the bend of the lip, suggesting the vessels were placed upside down during firing. Treatment: Both the interior and the exterior surfaces were polished with either a rounded pebble, that left wide striation marks, or a more pointed tool. T h e degree of polishing varies markedly depending on the tool used. The use of a pebble has compacted the surface producing an even, smooth texture in most of the sherds. In the others, incomplete polishing with a narrower tool (probably the point of a stone) left puck marks and air pockets between the smoother polishing striations. Hardness: Fairly hard, 4. FORM: Body wall thickness: 0.6-1.1 cm. Rim and Lip: All belong to outflaring plates or open bowls, differing slightly in contour and lip treatment. Major groupings: 1. Outflaring straight-sided direct rims with modified lips. Rim may be slightly thickened and end in a rounded lip (fig. 24/z), or it may be pronouncedly thickened and have a flat lip with a small hook in the interior or exterior (fig. 2^b,g). Thickening was probably done by raising the clay toward the tip with an upward pressure of the fingers, rather than by adding extra coils; no traces of coil inarks or junctures are visible anywhere. R im diameters range between 20 and 24 cm. 2. Rims belonging to open, shallow bowls. They are markedly thickened so that the rim is twice as thick as the body wall (fig. 24a). R im diameters range between 28 and 30 cm. 3. Unmodified rims belonging to deep bowls. T h e lip may be either flat at the top with an angular bent in the interior and rounded on the exterior (fig. 24c, d) or simply rounded (fig. 24^, / ) . Rim diameters range between 18 and 23 cm. Base and Supports: No whole vessels were found, but ringstands and pedestal bases of the same paste and decoration (bands and chevron designs in the interior) 36 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY V O L U M E 8 0 I 2 3 4 CM FIGURE 24.?Rim profiles and designs on the interior of Zapote Red Banded, San Lorenzo Phase. Hachure indicates red- painted bands. Arrows around rim profile indicate red slip. All rim interiors are oriented to the left. First drawing in group a is of an exterior. as Zapote Red Banded occur in levels with rims and body sherds of this type. They may belong together. See the illustrations and description of the following modes: (1) Ringstand, Mode b (p. 51) and (2) Pedestal Base, Mode i (p. 53). Reconstructed vessel forms: 1. Deep, outflaring, straight-sided bowls resting on ringstands or low pedestals. Walls are of fairly even width (8 m m ) . Rims are thickened (1.2-1.4 cm.) and have flat tops. R i m diameters range from 21-24 cm. 2. Deep open bowls with concave walls terminating in rounded or flat lip. Bowls rest on pedestal bases or ringstands. R im diameters are 18-23 cm. 3. Shallow open bowls with markedly thickened rims, twice as wide as the body wall, flat lips and ring- stands, or pedestal bases. DECORATION : Technique: Red-painted bands on the unslipped but polished exposed interior surface. Bands are of two general widths: 8-10 mm. (fig. 24c; pi. lOa-c, e-h) and 2-3 m m . (fig. 24^^; pi. \Qd). Individual bands are of uniform width. T h e paint varies in density from complete coverage of the underlying surface to a thin coat that reveals, in spots, the underlying surface. T h e edges of the bands are more t ransparent t han in the center. Some bands are fairly even and well d r awn , while in others the edges are jagged from careless application of the paint . Exteriors are mosdy unsHpped. A few sherds are slipped with a bright red on the exterior surface and on the l ip; the interiors of these are left undecora ted (fig. 24^). Motif: Evenly spaced, parallel bands , alone or combined with zones of solid color. Lips are always pain ted in red ; paint occasionally continues over the h p and merges with a band I cm. wide pain ted a round the exterior of the r im. Typical a r rangements a r e : 1. Parallel bands, d r awn at r ight angles to the lip (fig. 24a, c, d, h; pi. 10a, b, d, e, g, h). I n complete specimens, the bands would meet at the center to form chevron designs. T h e bands m a y be wide (0.7-1.0 cm.) or nar row (2-3 mm. ) as in figure 24(f, plate 10(f, and they m a y be closely spaced or far apar t . 2. Red painted bands placed obliquely to the r im so tha t they meet before the center of the vessel to form concentric t r iangular designs (pi. 10c , / ) . 3. A less typical a r rangement is for the interior of the bowl to be left undecorated and for the exterior to be covered with a rich red sfip tha t extends over a wide, flat Up. T h e slip is polished with a pebble, forming a hard surface and leaving pofishing stria? tions parallel to the r im (fig. 24 i ) . CHRONOLOGICAL POSITION: Zapote Red Banded occurs in all three Chiriqui Gulf Phases, but is most popular in the San Lorenzo Phase at SL-1 site. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: An open bowl of this type, decorated in the interior with chevron designs and resting on a perforated pedestal base, appears in MacCurdy ' s report (1911, fig. 159). T h e provenience he assigns to this vessel is El Banco. No other vessels of this kind have been por t rayed in the Chiriqui l i terature. Flaring bowls are a very common form, however, in the pottery of Sitio Conte , especially in the Early C o d e Polychromes (Lothrop, 1942, Vol. V I I I , p . 9). T h e r im profiles of the C o d e bowls are not unlike those of Zapote Red Banded; the C o d e bowls also rest on ringstands like the ones tha t occur associated with this type at I S - 3 . I t is not farfetched, therefore, to suggest that our flaring bowls may have been inspired by forms from the C o d e area. T h e geometric chevron pat terns may likewise be a simplified version of the curvilinear poly? chrome designs of Early and Late C o d e . Late and possibly Early t rade sherds occur in the San Lorenzo Phase of the Chiriqui Gulf. U n c l a s s i f i e d P l a i n W a r e s At S i te I S - 3 p l a i n she rds b e l o n g i n g to t h e levels p l a c e d in t h e S a n L o r e n z o P h a s e w e r e s o r t e d i n t o g r o u p s t h a t s h o w e d a d i s t inc t ive t r e n d in t h e p e r ? c e n t a g e f r e q u e n c y c h a r t s for t h a t site (fig. 4 ) . S o m e of these g r o u p s s h a r e d specific a t t r i b u t e s of p a s t e a n d WHOLE VOLUME CULTURAL REMAINS 37 sur face t r e a t m e n t w i t h c e r t a i n d e c o r a t e d types a n d in s o m e cases m i g h t b e t h e b a d l y e r o d e d surfaces of a s l ipped t y p e , w h e r e t h e sl ip h a s to ta l ly d i s a p p e a r e d . P L A I N W A R E D . ? I d e n t i c a l in pas t e compos i t i on to C a c o R e d S l i p p e d (p . 26). T h e c l ay h a s misce l l aneous pa r t i c l e s of v e g e t a b l e m a t t e r in it t h a t a p p e a r to b e t h e r e s u l t of n a t u r a l a d m i x t u r e r a t h e r t h a n t h e con? scious a d d i t i o n as a t e m p e r . T e x t u r e is g r i t ty w i t h l a r g e a i r pocke t s d i s t r i b u t e d t h r o u g h o u t t h e pas te . T h e ex t e r io r h a d b e e n c o m p a c t e d b y pol i sh ing w i t h p e b b l e s . Su r face co lor a n d p a s t e color is buff. P L A I N W A R E H . ? S a m e pas te as C e n t e n o R e d B a n d e d ( p . 31) . Dis t inc t ive ly th is c lay is ve ry p u r e . w i t h v e r y smal l r o u n d e d par t ic les of h e m a t i t e t h r o u g h ? o u t t h e w e l l - c o m p a c t e d pas te . Co lo r of surfaces a n d cross sect ion of she rds is a l i g h t c r e a m . P L A I N W A R E I . ? I d e n t i c a l i n al l fea tures to P l a i n W a r e D e x c e p t in color a n d firing. Cross sect ion of sherds is a d a r k g r a y . P L A I N W A R E J . ? S t r i k i n g l y different f rom all o t h e r p l a in w a r e types in several features . Al l sherds a r e th ick , some over 3 cen t ime te r s , w i t h a l igh t -co lored pas te m i x e d w i t h par t i c les of a n g u l a r , c r u s h e d q u a r t z r a n g i n g u p to 2 m m . in size. T h e t e m p e r is ve ry a b u n d a n t a n d a p p e a r s o n t h e surfaces so t h a t t h e surface t e x t u r e is g r i t ty a n d coarse . CHIRIQUI PHASE Cavada App l ique and Red Banded PASTE AND SURFACE: Like Zapote Red Banded; see that type description for details. F O R M : Body wall thickness: Ranges from 6-10 mm. Rim and Lip: O n e of the features by which the type can be differentiated from Zapote Red Banded is in the t reatment of rims and lips. Most of ihe rims of Zapote Red Banded are thickened in the interior and flattened at the top, and the exterior is left unmodified; with few exceptions, the rims of Cavada Applique and Red Banded are modified on the exterior. The lip is flattened on the outside so as to form a short (.8-1.5 cm.) per? pendicular shelf (or a sort of neck) around the rim {^g. 25 a-b,d,g,h). T h e rims may vary markedly in thickness, from 6 mm. to 12 mm. (fig. 25 ; cmp. a and g) ; variation is con? siderable within a single vessel. T h e r im may be thicker than the body wall (fig. 25a-c) or the reverse (fig. 25h). Bases and Supports: Like Zapote Red Banded, vessels may rest on ringstands and (or) pedestal bases, but this association has not been confirmed. Reconstructed Vessel Forms: 1. Deep, outflaring, straight-sided (fig. 25d) or curved- sided (fig. 25a) bowls with lips that are flattened on the exterior and unmodified in the interior. 2. Shallower bowls of the same contour as the deeper bowls (fig. 25^, c , / , h). DECORATION : Technique: T w o techniques are predominant : red paint and applique. They may be used separately or in conjunction with each other. 1. T h e painted designs are narrow bands, 4-6 mm. wide, put on ra ther sloppily with an applicator that was unevenly steeped in paint. 2. T h e applique designs consist of narrow rolls and pellets of clay that have been pressed into the vessel on the flat outside of the lip to form zoomorphic designs. Motif: 1. Crosshatch of red bands covers par t of the vessel interior and extends over the lip to cover the flat band on the r im exterior (fig. 25^, ^; pi. 11 f-h). 2. Most of the applique designs on the flat surface of the lip exterior are incomplete (fig. 25a-d\ pi. I h ) - W h a t remains are small, curved bands with per? pendicular incisions on them; they may have been arm motifs. The two complete applique elements are zoomorphic. One is a representation of a crablike animal (fig. 25^; pi. 11/), rendered by clay pellets incised in the center to represent eyes, mouth, and claws. The other is a birdlike head that protrudes from 4-6 mm. from the curvature of the bowl (fig. 25/). Incisions on the side of the head represent the beak; the eyes are also rendered by incisions. Designs on the vessel wall may represent wings. Additional decorations: Two of the rims decorated with applique designs also have red painted decorations. In one instance the paint is applied to the top and to the exterior of the lip (fig. 25^); in another instance the painted band is only on the top of the lip (fig. 25a). CHRONOLOGICAL POSITION: Cavada AppHque and Red Banded occurs in the most recent of the Gulf Phases, the Chiriqui Phase. I t occurs in very small percentages in the top levels of Pit No. 3 at IS -3 and in the middle levels of Pit No. 2 at I S - 7 ; it is slightly more popular in the bottom levels of Site l S - 1 1 . GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: The bowls of this type are probably an outgrowth of Zapote Red Banded. The Stirlings found a rim of a shallow bowl of this type (Stirling, 1964b, fig. 23, p. 274) in one of their sites in Bocas del Toro (site BOC-2 ) . It is identical to figure 25^. 38 S M I T H S O N L \ N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 0 1 2 3 4 CM FIGURE 25.?Rim profiles and exteriors of Cavada Applique and Red Banded, Chiriqui Phase. Rim interiors are oriented to the left. Hachure indicates red paint. T a r r a g o Bisquit W a r e PASTE: Method of Manufacture: Probably coiling; no traces of coil junctions remain on either surface. Temper: T w o slightly different kinds of temper ; one occurs in Villalba (Isla Muertos) and another at Isla Las Secas. 1. At Villalba ( IS -7 ) : Minute particles of pumice, also small pinpoint inclusions of a black and shiny, laminated mineral (probably inagnatite). 2. At Las Secas ( I S - l l ) : The added fraction is coarser and consists of particles of angular crushed quar tz and other crushed rocks of different kinds. Texture: Porous and friable. Loosely packed with only a few air pockets. Breaks irregularly with jagged contours. Color: A very homogeneous buff is characteristic of the variety occurring at IS -7 . Reddish and grayer shades occur at I S - l l . In any one vessel of either variety, color is uniform due to completely oxidized firing throughout the cross section. SURFACE: Color: Ranges from buff" to reddish orange to pale gray. No fire clouds, except in rare instances, occur in the IS-11 variety. Treatment: T h e exterior surface has been lightly smoothed by rubbing with some object, or with the hand, so that tiny temper particles have been dragged across, leaving hardly noticeable striations. The re is a sandpapery feel to the surface. T h e interior surface has been left totally unsmoothed ; temper particles, puck marks, and scattered imperfec? tions make the interiors coarse and abrasive to the touch. Hardness: Varies from very soft (2) to slightly ha rde r (2.5). F O R M : Body wall thickness: Characterist ically very t h in : 4 -6 m m . Rim and Lip: T h e variety of rims tha t occur in the Bisquit Wares is enormous (see "Armadi l lo W a r e , " M a c C u r d y , 1911, pp . 4 8 - 7 1 ; " T e r r a Cot ta W a r e , " Holmes, 1888, pp. 67-80) . Here I will describe only r im variants in Ta r r ago Bisquit Ware from IS -7 and at I S - l l : 1. Outflaring, short rims from globular j a r s : (a) Straight, short (2-3 cm.) rims, wi th exteriorly folded lips. R i m and neck m a y be thin (4 mm.) or thick (7 m m . ) ; (fig. 26 a, b; pi. \?> a, d,f). R i m diameters range from 8-12 cm. (b) Short (1.3-2.6 cm.) outcurved rims (fig. 26 c-e; pi. 13^). Lip may be exteriorly thickened or left unmodified. R i m diameters range from 6-8 cm. (c) Sharply everted and short (2-3 cm. ) , wi th markedly thickened ovoidal lips tha t have flat tops (fig. 26 n, o). R i m diameters 14-16 cm. 2. Outflaring rims from globular jars , sHghtly longer (3.5-5.0 cm.) than group I. (a) Rims may be gradual ly outflared with markedly or moderately tapered fips (fig. 26p-q; pi. \3e). R i m diameters 12 and 14 cm. (b) Sharply outflared with thickened lips tha t are flat at the top (fig. 26s) or unmodified fips (fig. 26t). R i m diameters are 12 and 14 cm. A var iant is a slightly less sharply outcurved r im (fig. 26u) with a diameter of 22 cm. (c) Long (5.5 cm.) , gradual ly outflared, wi th an exteriorly thickened lip and r im diameter of 30 cm. (fig. 26i;). 3. Outslanted rims belonging to fairly deep bowls with unmodified lips; rims are long (4 .0-5.5 cm. ) . T h e lip may be thicker than the vessel wall (fig. 26/) or it may be of uniform w i d t h through? out (fig. 26m; pi. 13^). 4. Direct rims belonging to bowls: (a) Rims from deep bowls with incurving sides. Lip may be unmodified (fig. 26/) or interiorly thickened with a flat top (fig. 26^). R i m d i a m ? eters are 16-18 cm. respectively. (b) Rims from shallow bowls with outcurved sides. Lip may be unmodified (fig. 26A:) or interiorly thickened (fig. 26;, j). R i m diameters are 14, 18, 15 cm. respectively. (c) A type that is popular at Site I S - l l , bu t absent from IS -7 , is an inslanted or notched r im (fig. 26/;; pi. I3b) with a d iameter of 16 cm. Base and Supports: I. Globular jars and deep bowls have rounded to fairly pointed bases of the same thinness as the vessel walls. Flat bases are very rare. WHOLE VOLUME CULTURAL REMAINS 39 I c H FIGURE 26.?Rim profiles, side views, and adornos of Tarrago Bisquit Ware, Chiriqui Phase. Rim interiors are oriented to the left. 40 SMITHSONLVN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 2. T w o kinds of supports have been found in our Ta r r ago Bisquit W a r e : (a) Hollow mammiform tripods with pellets in the interior (Mode f l ) ; turn to the section on tripod feet for a complete description, and see page 67 for the distribution, (b) Short pedestal bases (Pedestal Base, Modes f, m ) ; for a complete description turn to the section on pedestal bases, and for the distribution see page 67. Appendages: Vessels of the Bisquit Ware rarely have handles (see Holmes, 1888, or MacCurdy , 1911). None were found in the Ta r r ago variant of the Bisquit Ware . Reconstructed vessel shapes: T h e vessel shapes of the "Armadil lo-Terra-Cotta-Bisquit W a r e " (of which our Tar rago Bisquit Ware is part) are enormously varied. They have been described by Holmes (1888) and Mac? Curdy (1911). All of the following shapes occur: deep or shallow, open or constricted hemispherical bowls with incurving or outcurving sides; fairly deep, angular bowls with flaring sides; globular jars with constricted mouths and either short necks or no necks, and flaring r ims; some of these may have two orifices. Obviously, we did not find the full range of form variation in our small excavations. T h e following shapes are represented in our Tar rago Bisquit W a r e : 1. Globular jars with short, straight, or everted rims that have unmodified or exteriorly thickened lips. Rounded bases. 2. Globular jars with flaring, relatively long rims and rounded bases. 3. Deep bowls with insloping walls and either un? modified or thickened rims. Bowls of this type may end in rounded bases or may rest on short pedestals. 4. Shallow bowls with outsloping walls and either plain or unmodified lips. They may have rounded bottoms or may rest on either tripod feet or pedestal bases. DECORATION : Technique: Generally unslipped, though there are a few sherds with a light salmon-color polished slip that looks almost indistinguishable from the paste color. They are definitely in the minority. MacCurdy put them into a "Salmon-colored Sub-group," par t of the Armadil lo group (MacCurdy, 1911, pp . 71-72). All decoration is plastic, consisting of small modeled and applique zoomorphic and anthropomorphic ador? nos, applique fillets and nodes, incisions, and puncta? tions. Motif: T h e following adornos were found in our ex? cavations: I. Anthropomorphic figurine of the upper half of a creature with a tall headdress and some object in the mouth (fig. 26w; pi. 130- T h e face was built up by apphcat ion of pellets that were pushed in to form the ears, the nose, the eyes, and the mouth. A shallow incision divides the eyes and the mouth in half. T h e torso of the figurine is too eroded to reconstruct. 2. T iny figure of a h u m a n sitting down against a flat back, with arms crossed at the knees (fig. 26x). T h e body was built up by thin rolls of clay tha t were later incised to portray toes, arms, and legs. T w o very small (2 mm.) coffee-bean eyes and a p ro t rud? ing nose complete the face. Like all the adornos of this type, this figurine is stylized and crudely rendered. Its body parts and its position are nevertheless perfectly identifiable. 3. Tiny figurine of a m a n squat t ing down, wi th arms crossed on the knees (fig. 26;-; pi. 13A). The re are small punctations and incisions on the arms and on the legs. 4. Zoomorphic head, perhaps of an armadil lo (fig. 26^) and a tiny design of a frog seen from the back (fig. 26aa). CHRONOLOGICAL POSITION: Ta r r ago Bisquit W a r e occurs in the Chiriqui Phase, the most recent phase in the Gulf. It is present in small percentages in the top levels of Site I S - 3 . T h e largest percentages are at Sites I S - 7 and I S - l l . COMPARATIVE MATERIAL: In museum collections of Chi? riqui pots the world over, there are thousands of vessels of this type; their only rival in popular i ty is the Villalba Red Streaked type (called in the l i terature T h e Fish- Tr ipod-Handled Ware ) . Ta r rago Bisquit Ware was for the first t ime described in detail by Holmes (1888, pp . 67-80) , who called it " T h e Ter ra-Cot ta G r o u p . " M a c C u r d y (1911, pp . 48-72) renamed the type "Armadi l lo G r o u p " because he found the recurrent motif to be the whole or parts of the armadillo animal . He also subdivided the group into the "Armadi l lo" proper and the "Sa lmon Coloured Groups" (op. cit., pp . 71-72). Osgood (1935) combined their names into "Armadi l lo Ter ra Cotta W a r e . " According to Habe r l and (1958, p. 343): He [Osgood] also demonstrated that there is no reason to divide this ware into two sub-groups like MacCurdy's, since the difference in colour shade is due to variable firing temperatures. Neither Osgood nor Habe r l and are correct in affirming that there are no more than firing differences in the type. Some sherds are definitely slipped in a salmon color, bu t this distinction appears to have little tempora l or spatial importance. O u r Ta r r ago Bisquit W a r e is, beyond doubt , the same as the "Armadif io-Terra-Cot ta Bisquit W a r e " of Holmes (1888), M a c C u r d y (1911), Osgood (1935), and Haber ? land (1958). By retaining the term "Bisquit W a r e " and adding the name Ta r r ago , I hope to convey the fact that it belongs to the same group as the one described in the li terature. Osgood (1935, p . 240) summarizes the main land distribution of the Armadi l lo-Terra-Cot ta Bisquit W a r e as Bugavita, Divala, and a place 30 miles west-northwest of the city of David. W H O L E VOLUME CULTURAL REMAINS 41 T h e Ta r r ago Bisquit Ware bears undeniable affinities to ceramics that occur in the Diquis area of southern Costa Rica . Photographs of whole vessels from the General Valley, Potrero Grande , Cafias Gordas, and the Osa Peninsula (Stone, 1958, fig. 6, d~g and fig. 7a) are identical to the whole vessels of the Armadillo- Ter ra-Cot ta Bisquit Ware described in Holmes (1888) and M a c C u r d y (1911). Villalba Red Streaked PASTE : Method of manufacture: Coiling. Traces of coil junctions are clearly visible on the interior as marked depressions or ridges. Temper: Coarse, sand-sized fractions added to the clay compose about 10 percent of the paste. Pinpoint particles of angular quartz and feldspar, the most abundant minerals, show up in cross section as white specks clearly visible to the naked eye. They are distributed homo? geneously throughout the paste. Minor amounts of magnati te and hornblende are also represented. Texture: Friable, crumbly, and loosely packed, with fairly large (2?4 mm.) air pockets frequently visible in cross section. Breaks irregularly with jagged con? tours. Color: Ranges from orange-red to gray-buff or dark gray, sometimes almost black, depending on the degree of oxidation. Firing is variable. Some cross sections show a gray core up to 75 percent in width of the sherd body wall, with the exterior and interior outlined by a light-colored band, 1-2 mm. wide. SURFACE: Color: T w o distinct groups: 1. Charcoal gray to almost black due to large fire clouds that nearly cover the entire vessel. 2. Buff to light gray with small or no fire clouds. Treatment: Unpolished. T h e exterior of the vessel and the interior of the r im have been rubbed so that the surface is fairly smooth; the vessel interior has been left almost untouched. Surfaces are abrasive and have scattered imperfections. Both the interior and the exterior surfaces undulate irregularly, suggesting that coil junctions were incom? pletely worked into the clay by applying uneven pres? sure with the hand. Traces of coil junctions remain as crude folds on the exterior of the r im and under the shoulder. Hardness: Soft, 2 .5-3. F O R M : Body wall thickness: Two groupings: thin (5-7 mm.) or thick (9-13 mm. ) . An individual vessel may encompass the range of variation. Rim and Lip: All rims are sharply everted and their lips thickened exteriorly by the crude application of coils. Rims vary in length, lip treatment, thickness, and diameters. There are at least two groupings: 1. Large, everted rims (5.7-7.0 cm. tall) curving out? ward directly from the vessel shoulder (fig. 27 a, b). Opposite ends of the r im are frequently carried upward and extended into large strap handles; if the r im has broken off the vessel at the point where the straps commence, one side may be as much as 2.5-3.0 cm. taller than the other (fig. 27 a, c). In the majority of the cases, the lip has been thick? ened exteriorly by the addition of a coil. Another coil may have been added below the lip. Coil junct ions were often left untouched so that they show on the outside as folds varying in width between 0.7 and 2.5 cm. (fig. 27 b, e,f; pi. 13/). In other sherds there was some at tempt to obliterate the coils, (fig. 27^). A few of the tall rims have unmodified lips (fig. 27a), but these are definitely in the minority. Mou th diameters vary from 22 to nearly 28 cm. 2. Shorter rims (1.3-3.5 cm.) everted at a sharper angle than group I rims and thicker in comparison. There are several variations: (a) Sharply everted, flat at the top, with an angular, tapered, unmodified lip (fig. 27^). R im diameters average 17 cm. (b) Gradually everted with a tapered unmodified lip (fig. 27 -^, t). R im diameters 16 and 18 cm. (c) Very thick and short with either unmodified (fig. 27z, k) or thickened lip (fig. 21 h, j , I, m). Rim diameters 14 and 16 cm. (d) Short and thinner rims with modified lips and diameters averaging 14-15 cm. (fig. 21n-p, x,y). (e) Slightly longer shafts (average 3.0 cm.) and modified lips (fig. 27^, r, u-w). R i m diameters range from 16 to 20 cm. Base and Supports: Two clear-cut groups: 1. Concave, rounded bases devoid of any supports. A certain stability to the vessels is afforded by a thickening of the base. 2. Vessels with tall, hollow tripod feet in the shape of fish (Mode e, variations e l - e4 ) . See description under Tripod Feet. Appendages: 1. Round Handle, Mode b ; for a complete description, see p. 56. 2. Strap Handle, Mode b ; for a complete description, see p. 55. Reconstructed vessel forms: Two general groups: 1. Globular jars, wider than they are tall, with rounded, thickened bases, and sharply everted rims that may be tall (5.7-7.0 cm.) or short (1.3-3.5 cm.). A few have no handles. The majority have strap handles that connect the lip with the shoulder or round handles vertically placed between lip and vessel shoulder; some round handles are placed horizontally between two points on the shoulder. 2. Tall tripods (20-30 cm.) with small bodies and tall hollow legs in the form of fish. Short, sharply everted rims. 42 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 FIGURE 27.?Rim profiles, exterior and interior views, Villalba Red Streaked, Chiriqui Phase. Hachure indicates red bands. Rim interiors are oriented to the left. WHOLE VOLUME CULTURAL REMAINS 43 DECORATION : Technique: A bright orange-red paint was sparingly ap? plied with a wide brush to the exterior of the vessel, rarely covering the whole surface. In some the paint rubbed down with the hand before it dried, to produce a streaked appearance. Painted bands are carelessly exe? cuted; individual ones vary in width between l . l and 2.0 cm., and their edges are irregular. T h e r im exterior is rarely streaked in red; interiors of rims are painted completely or in part. The paint extends beyond the angular curvature of the shoulder. Motif: 1. Either red bands or streaks are painted on the exterior of the vessel walls. Variations include: (a) Wide, carelessly drawn bands varying in width between 1.2 and 2.0 cm., applied singly or in pairs, encircling the vessel at its inaximum diameter ; large loops or triangles, facing up or down, are drawn from the horizontal bands to the neck of the vessel. (b) A series of short streaks or smudges. 2. T h e interior of the rims is always decorated, either covered entirely with red, or with a wide (1.5 cm.) band of red that extends over the lip (fig. 27^). This band may be placed around the bottom of the r im; it may be extended for fractions of a centimeter below the vessel shoulder. Some rims have slightly more complicated arrangements of intercrossing bands (fig. 27a, interior; pi. 13;72) or parallel bands (fig. 27^). CHRONOLOGICAL POSITION: Villalba Red Streaked is one of the two most popular types diagnostic of the Chiriqui Phase, the last or most recent of the phases in the Gulf. It occurs in largest percentages at Site I S - 7 ; it appears, though in smaller quantities, from bottom to top of the stratigraphic column at Site IS-11. At Site IS -3 , it occurs in the two top levels; it is absent from S L - 1 . GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: Villalba Red Streaked was first described by Holmes (1888, pp. 91-109), who divided the type into two groups: the "Handled G r o u p " and the "Tr ipod Group ." MacCurdy (1911, pp. 76-92) kept Holmes' distinction between the two groups and renamed them respectively "Pa in ted" and "Unpa in ted Handled W a r e " and the "Fish Tr ipods ." Keeping the two groups separate, he admitted to be merely a matter of descriptive convenience (op. cit. p . 77): It is true that many vessels of similar material, form and finish are mounted as tripods; but these are placed for convenience with the tripod or fish group. Osgood (1935, p . 237) recognized the affinities between these two groups and combined them into one type which he called "Fish-Tripod-Handled Ware . " Haberland (1958, pp. 342-343) reaflfirmed the fact " tha t the groups, formerly used by MacCurdy, are only aspects of the same pottery type." Sherds of our Villalba Red Streaked type ai-c indisdnguishable from the "Fish-Tripod-Handled W a r e " that occurs in mainland graves. I have changed the name because most of the specimens are not fish tripods, and some have no handles. T h e distribution of this ware on the mainland has been summarized by Osgood (1935, p . 240) as being Bugavita, Divala, and a point 30 miles west-northeast of the city of David. I t occurs in other areas as well. U n c l a s s i f i e d P o t t e r y T y p e s T h r e e of t h e fol lowing types h a v e n o t b e e n g iven n e w type n a m e s because t h e y e i t h e r r e p r e s e n t local im i t a t i ons of foreign w a r e s or a r e t r a d e i t ems . T h e o t h e r two , " A l l i g a t o r W a r e " a n d " B l a c k a n d R e d N e g a t i v e , " a r e w e l l - k n o w n "C la s s i c C h i r i q u i " p o t t e r y types . S ince t h e y a p p e a r in o u r s e q u e n c e in such sma l l n u m b e r s , t hey m a y h a v e b e e n b r o u g h t in as t r a d e s f rom a d j a c e n t m a i n l a n d coas t a n d h i g h l a n d s of C h i r i q u i w h e r e t h e y a r e so a b u n d a n t . S t r a t i g r a p h i c work in these o t h e r a reas is necessary before r e n a m i n g or redef in ing these types . T h e y a r e b e i n g cursor i ly desc r ibed for t h e p u r p o s e of c o m p a r i s o n s . All igator W a r e Only one sherd of this type occurs in Level 60-70 cm. of Pit No. 1 in Site IS-11 (fig. 28; pi. 14A). Paste and surface treatment are like Cavada Applique and Red Banded, which also belong in the Chiriqui Phase. DECORATION : The decoration is its most salient character? istic. The background color is a white or cream-colored slip on which designs are painted in red and black. The lip and the designs on the exterior under the lip are painted with a black band. Below the design is another band circling the vessel; it is painted in red. The motif consists of black bands that are about 4 mm. wide, arranged in triangles with pendant round dots in the center. 0 I 2 CM FIGURE 28.?Sherd of Alligator Ware, Chiriqui Phase (especially diagnostic of "Classical Chiriqui Culture"). Red areas in? dicated by horizontal areas, white by perpendicular broken lines, and black areas are filled with black. 44 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: T h e alligator group is one of the four most abundan t of the Classical Chiriqui Wares. I t has been described in detail by Holmes (1888) and MacCurdy (1911). Osgood (1935) worked out the distribution as being Bugavita, Divala, and an area 30 miles west-northwest of the city of David. T h e affiliations of this ware are to the east with the Diquis area of Costa Rica. Lothrop (1963, p . 78) has named the Costa Rican variant the Polychrome Group (Black-and-Red-Line), rather than Alligator Ware as in Chiriqui , because he sees slight divergences in design, size, and shape. His report (op. cit., pp . 78-87) should be consulted for a detailed comparison of the Alligator Ware of Chiriqui and the Polychrome Group (Black- and-Red-Line) of the Diquis Delta. Black-on-Red Negative This is another of the popular "Classical Chir iqui" po t te ry types. In the Gulf its populari ty is very small; it occurs only in the top levels of Site I S - l l . In paste and surface t reatment , it is exactly like Villalba Red Streaked; see t h a t type description for details. DECORATION : The technique of Negative Paint ing is well known in the New World. MacCurdy (1911, pp . 104-106) describes it in detail in reference to the Chiriqui Lost Color Group. T h e motif of the few sherds found in the Gulf Site IS?11 is simple and geometrical. I t consists of black horizontal and vertical dipped bands on a red-orange slipped surface. The bands alternate with larger zones of black (fig. 29; pi. 14/). G E O G R A P H I C A L DISTRIBUTION AND COMPARATIVE MATERIAL: W h a t we call Black-on-Red Negative is the old Lost Color Group of Holmes (1888) and M a c C u r d y (1911). They describe complete vessels of this ware in their respective reports. Osgood (1935) kept this ware as one of the basic four pottery types of the "Classical Chiriqui cul ture ." He worked out its distribution as being the same as the other three basic Chiriqui Wares : Bugavita, Divala, and an area 30 miles west-northwest of the city of David. Haber land (1958, pp. 340-341) argues that the Lost Color Ware is a catchall category which includes four different types: Negative Ware , Black-on-Cream Neg? ative, Black on Red Negative, and Red on Red Negative. T h e significance of all of these divisions is not clear; they seem unnecessary unless they convey geographical or chronological information. In the Diquis region there is a ware which is very similar, if not identical, to the Negative Painted Type of Chiriqui . Of the Costa Rican ware. Stone (1958, p. 47) remarks: Globular ware with negative painting is common throughout the region (Diquis), particularly in the uplands. Here it is found with a red slip and black dip as in the delta, and with a cream slip and black dip. 2 CM FIGURE 29.?Sherd of Black-on-Red Negative, Chiriqui Phase (especially diagnostic of "Classical Chiriqui Culture"). Ha? chure indicates red area; black areas are left blank. Chocolate Incised PASTE: Method of manufacture: Probably coiling, though no traces of coil junctions remain on either surface. Temper: Pinpoint particles of angular quar tz distributed homogeneously. Texture: Fairly smooth; breaks cleanly, bu t not along coil junctions. Color: An even light buff caused by complete or almost complete oxidation. S U R F A C E : Color: Light gray to beige. Treatment: Polished to produce an even surface. Hardness: 2.5 to 3. F O R M : Body wall thickness: Characteristically thin, averaging between 4 and 6 m m . Rim and Lip: Short rims belonging to open bowls. Rims are fairly straight, curving inward only when they join the vessel shoulder. T h e lip is slightly enlarged so that it forms a small ledge a round the mou th of the vessel. Base and Supports: Tr ipod feet in the shape of feline heads (see description of Tr ipod Feet M o d e i l ) . Major vessel forms: A shallow open bowl wi th very thin walls resting on hollow tr ipod feet conta ining pellets. WHOLE VOLUME CULTURAL REMAINS 45 DECORATION: A very thick, reddish chocolate slip is appHed on both surfaces of the bowl to cover the entire area. Then the outside of the rim above the vessel shoulder is decorated with fine shallow incisions drawn when the surface was very hard prior to firing. They are arranged in a motif of concentric triangles around the entire exterior of the vessel. The front part of the tripod foot is also decorated with shallow incisions (see description of Mode i l , p. 50). CHRONOLOGICAL POSITION: This type occurs in sizable percentages in the top half of Pit No. I at Site IS- l l (fig. 10). The whole of this pit belongs in the Chiriqui Phase. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: The first complete description of vessels belonging to this type appears in MacCurdy (1911, pp. 95 and 96). He describes a small group of eight specimens, all tripods, which he named collectively "The Chocolate Incised Group." Two of the three vessels that are illustrated in his report come from the area of Divala, directly west of the city of David, between this city and Puerto Armuelles in the Burica Peninsula. Osgood (1935) disposed of this ware as being trade from Costa Rica where a type of pottery very similar to it occurs in large numbers. The Costa Rican ware takes its name from its Chiriqui counterpart; it is named "Choco? late Ware." Lothrop (1926, vol. II, p. 227) says the following about the Costa Rican ware: Vessels of this class are found most abundandy in the Nicoya peninsula, whence they passed freely in trade to the Lake of Nicaragua, the Highlands of Costa Rica, and even to the distant Province of Chiriqui in Panama. In a recent article, Haberland (1958) argues the reverse from Osgood and insists that the Chocolate Incised Ware is a local product of Chiriqui. I agree with him, but feel that its inspiration is the widespread Chocolate Ware of Costa Rica. Red on White PASTE AND SURFACE: Similar in all respects to Villalba Red Streaked; see that type description for details. FORM: Body wall thickness: Ranges from 6-10 mm. Rim and Lip: Unknown. Base and Supports: Unknown. Major vessel forms: The fragment of a shoulder indicates that the shape may have been that of a carafe, but this is not certain without a more complete specimen. DECORATION : Technique: A white slip is put on the vessel first, and then a bright orange-red paint is applied in either wide bands or rectangular zones. The decorated surface is then polished. Motif: Orange zones in either wide (1.2 cm.) bands or rectangular zones at least 2.5 cm. wide and 4.0 cm. long. These zones are placed on the upper part of the shoulder. The exact placement of the band motif is unknown. CHRONOLOGICAL POSITION: This type occurs in Site IS-7 in substantial percentages in almost all of the levels of Pit No. 2 and in the middle three levels of Pit No. 1. GEOGRAPHICAL DISTRIBUTION AND COMPARATIVE MATERIAL: I have been unable to find the same exact pottery type in any of the Panamanian regions or in Costa Rica. It probably is a local imitation, in a very simplified version, of painted carafe forms that occur in the Cocle-Azuero region. Red Slipped Bisquit Ware This type resembles, in all attributes, the Tarrago Bisquit Ware. The only difference is that it is slipped on the exterior with a heavy red slip that peels off easily. It occurs only in the upper levels of Pit No. I in Site IS-11. All of the occupation of this site has been placed in the last of the Gulf Phases, the Chiriqui Phase. Red Slipped Bisquit Ware represents a localized variant of the Tarrago Bisquit Ware. Unclassified P la in W a r e s The .plain sherds from Pit No. 1 at Site IS?11, whose occupation lies entirely within the Chiriqui Phase, were separated into rough groupings called wares, differentiated from each other on the basis of temper and set apart from the decorated types by their plainness. It is possible that some of these are nothing more than eroded decorated types. PLAIN WARE A.?Identical in paste composition to Linarte Red Line (p. 32). The distribution in the sequence is also similar. PLAIN WARE B.?Probably a variant of Villalba Red Streaked (p. 41) because in all features of paste it is identical. I t is possible that the red slip had merely eroded off some of the sherds. PLAIN WARE E.?BufF-colored paste, compacted but with air holes resulting from small particles of vegetal matter in the clay; buff-colored surface; oc? casionally darker areas from fire clouds. Sand particles too fine to call intentional temper. There is no resemblance to the decorated types. PLAIN WARE G.?Similar in all paste and color features to Plain Ware E and F with only difference in the amount of quartz sand temper particles. Temper abundant , causing a friable texture and a puckered irregular surface. There is no resemblance to the decorated types. BEIGE PLAIN.?In all paste features this ware is like Tarrago Bisquit Ware. Unlike its counterpart, how? ever, the surface has been compacted by the use of a pebble polisher. Nevertheless, it may be simply a variant of Tarrago Bisquit Ware. POLISHED BLACK PLAIN.?Paste is hard, homoge? neous clay with practically no visible temper particles. Sherds are consistently 5 to 10 mm. in thickness with both the surface and cross section color a dark gray to black. 46 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 APPENDAGE MODES Vessel supports (tripod feet, ringstands, and pedestal bases) and handles from all pottery types are described following an outline that permits becoming progres? sively more specific. For example, the category "tr ipod feet" is broken into different modes which, in turn, are further subdivided into modal variations that take into account minute differences between individual specimens. The modes under each category are given letters, and the variations of each mode are designated as numbers attached to the letters. A system of dual numbering has the advantage of allowing the incorporation of new modes in future studies. The number of supports and handles per level is given in Appendix 1, table 5. The concept of modes used in this approach is defined as a small cluster of attributes. Together these attributes may be called a technique (applying clay stripes), a decorative effect (efiigy legs), or an element of shape (outslanted feet). Since these cate? gories of analysis crosscut the three phases, the modes and their variations have been described independ- endy of the pottery types. The distribution of append? age modes in relation to the phase divisions (p. 67), however, breaks down as follows: 1. Modes mainly in the Burica Phase: Tripod Foot a l . 2. Modes mainly in the San Lorenzo Phase: Tripod Feet a2-a6, b l - b 3 , c l - c 2 ; Pedestal Bases b, e, f, g, h, k, 1; Strapped Feet a, c; Ringstand c; Ladle Handle a; Strap Handle b, c; and Double Handle b. 3. Modes that begin in levels placed at the very end of the San Lorenzo Phase, but whose main distribution is in the levels placed within the Chiriqui Phase: Tripod Feet e l - e 4 ; Pedestal Base m ; Round Handle a; and Double Handle a. 4. Modes that occur equally in the San Lorenzo and Chiriqui Phases: Tripod Foot g; Pedestal Base a; Ringstands b, d; Strapped Handle a; and Horizontal Handle a. 5. Modes in the Chiriqui Phase only: Tripod Feet f, h ; Pedestal Bases c, d, i, j ; Strapped Foot b ; Strap Handle d; Round Handle b ; and Double Handle c. 6. Modes occurring in all three phases: Ringstand a. Supports Tripod Feet Mode a FIGURE 30a-/; PLATE 156, c Variations al to a6: Mode a with all of its variations consists of effigy tripod feet decorated to represent zoomorphic or anthropomorphic figures. 1. Modal variation a l shows a modeled anthropo? morphic figure added to the top of a tall (about 24 cm.) tripod foot. T h e legs of the figure are rendered as applique bands 1 cm. in width, ending in unin- cised round pellets 1.5 cm. in diameter. T h e thighs are depicted with oval pellets measuring 1.5 by 2.5 cm. The torso itself is a solid tube 2 cm. in diameter ; the top has been broken off (fig. 30a). Only one specimen of this tripod foot has been found; its provenience is Pit No. 3, level 150-160 cm., Isla Palenque. This level corresponds to the Burica Phase. Mode a l shares with the pottery type Isla Palenque Maroon Slipped the same unmistakable attributes of paste, surface treatment, and red slip. Comparisons with other areas show interesting parallels. The Veraguas Tall Tripods have modeled figurines added to the top (Lothrop, 1950, figs. 40 and 41). These are figurines of animals, however, and only their heads, not their bodies, are shown. They are not very similar to our Mode a l . The exact duplicate of our Mode a l appears in a vessel which Stone places somewhere in the Adantic Watershed and the Highland region of Costa Rica (Stone, 1958, fig. No. 11, g). A tripod similar to our Mode a l has been illus? trated recendy by Stirling in his report on Almirante Bay, Bocos del Toro (Stirling, 1964b, Plate 30b). T h e Almirante Tr ipod has red horizontal lines painted on the exterior, which are missing from our specimen; otherwise they are alike. 2. Modal variations a2-a6 consist of lumps of clay and small strips and pellets applied to the exterior surface of solid, long, tripod legs in order to convey the idea of body parts. There was some lati tude in the way that this decora? tive technique was put into effect. T h e idea of a FIGURE 30.?Tripod feet modes from all three Chiriqui Gulf Phases: Burica, San Lorenzo, and Chiriqui. Horizontal lines indicate areas painted red. Distribution by phase is shown in figure 43 and occurrence by pit level in table 5. a. Mode al (front and side views) b. Mode a2 (front and side views) g, Mode bl (front and side views). /, Mode el (side and front views). q, Mode e4 (front and side views) V, Mode 12 (front and side views). c, Mode a3 (front and side views), d. Mode a4. e. Mode a5. / , Mode a6 (front and side views). h, Mode b2 (front and side views), i, Mode b3 (front and side views), j . Mode cl . k. Mode c2. m, Mode e2. n, Mode fl. o, Mode f2 (side and front views), p. Mode f3 (side and front views). r, Mode g. s, Mode h (side and front views), t. Mode e3 (front and side views), u. Mode i l . WHOLE VOLUME CULTURAL REMAINS 47 m 48 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 face could be simply rendered by the application of slit pellets to form coffee-bean eyes, a strip to form a nose, and slits to render the mouth (fig. 30/). On the other hand, the face could be built up first by shaping a lump of clay on the top of the tripod foot into a circular or semilunar form that stood out from the leg wall. T o this protuberance were added the face features (fig. 30^, c, e; pi. 15^, c). The manner of rendering arms and legs is strict: long strips for the arms and legs; incised, flattened pellets for toes and fingers. There was freedom in the position in which the extremities could be portrayed. The legs could be rendered curled upward (fig. 30^) or downward (fig. 30c). The hands could face downward (fig. 30^), or one a rm could be extended with its hand face back, while the other hand was forward (fig. 30/). These anthropomorphic effigy legs occur at IS -3 in the bottom levels. In Pit No. 3 of S L - 1 , they occur in the top two levels and in Level 50-60 cm.; in Pit No. 1 of the same site, there is one of these effigy legs in the bottom level. They are absent from the other sites. In terms of phases, these modes, except a l , occur at the beginning of the San Lorenzo Phase. An exact duplicate of Mode a2 appears in a com? plete vessel described by MacCurdy (1911, fig. 155). This was reportedly found at Caldera, situated be? tween the cities of David and Boquete, in the Province of Chiriqui. MacCurdy (op. cit. p. 92) notes that this vessel resembles the pieces found by Har tman in Orosi, Costa Rica. Turning to Har t man's report (1901, Plate 65), we can verify some resemblance of Modes a2-a6 to vessels from the Orosi region of Costa Rica. This resemblance is based not only on the use of body-part motifs, but also on the shape of the legs (long and pointed), on their being solid rather than hollow, and on being decorated with red paint. M o d e b FIGURE 30^-/y PLATE 15^-f Variations bl-b3: Mode b consists of a feature of shape: strong outcurving of the tripod feet so that they flare out, probably protruding beyond the vessel walls. This outcurve appears on very short feet (fig., 30?; pi. 15^) or in longer feet (fig. 30^, h; pi. \bd). The foot may be nearly circular (fig. 30A; pi. \bd) or it may be flat on the exterior and tapered toward the back (fig. 30^). Mode b, variations b l - b 3 , is restricted in time to the San Lorenzo Phase, and in place to Isla Palenque. Strongly outcurved solid tripods are not found in the "Classical Chiriqui" pottery types reported by Holmes (1888) and MacCurdy (1911). The only specimen I know of was found in the main land site of Las Tinajas, south of the town of Divala, on the coast halfway between the city of David and the town of Puerto Armuelles. This specimen is in the Smithsonian Institution ( U S N M Archeology Cat . No. 108272); it is exactly like Mode b3 . T h e tripod feet recently found in Tinajas, associated with the Burica Phase urns (de la Guardia , 1966), a re very similar to Mode b3 in representing closely spaced female legs and thighs. T h e Tinajas tripods are, however, tall and straight and decorated with pellets; these features are not present in Mode b3 . One tripod foot that is very similar to Mode b3 , in spite of not being outcurved, also has been reported from the Diquis D d t a (Lothrop, 1963, fig. 22, f). I t belongs in the Early Monochrome Pottery horizon. M o d e c FIGURE 30j-k; PLATE 15a Variations cl and c2: Mode c consists of incisions used as a technique for portraying either body parts (as in the eyes and mouth) or items of clothing, such as headdresses (fig. 3C);). The incision may be long and so shallow as to be almost imperceptible (Mode c l ) or short (1.0 cms. on the average) and deep (Mode c2). The use of incisions is widespread in the pottery of many areas, including that of Veraguas and Chiriqui, but I have been unable to find exact duplicates of Modes cl and c2 in any published literature. I t should be observed, however, that the use of perpendicular incisions on a horizontally applied band around the top part of the tripod foot and coffee-bean eyes ap? pears also in the e3 and e4 variants of the Fish Tr ipod Feet. It is conceivable that the cl and c2 Modes, which belong in the San Lorenzo Phases, are some? how the precedent of the e Modes, which belong to the later Chiriqui Phase. Mode e FIGURE 30/, m, q, t; PLATE 15/ Variations el-e4: A technique of applying incised fillets and incised flattened clay pellets on specific spots on the tripod legs to convey a very schematic impression of the body parts of an animal (fish or armadillo). Pellets pinched laterally and then incised represent the fins of the fish (fig. 30m); a flattened incised area at the tip creates the effect of a fish's tail (fig. 30/); incised pellets at the top of the tripod create a mouth or two eyes (fig. 30i; pi. 15/). Horizontal , shallow incisions combined with punctated bands and perpendicular incised bands create the effect, accord? ing to MacCurdy (1911), of an armadillo (fig. 30^). These kinds of decorations occur on tripod feet that WHOLE VOLUME C U L T U R A L R E M A I N S 49 do not necessarily share the same general shape. For example, Mode el is on a tripod leg that is sharply curved inwardly, while Mode e3 is on a perfecdy straight tripod foot. Modes e l - e 3 are on tall (11-15 cm.) hollow tripods; Mode e4 is on a short (6 cm.) solid tripod. Tr ipod Modes e2 and e3 can be assigned with certainty to the pottery type Villalba Red Streaked because they share added features of paste and surface treatment. The hundreds of complete tripod vessels described in the literature confirm this association (Holmes, 1888; MacCurdy, 1911). Mode e is, like the type Villalba Red Streaked, restricted to the Chiriqui Phase. It occurs in the top levels at IS -3 and IS-7 and in most of the levels of IS -11 . Variant el is aberrant in one respect. Neither the paste nor the surface treatment links it with the type Villalba Red Streaked, to which the other modes in this group belong. The link is with Tarrago Bisquit Ware ; they share all features of paste, surface, and decoration. Exactly the reverse happens with Mode e4. The paste and surface treatment links it with Villalba Red Streaked, but in all of the illustrations and in the whole vessels that I have inspected, this exact mode occurs associated with the pottery type Tarrago Bisquit Ware (MacCurdy, 1911, plate I X ) . The tall tripods of Veraguas (Lothrop, 1950) have legs that are decorated with techniques bearing little resemblance to Modes e l - e4 . The relationships are, rather, to the east. In Costa Rica our modes occur in even greater proliferation than in Chiriqui. A variant of the Tripod Ware has been appropriately designated the "Chiriqui Group (Fish W a r e ) " by Lothrop (1926, Vol. I I , p. 338 and Plate C L X X V ) . O n this ware. Modes e l - e4 occur abundantly as incision on lateral flaps to create fin motifs, incised bands at the top to render the mouth of a fish, coffee-bean eyes, and so forth. Mode f FIGURE 30n-p; PLATE \6b, c Variations fl, f2, f3: This mode consists of an ele? ment of shape often called "mammiform": a bulbous tripod foot with a width at the top of approximately two-thirds of the height of the leg and a tapered point. This shape occurs in otherwise disimilar tripod feet. Mode fl is plain and hollow, having formerly contained clay pellets in the inside; the paste and sur? face treatment of this tripod is that of Tarrago Bisquit Ware. Mode f2 is solid inside, with occasional red splotches of paint. In paste and surface treatment it resembles pottery types that are popular in the San Lorenzo Phase. Mode f3 is solid, but with added incisions at the end to create the effect of toes. There is only one instance of Mode f3 in Level 40-50 cm. of Site I S - 1 1 . To my knowledge it is not common in Veraguas or in Code . In Chiriqui mammiform supports are one of the most common forms accompanying the tripods of the Armadillo-Terra-Cotta-Bisquit Ware, which is our Tarrago Bisquit Ware (MacCurdy, 1911, Plate VI ) . Mammiform feet are not typical of vessels of Veraguas or of Cocl6. The immediate ties are again with Costa Rica, where they occur in a large, number of pottery types like the Black Line Ware (Lothrop, 1926, Plate CLX) or Chocolate Ware Bowls (op. cit. Plate CI I ) , etc. The exact mammiform feet that occur in our Tarrago Bisquit Ware (Mode fl) are found in the Diquis region on vessels that correspond, on the basis of added features of paste and surface treatment, to our Mode fl (Stone, 1958, fig. No. 6 d, e). Our Mode f2 is more similar to modes occurring in the other Costa Rican areas. Modal variations fl-f3 are restricted in time to the Chiriqui Phase, the most recent of the Gulf Phases. Mode g FIGURE 30r Long and hollow tripods with thin walls and probably pellets in the interior. Only one variant has been illustrated, but Mode g also includes slightly different shapes, less wide at the top and completely undecorated. The description of this mode is impaired by the fact that no complete tripods of this kind remain. It is difficult to trace its geographical relationships. Mode g occurs in the Chiriqui Phase. Mode h FIGURE 30J-; PLATE ISa A short (6 cm.) and thin (2.5 cm.) incurved tripod foot with a deep incision down the center. In shape, size, paste surface treatment, and in red overall slip covering the interior and the exterior. Mode h is identical to Mode variant e4. In time and place they are also contemporaneous, being restricted to the Chiriqui Phase and to IS -11 . The idea of dividing the foot into two vertical halves is also inherent in all of the tripods where Mode e occurs, except in the above-mentioned el variant. There is, however, a difference. The Fish Tripods (or e Modes) are hollow inside, and the slit is actually a wide cut into the interior of the leg where the 284-757 0-68?5 50 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 pellets are lodged. In contrast, the h Modes are solid, and the slit is only a deep incision on the exterior. The exact duplicate of Mode h is found in the Diquis Delta in the Red Ware which belongs to the Early Polychrome Horizon (Lothrop, 1963, fig. 43 e). Mode i FIGURE 30U, V; PLATE I6d Variations il and i2: A feature of shape consisting of tripod feet in the form of animal heads (felines). They measure 5.0 cm. in length by 3.4 cm. at the widest point. The feet are hollow and contain pellets inside. A wide slit on each side marks the mouth, while the eyes are rendered as smooth bumps on the sides of the head. The two variants of Mode i (il and 12) are exactly alike in shape, but differ in other attributes. Mode il is slipped red-brown and bears horizontal lines drawn when the surface was very hard prior to firing. Mode il is slipped in white and bears on the front two wide (6-8 mm.) bands painted black. Neither variant of Mode i is indigenous to Panama. None of the Chiriqui tripods or the tripods of Veraguas and C o d e have legs in the shape of feline heads. Mode i is definitely of Costa Rican inspiration, if not of origin, appearing in two pottery types that are in? digenous to that area. The exact relationships follow. The variant il is part of a shallow bowl slipped in a dark reddish-chocolate color, bearing incisions around the exterior of the rim and on the front of the tripod feet (fig. 30w; pi. 16^'). Bowls of this kind belong to the Chocolate Incised pottery type, which occurs with substantial frequencies (up to 5 percent) in the top five levels of IS -11 , an occupation that corresponds to the Chiriqui Phase. The Chocolate Incised type has been directly imported from Costa Rica or is a local product of Costa Rican inspiration. Mode il is duplicated in an illustration fo the Chocolate Ware that occurs in the Diquis Delta (Lothrop, 1963, fig. 20, d) . Mode 12 is also indigenous to Costa Rica where it appears in pottery type Red and Black in the Diquis region. A vessel bearing the exact same tripod foot as 12 has been found in the General Valley of Costa Rica in a grave side by side with an iron knife and an iron ax (Stone, 1958, p. 4 8 ; fig. 7b). According to Stone (1958), this same type of pottery occurs in the upland parts of the Diquis region as well as on the coast and on the Osa Peninsula of Costa Rica. This find in Costa Rica of a vessel bearing the same tripod feet as our Mode 12, together with artifacts of European origin, lends a tentative terminal date to our sequence. Our Mode 12 was found in the middle levd (40-50 cms.) of Pit No. 1 at Site I S - 1 1 . This level, together with all the occupation at Site I S - 1 1 , has been placed in the Chiriqui Phase, the most recent of the Gulf Phases. By implication, the Chir iqui Phase must have a terminal date at the Conquest or shortly following it. Ringstands Mode a FIGURE 31a, b; PLATE 6a, BASE A very short ringstand (height 2.0 cm.) with a diameter of 8.7 cm. Wall thickness is 7 m m . Other associated features are a slighdy bevelled underside and a rich red slip in the interior. The time span of Ringstand Mode a is as follows: it is the only kind of ringstand or support t h a t occurs in the three bottom levels of Pit No . 3 at Site IS-3 (excepting the bottom level, 150-160 cm., which has only 11 sherds in it) . In these three levels there are seven ringstands of Mode a, an impressive number considering that these are the levels with the smallest total sherd sample. These levels correspond to the oldest phase in the Gulf, the Burica Phase. The occurrence of this mode in other levels of this pit (one in Level 60-70 cm. and two in Level 10-20 cm.) is somewhat puzzling. It may be considered mixture, revivalism, or simply persistence. I would argue for mixture. At any rate, the greatest popularity of Mode a is undoubtedly in the lower levels at IS -3 . Mode a also occurs at Site IS -7 . In Pit No . 2 (Pit No. 1 is not discussed) this ringstand has a scattered distribution; four single modes occur in widely sepa? rated levels throughout the pit. As in Site I S - 3 , traces of pottery type Isla Palenque Maroon Slipped appear 0 2 4 CM FIGURE 31.?Ringstand Modes of Burica, San Lorenzo, anc Chiriqui Phases. Hachure indicates areas painted red. a Mode a (interior), b. Mode a (side view), c. Mode c (interior) d, Mode d (exterior). W H O L E V O L U M E C U L T U R A L R E M A I N S 51 from top to bottom of the I S - 7 ; the scattered occur? rence of this mode is, therefore, not surprising. Mode a is absent from Sites SL-1 and IS-11 where the pottery type Isla Palenque does not appear. Recent finds in Chiriqui show that these shallow ringstands occur on bowl-shaped lids for funeral urns (fig. 13). M o d e s b , c, d FIGURE 31C, d; PLATE \lk These ringstands are slightly taller (2.5 to 3.0 cm. high) and have bigger diameters (9.0-9.5 cm.) than Mode a. An additional feature consists of a red slip on the exterior surface. The interior may be plain, sHpped in red, or painted with red bands (pi. 17/, g). Modes b , c, and d occur in Site I S - 3 in scattered levels down to 110-120 cm. They are absent from Sites SL-1 and I S - 1 1 . At Site IS-7 one Mode b and one Mode d occur in level 40-50 cm. of Pit No. 2. Their scattered distribution renders these modes use? less for making chronological distinctions. As for geographical comparisons, these same ring- stand modes occur in the Red Line Ware of MacCurdy (1911, fig. 157), but are very scarce in the more popular types that he discussed. Farther to the east, ringstands such as Modes b , c, and d are not particularly popular in the ceramics of Veraguas. Ringstands are a well-known feature of the Code pottery, but do not closely resemble the Chiriqui ones. Pedestal Bases All these are not of equal distribution and chrono? logical importance. Some are isolated examples, occur? ring only once in our stratigraphic columns. The best time markers of the San Lorenzo Phase are Pedestal Base Modes b, e, f, g, h, k, and 1, and for the Chiriqui Phase, Pedestal Base Modes c, d, i, and j . See p. 67 for distribution by phase. M o d e a FIGURE 32<7; PLATE 17C A flaring-sided pedestal base with red slip on the outside and in the interior; parallel lines and wedges are incised on the exterior. M o d e b FIGURE Z2b A short, sharply everted pedestal foot, punctured on the sides by circular holes and decorated around the top and on the sides with red-painted bands. One of these occurs in Level 70-80 cm. of Site I S - 3 . M o d e c FIGURE 32C It is a tall, fragmentary pedestal base that is too incomplete to merit description. It occurs in the Level 10-20 cm. of Site I S - 3 . I t is decorated along the sides with red bands. M o d e d FIGURE 32d; PLATE 17b A short, plain, thin-walled ringstand, 3.5-4.0 cm. high, with a constricted neck (5.5 cm. wide) and a flaring base (9.0 cm. in diameter). In paste composi? tion and surface treatment this pedestal base belongs in the Tarrago Bisquit Ware . An illustration of a similar base appears in MacCurdy's (1911, fig. 72) report. Mode d occurs in the Gulf islands exclusively in the Chiriqui Phase. One specimen was found in the top level of I S - 3 , Pit No. 3. They are most popular, how? ever, in the middle to top levels of the Pit at I S - 1 1 . Mode e FIGURE 32ey PLATE \7e Shallow pedestal (could be called a ringstand) with a notched base. One occurs in Level 120-130 of Site IS -3 . M o d e f FIGURE 32/ Tall (8.0 cm.) pedestal bases with flaring sides. Their diagnostic feature is a red paint, usually in bands, on the outside. These bands may be arranged in different motifs. Time-place distributions are as follows: one occurs in Level 60-70 cm. of Site IS -3 in the San Lorenzo Phase. They are more popular at Site SL-1 in the same phase; in total four are found in the top levels of both pits. Mode f is absent from either of the two other sites. Exactly the same Mode f occurs in the Red Line Ware of MacCurdy (1911, Plate X X V , fig. f). M o d e g FIGURE 32^; PLATE \7d A tall (10 cm.) plain pedestal base with a rectangu? lar, wide incision on the middle of either face. The interior may be plain or decorated with red bands in geometrical patterns (pi. 17/, g). Paste and surface treatment equate with the pottery type Zapote Red Banded. Two of these modes occur in Level 40-50 cm. of I S - 3 , corresponding to the San Lorenzo Phase. 52 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY V O L U M E 8 0 2 4 6 8 10 CM ^'^^A ^U'^A^^^^'A^^ ^If ^ ? ' ' ? ?^ l^^ ^^" ^?'"^"'? ^"^ Chiriqui Phases. Hachure indicates red-painted areas a Mode a 6 Mode b (side and profile views), c, Mode c. d. Mode d e Mode e /? Mnrle f . \/r^^? Y \ ? J u ? , J ' . ^ ? ? ^ , ^ - , .' (side and interior views). Modes k-m not shown. ,^ Mode e. / , Mode f. ^, Mode g. A, Mode h. ., Mode i. 7, Mode j WHOLE VOLUME CULTURAL REMAINS 53 Mode g is restricted in time to the San Lorenzo Phase. It occurs mosdy in Site SL-1 where it is quite popular in the middle to top levels. Its infrequency at IS-3 can only be explained on the basis of insufficient sample. Mode b FIGURE 32;^ A tall (9 cm.), plain pedestal base, in paste and surface treatment like Mode g, only without the cutout design on either of its phases. The interior may be plain or decorated with red bands (pi. 17/, g). In distribution it is restricted primarily to the San Lorenzo Phase. Mode i FIGURE 32i A short, thin-walled ringstand, with straighter walls than Mode d, decorated with incisions arranged as follows: short (2-3 mm.) shallow oval incisions no both sides of a shallow perpendicular incision. Mode m resembles Mode d in paste composition and surface treatment. It also belongs to the Tarrago Bisquit Ware in the Chiriqui Phase. Mode i is found only at Site IS-11, where it is very popular. Mode j FIGURE 32j A heavy set pedestal with a rounded notch at the bottom of either side. The interior is decorated with very thin lines and dots. One such mode occurs in Level 20-30 cm. of the pit at IS-11. Mode k A variant of Mode e, only without the side punc? tures. One occurs at Site IS-3, Level 60-70 cm. Mode 1 A variant of Mode c, only with bright red bands in the interior and a beige slip around the rim. Mode m A "heavy set" pedestal with thick walls and no decorations. The mode that occurs in Level 50-60 cm. of IS-3 may or may not be the same as the Mode m from Level 80-90 cm. of Site IS-11. Their form is generally similar, but the paste is not. Category c includes the bases of vessels bearing, on three different spots, marks where straps had once been attached; the straps have since been broken off. There is little doubt that category c represents the remains of strapped feet. Category d, however, may include straps that served as handles and not as supports. A strong indication that this may be the case, at least in Sites IS-7 and IS-J1, is afforded by the absence here of pottery type Cangrejal Red Line. Bowls of this type always have strapped feet. Mode d will, therefore, be disregarded for descriptive or chronological purposes. Modes a, b, c FIGURE 33a, b; PLATE 17a Flat, ribbon-like straps (1.0 to 1.4 cm. thick and 5.0 cm. high) are placed underneath the vessel in a tripodal arrangement. This renders the vessel very stable. The interior of the vessel may be plain (fig. 33b) or decorated with red bands (fig. 33a). Mode c, which does not appear in figure 33 consists of vessel bases bearing the marks where straps had once adhered. In the Gulf, Strapped Feet Modes a-c belong in the San Lorenzo Phase and are popular at Sites IS-3 and SL-1. They are absent from Sites IS-7 and IS-11. Strapped Feet The strapped feet have been divided into four categories to insure their correct classification. Only Modes a and b have been found still attached to the underside of vessels. Undoubtedly these served as supports. I I L J 0 2 4 CM FIGURE 33.?Strapped Feet Modes of the San Lorenzo and Chiriqui Phases. Hachure indicates areas painted red. a, Mode a (side view). b, Mode b (interior). 54 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 Strapped feet are practically unknown in the most popular pottery types of the Classic Chiriqui Phase such as Fish-Tripod-Handled Ware (our Villalba Red Streaked) or Armadillo-Terra-Cotta-Bisquit Ware (our Tarrago Bisquit Ware). The only type in Chiriqui that has strapped feet is the Red Line Ware of Holmes (1888) and MacCurdy (1911). Most of the types into which I have subdivided their Red Line Ware are diagnostic of the San Lorenzo Phase. Veraguas exported to our region the idea of strapped feet. Costa Rica could not have done this because the looped legs that occur in the northern sector of the country as well as in western Nicaragua are very different from ours (Lothrop, 1926, Plates CLXVII b, c; CLXXIIId ; f ig . 225c). Exactly the same strap feet as the ones in th^ Chiriqui Gulf occur in Veraguas. Lothrop (1950, p. 38) describes them as follows: Tripod vessels of various sizes and shapes, supported by looped ribbons of clay, are characteristic of Veraguas and of no other region. Strapped feet are not popular in the Cocle-Azuero region, although they do occur sporadically (Lothrop, 1942, Fig. 252a; Plate Il lk). Handles Double Handles Mode a FIGURE 34a-c; PLATE 18j-l Variation al-a5: Large, looped handles that curve upward from two spots on the shoulder of the vessel and join the rim at the level of the lip. The width of the handle across to the two points on the shoulder may be anywhere from 9.5 to 11.0 cm.; the height varies from 6.6 to 8.0 cm. The handles may be round and completely plain (Mode al , fig. 3Aa; pi. 18^) or they may be nearly plain, except for two pellets at the top (Mode a4; fig. 3^d). Incisions also may be added as decoration. The incisions may vary in length from 1-2 cm., but they are always fairly deep. They may cover a large section of the handle and also decorate the flat shelf that joins the rim to the lip (Mode a2; fig. 34^; pi. 18/). Two rounded pellets are commonly placed on the spots where the handle loops join the lip. The pellets may be left plain, or they may be incised (Mode a5; fig. 34^). The lower parts of the handle loops adjoining the vessel shoulder are usually round. In at least one of the variations, however, the handle straps are fairly FIGURE 34.?Double Handle Modes of the Chiriqui Phase. a, Mode al (exterior), b, Mode a2 (interior and side views). c, Mode a3 (exterior), d, Mode a4 (interior and exterior views). e, Mode a5 (interior view). / , Mode b (slanted view). g. Mode c (exterior). flat (Mode a3, fig. 34c). In this same specimen the decoration on each strap consists of two long parallel incisions with a row of punctations in the middle and short (4 mm.) incisions at the top. At Site IS-3, Pit No. 3, Mode a first appears in levels belonging to the later part of the San Lorenzo Phase, but its largest occurrence (12 in total) is in the topmost level belonging to the Chiriqui Phase. This mode is also found in the Chiriqui Phase levels of Site IS-7, Pit No. 2, and in the middle to top levels of Pit No. 1 at Site IS-11. It is absent from Site SL-1 where the occupation belongs solely to the San Lorenzo Phase. Double Handle Mode a has not been previously reported from Chiriqui. Distribution is farther to the east in the district of Remedios. In the surface collections from this area double handle modes are common (see Ranere, Appendix II). We were fortu? nate to find two complete, very large jars with Double Handles Mode a in a private collection in the city of Remedios. These jars measured 53 cm. from the lip to the pointed base. They were 38 cm. in maximum diameter and 21 cm. in diameter at the WHOLE VOLUME C U L T U R A L R E M A I N S 55 mouth. T h e double handles were 21 cm. across at the top. How far east along the coast these handle modes spread is hard to ascertain. Lothrop (1950, fig. 11 a-b) includes the picture of two double handles that are different from the Gulf ones in bearing zoomorphic motifs. They were found in southwestern Veraguas. In a collection from Utiv6, about 25 miles northeast of Panama City, made by M . W. Stirling in 1951, and now stored at the Smithsonian Institution ( U S N M Archeology, Cat. No. 398456^75) , there is a plain double handle exactly like Mode a l . I t is reasonable to suppose, therefore, that Double Handle Mode a represents an influence into the Gulf area from farther to the east. M o d e b FIGURE 34/ A much smaller double handle than Mode a. I t measures 5 cm. across the bottom of the loop and in height is only 4 cm. Otherwise shape is the same. The decoration consists of three rows of short and wide incisions along the entire loop. There is only one example of Mode b in Level 80-90 cm. of Site I S - 3 . Mode c FIGURE 34^ A small double handle of the same general dimen? sions as Mode b , only plain. Nothing is known about its geographical distribution. Chronologically it is con? temporaneous with the other double handles since it occurs in the top levels of Pit No. 3 at IS -3 belonging to the Chiriqui Phase. I t is absent at Site IS-7 , but occurs at Site I S - 1 1 . No double handle modes are found in Site S L - 1 . Strap Handles This category includes variations on one theme: handles that are flat in cross section and extend from one side or from opposite sides of the vessel wall to merge with the rim. All of these straps bear some form of decoration, either painted or in the form of applique designs. Mode a FIGURE 35a-g; PLATE 18a-c Variations al-a7: All strap handles in this group are fairly flat in cross section (thickness 1.2-1.5 cm.). They are broader at the top, where they meet the lip, than at the bot tom where they join the vessel wall. T h e arch where the straps bend down to meet the vessel wall is usually sharp; this is the spot that is usually decorated. Regardless of any other decoration, the edges of all these straps are always painted in red. Mode a l (fig. 35fl) has no decoration except red bands on the sides and a wider (2.5 cm.) red band at the top. The other variations have some form of applique incised decoration in addition to red bands. Mode a2 (fig. 35^) has a winding incised fillet, and in shape is much longer and flatter than the other strap handles. In Modes a3 to a6 the applique fillets are not winding, but are placed fairly straight; they are also incised (fig. 35c-/; pi. 18^-/). Round incised pellets are often placed on the same area as the applique fillets. Punctations (Mode a3) or incisions (Mode a6) may be added to decorate this spot. In at least one of the modes (Mode a7, fig. 35^; pi. 18(2) the decoration is zoomorphic, consisting of a frog measuring 5 cm. across the arms with a body 2.2 cm. long and 1.6 cm. wide. The wrinkled skin of the frog has been clearly rendered by closely spaced punctations made with a hollow reed. Strap Handle Variations al?a6 are found in all four of the Gulf sites. At IS -3 strap handles occur in both the San Lorenzo and Chiriqui Phases, but are absent from the Burica Phase. At Sites S L - 1 , IS-7 , and IS-11 they occur in all the levels. The Strap Handle Mode a with all its variations is characteristic of several pottery types of the Gulf: Arayo Polished Line, Horconcitos Red Banded, Banco Red Line, and Linarte Zoned Red Line. Since no whole vessels belonging to these types were found in our excavations, it is difficult to match the variations with the specific pottery types to which they belong. Reference should be made to illustrations in the reports of Holmes (1888, fig. 160) and MacCurdy (1911, Plate X X V ) . Undecorated strap handles occur in the pottery of Veraguas (Lothrop, 1950, figs. 58 and 59). Lothrop (op. cit., p. 44) is of the opinion that they probably are copies of Chiriqui prototypes. Strap handles are very rare in the pottery of Code . M o d e b A narrow (2.5 cm.) strap handle with a small applique decoration in the exterior came from Level 80-90 cm. at Site I S - 3 . As an isolated mode it is of limited chronological or geographical importance. Mode c FIGURE 35/i A narrow (2 cm.) strap handle with red-painted bands along the sides; the rest is broken off. It belongs to pottery type Centeno R e d Banded. Two of these 56 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 FIGURE 35.?Strap Handle Modes of the San Lorenzo and the Chiriqui Phases. Parallel lines indicate area painted red. a, Mode al (front view and profiles of either end), b, Mode a2 (top and side views), c. Mode a3 (top and side views), d. Mode a4 (profile and top view), e, Mode a5 (profile of top view). / , Mode a5 (profile of top end, of side, and front view), g, Mode a7 (top view), h, Mode b (top view), i, Mode d (top view). modes occur in Levels 100-110 and 80-90 cm. of Pit No. 3 at Site IS-3. Mode d FIGURE 35e A very flat strap, varying in width from 0.8 to 1.0 cm., painted all over with a red slip. The paste composition, surface treatment, firing, and slipping make this mode a part of the pottery type Villalba Red Streaked and thus belongs in the Chiriqui Phase. It is found in IS-7, which has the largest percentages of this pottery type. Round Handles These are handles that are nearly circular in cross section. They are arranged in a tight loop and placed on the body wall usually horizontal to the rim. Mode a FIGURE 36a A plain, round handle with a thickness of 1.5 to 2.0 cm. It bears no decoration. How it was placed in relation to the rim is difficult to determine since no sherds with both handle and rim have been found in any of our sites. Mode a, together with all of the other round handles, is a mark of the Chiriqui Phase. At Site IS-3 it occurs in the top seven levels. At Sites IS-7 and IS-11 it occurs throughout the deposit. It is absent from SL-1. Mode b FIGURE 3^b-e; PLATE 18/, h Variations bl to b4: This mode encompasses several round handles which have in common a red slip; most of them also bear incised pellets for decoration. WHOLE VOLUME CULTURAL REMAINS 57 I I I I I 0 2 4 CM FIGURE 36.?Round Handle Modes of the Chiriqui Phase, a. Mode a. 6, Mode bl (side and end views), c. Mode b2 (side, top, and underside), d. Mode b3 (side and end views), e, Mode b4 (end and side views). Mode b l (fig. 366) is flat at the edges, but round on top. I t is sharply curved and has two incised pellets on either side of the top section. Mode b2 (fig. 36c; pi. 18/) is similar to Mode b l , only it is round in cross section at all points. Three incised pellets adorn the top part. Mode b3 (fig. 36d) is the flattest of all these modes; it has been put in this group because it shares all the other attributes with the rest of the b modes?red paint and incised pellets. Few incised pellets decorate the exterior at the top and on the sides. Mode b4 is plain. I t measures 2.2 cm. in width and is 4.0 cm. at the top. I t was placed vertically from the vessel wall to the rim. Mode b4 (fig. 34^; pi. 18A) is a vertical round handle joining shoulder to vessel wall. Mode b with all of its variations is restricted to the latter par t of the Chiriqui Phase. I t is absent from both IS -3 and S L - 1 , but occurs in substantial numbers at Sites IS-7 and I S - 1 1 . T h e paste, surface treatment, and decoration leave little doubt that Mode b belongs to the pottery type Villalba Red Streaked (or Fish-Tripod-Handled Ware, as it is called in the literature). The exact Modes b l , b2, b3 , and b4 are portrayed in vessels of the Handled Group of Painted Ware (MacCurdy, 1911, fig. 122, Plate X X ) . Round handles are absent from Veraguas or C o d e pottery. The relationships are to the east with Costa Rica. In fact, there is a pottery type here which bears the same name as the Handled Group of Painted Ware and has very similar handles to our Mode b (Lothrop, 1926, Vol. I I , fig. 240, Plate C L X X X I I ) . Whether this Panamanian ware influenced the Costa Rican variant, or the reverse, is hard to ascertain from distributional data alone. Ladle Handles Ladle handles is my term for broad and flat projec? tions attached to one side of a plate. These plates are known from other regions as incense burners; this may have been their original use. Mode al (fig. 37a; pi. \^d) measures 8 cm. at the top and approximately 9.5 cm. at the point where it joins the plate. It is decorated with two raised parallel ridges 1.5 cm. apart on each side of the front surface; six coffee-bean pellets are placed between these raised ridges. The edges, as well as the area in the center, are 58 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 FIGURE 37.?Ladle Handle Modes of the San Lorenzo Phase a. Mode al (profile and front views), b, Mode a2 (front view) 4 CM slipped in red. It was found in Level 90-100 cm. of Pit No. 3 at I S - 3 . Mode a2 (fig. 37b; pi. 18^) differs slightly from Mode a l in being unpainted, in having the ends more pointed, and in having the decoration in a V-arrange? ment rather than at both sides of the handle. One example came from Level 20-30 cm. of S L - 1 , Pit No. 2. Both ladle handles belong to the San Lorenzo Phase. Incense burners are primarily a Middle American rather than a South American trait. They have been found from Michoacan in Mexico all the way down to Colombia and Ecuador (Lothrop, 1942, pp. 174? 176). No incense burners have yet been reported from the Veraguas area, but they occur abundantly in Sitio Conte. These may have round handles or flat rectangular handles very similar to ours (Lothrop, 1942, figs. 354, 355, and 356). The incense burners that occur in Costa Rica have large handles modified into zoomorphic forms (Lothrop, 1926, Vol. I I , figs. 255 and 256). They do not look at all like Gulf of Chiriqui, Panama, ladle handles. Horizontal Handles These are fairly flat, undecorated looped handles that are placed one on each side of a globular neckless vessel, either parallel to the vessel lip or at a slight angle to it (fig. 38a-c). In length they average 8 cm. and in thickness anywhere from 0.75 to 1.75 cm. They are uncommon, being restricted solely to Site IS-3 in levels of the San Lorenzo and Chiriqui Phases. FIGURE 38.?Horizontal Handle Modes shared by the San Lor? enzo and Chiriqui Phases (exterior and side views), a, Mode a l . b, Mode a2. c, Mode a3. ' ' ' I I 0 2 4 CM FIGURE 39.?Miscellaneous Handle Modes from the San Lorenzo and Chiriqui Phases, a, Mode al (profile and front views). b, Mode a2 (profile and front views), c. Mode b (front and side views), d, Mode g. e, f, Mode j (two side views). WHOLE VOLUME CULTURAL REMAINS 59 Miscellaneous Handles Miscellaneous Applique Motifs Some handles not classified into the mode groupings occurred in scattered levels of the San Lorenzo and Chiriqui Phases and are illustrated in figure 39a-/. Each handle will be described separately. Handle a is small and flat, decorated by a n applique design, probably a reptile motif. I t is restricted in time to the San Lorenzo Phase. Handle b is a var iant of a, showing an incised design rather than an applique design. It also occurs in the San Lorenzo Phase. Handle c is a small knob placed near the rim. One occurs in Level 10-20 cm. of Pit No. 3 at Site S L - 1 . Handle d is a very large, rounded handle, similar to some that occur in Veraguas. It comes from Vill? alba Site (IS-7) in a level of the Chiriqui Phase. Handle e is a lobe representing either an extension of the lip or a raised shelf under it. Two specimens occur in a level of Site IS-11 belonging to the Chiri? qui Phase. These are applied ribbons of clay on coffee-bean motifs, incised or punctated. They occur in several types from San Lorenzo and Chiriqui Phases. Modfs a, b, and g (fig. 40 a, b, g) occur at IS -3 and SL-1 in levels belonging to San Lorenzo Phase. In paste composition they are identical to the types of these phases. Motifs a and b also have traces of red paint in the exterior edges of the applique motif. Motif d (fig. 'iOd) is different; it is applied to a black, polished ware not found in the Gulf. It is probably a trade item. Motifs c, e, f, h, and i (fig. 40 c, e, / , h-i) are found at IS-7 and IS-11 in levels belonging to the Chiriqui Phase. The paste composition of these motifs is in? disputably Tarrago Bisquit Ware. The coffee-bean motif and the incised wavy bands are common in the Classical Chiriqui pottery illustrated by Mac? Curdy (1911). J L J 4 CM FIGURE 40.?Applique motifs of the San Lorenzo and Chiriqui Phases. 60 S M I T H S O N I A N C O N T R I B U T I O N S TO ANTHROPOLOGY VOLUME 8 STONE ARTIFACTS Twelve kinds of stone implements have been found in the Chiriqui Gulf sites. Their distribution by levels is summarized in Appendix 1, table 6 and by phases on p . 69. They correlate with the three phases as follows: Burica Phase: No stone artifacts of any kind. San Lorenzo Phase: At Site IS-3 in levels of this phase, stone artifacts occur only toward the end of the phase and not at the beginning. The same thing happens generally at Site SL-1 of the same phase. Chiriqui Phase: The stone inventory is almost the same as it is for the latter part of the San Lorenzo Phase. The only artifact types exclusively associated are sandstone saws and scrapers. Celts Celts are rare in levels belonging to the San Lorenzo Phase, but occur in respectable numbers in levels belonging to the Chiriqui Phase, especially at Site IS-11 which yielded nine of them. At Site IS-3 they occur in the upper half of Pit No. 3. Only one was found at SL-1 in Pit No. 1, Level 10-20 cm. At IS-7 three came from the top levels of both pits; one was found in Levd 90-100 cm. of Pit No. 2. There are some differences in the shape. The follow? ing groupings are based on these differences. GROUP 1 Material: Light to dark bluish-gray porphyritic rock with phenocrests of feldspar and an aphanitic groundmass. Description: Fairly large (9-10 cm. long) and broad (5.7 cm.), biconvex in cross section, with either or both ends battered from reuse as a hammerstone. The sides may be fairly straight or slightly convex (pi. 19/). The butt end is blunt. Initially they had been polished all over. These celts are typical of Site I S - 3 . GROUP 2 Material: Several kinds?a dark gray aphanitic basalt, a phyllite, high in silica, and a light bluish-gray aphanitic cryptocrystalline rock. Description: Smaller than Group 1 celts (average length, 8 cm.; average width, 4 cm.). Two are roughly diamond-shape in outline (pi. 19^, / ) . In the larger ones the ends have been used for either pecking or hammering. Rough marks of percussion chipping are evident along the longer sides and occasionally large concoidal chips have been removed from the ends (pi. \9h, j). Cross sections vary from biconvex to diamond-shape. Most had been polished originally. These cdts are typical of Site I S - 1 1 . T h e larger ones (pi. I9k, j) resemble closely the Group 1 celts from Site I S - 3 . GROUP 3 Material: A gray, aphanit ic microcrystalline equi? angular rock; one specimen has been identified as basalt. Description: Long (average 13 cm.) and fairly nar? row (average 6 cm.) with straight edges tapered toward the butt and a cross section that is convex on one side and a truncated pyramid in shape on the other side. Tiny to large chipping scars, which in some cases have been smoothed over, are evident along the longer sides (pi. 19^). The prominent feature of this group is the long flat facet on one of the polished faces (pi. \9b,e). This type of faceted celts is characteristic only of Site IS -7 . In the same site was found a very well-finished fragment (the bit end only) of a highly polished celt with a sharp bevelled cutting edge (pi. 19*;). T h e face may have been faceted, but this is hard to tell from its fragmen? tary condition. GROUP 4 Material: Basalt porphyry. Description: Polished on both surfaces with a bevelled broad bit and a sharp, pointed but t (pi. \9d). Large chipping scars are evident along the longer sides. Only one occurs in Site SL?1. GROUP 5 Material: A dark, gray or blue-black aphanitic basalt. Description: Small, biconvex, highly polished celts. Two were found at Site I S - 1 1 , one is Level 80-90 cm., and one in Level 40-50 cm. Very small elongated celt with a small bulge on the lower par t nearest the bit end (pi. 190- Biconvex in cross section, with a rounded but t and a straight, well-polished tapered cutt ing edge. Chipped all over by percussion. Size: length 7 cm. and width 2 cm. Small, tubular celt, biconvex (almost round) in cross section with sides that taper gradually toward the ends (pi. 19^). Cutt ing edge is very small and sharp. I t has been beautifully polished all over. Size: 9 X 2 cm. Flakes Only five flakes occur, all of them from Site SL-1 , Pit No. 1. Four were found in Level 10-20 cm. and one in Level 70-80 cm. They will be described together. WHOLE VOLUME CULTURAL REMAINS 61 Material: Four are flakes from quartz fragments: crystalline quartz, milky quartz, and jasper. There is only one basalt flake. Description: Flakes with either use chipping along an edge or a corner, or unmodified except for the blow which detached it from the core. T h e flake shown in plate 20/ has a long area of use retouch chips on one corner. T h e flakes can be as long as 3.0 cm. or as short as 1.5 cm. Manos In the San Lorenzo and Chiriqui Phases; none asso? ciated with the Burica Phase. At Site I S - 3 manos were found in the top two levels and in the fifth level. At Site S L - 1 , Pit No. 1, there are two from Level 20-30 cm. and one from the bottom level. No manos were found at Site IS -7 . At Site IS-11 there was one mano in Level 10-20 cm. and another in Level 90-100 cm. Two varieties occur with the same chronological distribution. GROUP 1 Material: Several kinds of rock?a brown, aphan i t i c microcrystalline rock composed chiefly of feldspar, a light green granite rock rich in quartz and in feldspar, and an equiangular gabbro rock composed equally of a white intermediate plagioclase feldspar and a green mineral. T h e majority are granite or gabbro. Description: In this group belong the manos that were shaped before being used for grinding (pi. 200- The shape or size of many specimens is hard to reconstruct because only the end portions remain. Cross sections are planoconvex or ovoid; only one is triangular in cross section. Both faces, or in a few cases only one face, were used for grinding; they are smooth to the touch. T h e grinding face in some is slightly convex front to back. Length may reach 24 cm. (pi. 20t). Thickness varies between 4?7 cm. GROUP 2 In this group are elongated, irregular-shaped cobbles, unmodified except for the faces which have been smoothed in grinding. T h e grinding surface is flat so that the cross sections are planoconvex. In at least one specimen the end has been lightly used for pecking. T h e only complete one found (pi. 20M) measures 22 x 7 x 8 cm. several manos were found here. The reverse is true of Site IS-7 where no manos were found, but one metate fragment was uncovered in Level 50-60 cm. of Pit No. 2. Las Secas produced manos but no metates. Of the four specimens, three are shaped legs (pi. 20r) aijd the fourth is a large fragment of a trough-shaped metate from Site IS -3 (pi. 20.y). They will be grouped together for description. Material: Different kinds of rocks?a dark brown phaneritic, granitic rock composed mainly of acidic to intermediate plagioclase feldspar, a light, gray, vesicular, aphanitic rock with macrocrystalline crys? tals, a brown porphyritic rock with large phenocrests of intermediate plagioclase feldspar, and a phaneritic groundmass. Description: Three of the metate legs were reused after they had broken off, two as pecking stones, and a third for abrading (possibly to enlarge a hole). T h e shape of the legs varies: one is roughly triangular with rounded edges, a second one is a truncated cone in outline and nearly circular in cross section, and a third is planoconvex in cross section. The largest measures 12 cm. in height. The fourth (pi. 20s) consists of a boulder unmodified except for the deep (8 cm.) depression at the center. Only portions of the sides in the depression have been smoothed by usage. One of the short ends is closed off for the obvious reason of preventing the grain from spilling out. Net or Line Weights These occur only at Site IS?3 in scattered levels between 10 and 100 cm. These levels represent the San Lorenzo and early Chiriqui Phases. Material: Most are made of a green aphanitic rock composed of rounded to angular fragments or grains of crystalline and amorphous material. Description: Small and flat ovoidal cobbles gen? erally unmodified except for a slight pecked notch on the same place in both of the longer sides. The notches have not been smoothed but are, instead, left rough as if to hold a cord from a line or a net. Some are as large as 6 by 10 cm. (pi. \9m, n); others are fairly small (pi. 19o-^). Metates Four occur in the San Lorenzo and the Chiriqui Phases, but like all other stone artifacts, they are absent from the Burica Phase. At Site IS -3 there was one metate in Level 10-20 cm. and two in Level 90-100 cm. There are no metates at S L - 1 , although Pebble Polishers These occur almost exclusively in the San Lorenzo Phase at Sites SL-1 and IS-3 and are most popular at Site S L - 1 . No pebble polishers were found at Site IS-7 . Only one doubtful specimen came from the top level of the pit at Site IS -11 . 62 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 This distribution fits in very well with the fact that the pottery types with polished exteriors are typical of the San Lorenzo Phase and are most popular at Site S L - 1 . Sites IS-7 and I S - 1 1 , which lack pebble polishers almost entirely, have large percentages of pottery types Villalba Red Streaked and Tarrago Bisquit Ware, both of which have unpolished surfaces. Material: Porphyritic rock with an aphanitic groundmass of intermediate composition; one pebble has been identified as granite. Description: Small, irregularly shaped, smooth, and shiny pebbles with parallel striation marks across one or both faces (pi. 20^-?). Slight polishing facets can be observed in some. Average diameter is 3 cm. Pecking Stones These occur in the top half of the pit at Site I S - 3 , in Level 20-30 cm. of Pit No. 2 at S L - 1 , in Level 50-60 cm. of Pit No. 2 at IS -7 , and in scattered levels down to 70-80 cm. in the pit at I S - 1 1 . They span both the San Lorenzo and the Chiriqui Phases. All will be described together. Material: Very fine-grained aphanitic, cryptocrys? talline rocks composed of feldspar, magnati te, and hornblende. Some are metamorphic ; others are igneous. Description: Unmodified cobble, usually ovoid in shape and biconvex in cross section, with pecking marks at the ends and occasionally on the long sides (pi. 20/). They were probably used to shape tools. Length 6-14 cm. Pounding-Anvil Stones Represented in both the San Lorenzo and the Chiriqui Phases at all sites except Villalba ( IS-7) . Material: Several kinds of rocks?a light gray aphanitic microcrystalline rock, a greenish, phaneritic, medium to fine-grained rock, and a rusty-brown aphantic groundmass with a few large phenocrests of light feldspar. Description: Natural cobbles, round to oval, usually planoconvex in cross section so that one of the faces is nearly flat; some have both faces flat. O n the center of the flat face (or faces) there is a pecked depression possibly used to hold small nuts while they were being cracked (pi. 19a:). The ends and the long sides show extensive evidence of bat ter ing; this indicates these stones were used also for pounding. Projectile Points Only five were found in our excavations, all from Pits No. 1 and No. 3 at Site SL-1 in the San Lorenzo Phase. They divide into two groups on the basis of material. GROUP 1 Material: A dark gray to black aphanit ic rock of high density, probably a basalt. Description: Each point is described separately. 1. Essentially straight sided (one side is slightly convex) with small angular tapered shoulders and a contracting stem (pi. 20a). Bulbar face flat and es- sendally unworked in the body portion, though several accidental (?) flakes have been removed. T h e bulbar face of the stem was chipped to reduce the size of the bulb of percussion. Truncated triangle in cross section with previous flake scars evident in upper face. Steep sided with edges modified by irregular, steep secondary percussion chipping. T ip broken off. Size: 7.0 X 1.9 X 1.2 cm. 2. Straight sided with angular tapering shoulders and straight contracting stem not quite coming to a point (pi. 20b). Bulbar face flat and unmodified; bulb of percussion not present. Opposite face exhibits two long flake scars resulting in a ridge down the center of the artifact. Edges and base are shaped by steep-sided percussion flakes removed from the non- bulbar face. Thin in cross section. Size: 5.5 x 2.0 x 0.6 cm. 3. Appears to be a fragment of a steep-sided projectile point with a flat t runcated pyramid cross section and an unworked lower (bulbar?) face (pi. 20^). Size: 2.1 x 1 .4x0 .6 cm. GROUP 2 Material: Jasper , a variety of quartz. Description: Each point is described separately: 1. Convex, irregular sides, rounded, tapering shoulders (more pronounced on one side), and a straight-sided contracting stem (pi. 20d). Bulbar face generally flat and unworked. Upper face shaped along steep sides by irregular percussion. Small secondary chipping along edges. Size: 3.3 x 1.4 x 0.8 cm. 2. Equilateral triangle in cross section (pi. 20c). Shaped by percussion flaking, almost entirely primary. Tapers gradually to a point with the but t broad and blunt. Edges at but t end have been dulled for 1.5 cm. from the base. Size: 5.5 x 1.1 cm. Sandstone Saws These occur exclusively in the Chiriqui Phase at Site IS-11 where they come from almost all levels. They were probably used to cut ornaments and fishhooks out of shell. Material: A brown siltstone or fine-grained sand? stone composed of an igneous mineral . WHOLE VOLUME CULTURAL REMAINS 63 0 1 2 CM FIGURE 41.?Shell pendants from Las Secas Site (IS-11), Chiriqui Phase. Description: Th in in cross section (4-5 mm.) and tubular in shape, with flat and smooth surfaces. The most characteristic feature is the bevelling, through polishing, of one or both sides so as to form five abrading edges that may be used for cutting soft materials like shell. In specimens with two cutting edges these may be parallel to each other (pi. 20m, n) or at angles to each other (pi. 20p, q). This is also true of the specimens in which only one of the edges has been bevelled for cutting (pi. 20/). Width ranges from 2-4 cm.; thickness from 4?6 mm. ; length is unknown. Scrapers A few scrapers occur in sites of the Chiriqui Phase. They are of little use in chronology or distri? bution and have thus been put into a single group. Material: A cryptocrystalline quartz that in some specimens looks like agate. Description: Angular with what appears to be chipping marks along one edge; otherwise unmodified. Specimens are sharp enough to have been used for scraping hides and so forth. Sharpening Stone Only one found in IS-3 , Pit No. 3, Level 70-80 cm. Material: A light, yellow-green aphanitic, vesicular rock, which is composed mainly of feldspar with very minor amounts of magnatite. Description: Texture makes it excellent for polishing celts and sharpening the edges of stone objects. Both faces used. One face has four concave depressions, resulting from polishing objects like celts. Length is broken off at 16 cm.; thickness is 2.5 cm. BONES AND SHELL ORNAMENTS Three small ornaments, probably pendants, were found only in the Chiriqui Phase at Site IS -11 . Two are shell and one bone. Each of these pendants will be described separately. A small figurine (fig. 41a) has the shape of a lobster, if looked at from the narrower end; if turned around, one may conceivably see a female figure instead. Length is 2.2 cm. and width 1.6 cm. at the widest point. The carving has been done by careful incising on very soft shell, probably oyster. The back part of the pendant is left unworked. The specimen was found in Level 20-30 cm. at Site IS -11 , Pit No. 1. A small pendant (fig. 41^) is in the shape of a flat (5 mm. thick) celt, measuring 2.3 cm. in length and 1.6 cm. in width. The simulated cutting edge (called a bit in a stone celt) has been bevelled. The top, or butt end, is slightly thicker than the bit end as in most stone celts. This specimen was found in Level 30-40 cm. of Pit No. 1 at Site IS-11 in the Chiriqui Phase. A small bone pendant is beautifully carved in the form of a frog (fig. 41c). The details are amazing: tiny toes, legs curled upward, and bulging eyes that are split in the middle. Length measures exactly 2 cm., and width is 1.3 cm. across the legs. The back is flat and shows no details. A perforation through the body leaves little doubt that it was used as a pendant. The specimen was found in Level 30?40 cm. of I S - l l , together with the preceding specimen. Stone (1963) describes shell ornaments, most of which represent a frog, from the Diquis region of Costa Rica. They are very similar to this specimen. The frogs in the Diquis have been found with vessels of the Red-and-Black Ware (or Polychrome Ware) which in Chiriqui is called Alligator Ware. The same association holds for the Gulf. The Sequence, Cultural Inventory, and Way of Life of the Gulf of Chiriqui Inhabitants The various layers of cultural materials in the refuse deposits of the various sites were interdigitated into a single chronological sequence (fig. 42). The framework was provided by the stratigraphy of Site IS-3 with its deep refuse deposit in Pit 3 (fig. 4). Certain pottery types were restricted to the earliest, the intermediate, and the most recent of the occupation levels. The percentage frequencies, at each site, of the pottery types that form the basis for the sequence at the Gulf of Chiriqui are presented in figures 4, 6, 8, and 10. Presence or absence of Appendage Modes (supports and handles, fig. 43), and, to some extent, the time distribution of stone artifacts (fig. 44) were also useful criteria. These, as well as the presence or absence of pottery types, are considered the "internal" criteria of chronology. External criteria, on the other hand, are provided by trade wares; in our case, sherds that are important for chronology come from the Cocld-Parita area. John Ladd, who has made an exhaustive classification of the pottery types in that area (Ladd, 1964), personally identified each of the trade sherds. The division of the phases on the basis of the "internal" criteria was carried out independently of the information based on trade wares. In fact, it was not until the very end of the analysis that trade sherds were plotted by levels and matched with the already established phases. The result was very satisfactory; Ladd's phases correlated well with my San Lorenzo and Chiriqui Phases. Each will be summarized, beginning with the oldest and moving to the most recent. 284-757 0-68?6 65 66 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY THE SEQUENCE VOLUME 8 Burica Phase It was named after several sites on Punta Burica (Appendix 2, p. 107) that produced a high fre? quency of the pottery type Isla Palenque Maroon Slipped, the diagnostic pottery of the Burica Phase. While this report was in preparation, the same pottery was isolated in other mainland cemeteries. Since the Burica Phase name has already been used in a pre? liminary publication (Linares de Sapir, 1966, pp. 405-414), it has therefore been retained here. The Burica Phase is represented by the two bottom levels of Site IS-3 (Isla Palenque) and the bottom level of Pit No. 2 at Site IS-7 (Isla Villalba), with the diagnostic pottery type reaching a frequency as high as 40 percent in some levels. It is possible that Plain Ware C is actually part of Isla Palenque Maroon Slipped (p. 23). If the badly eroded surfaces of this type actually caused misclassification and the group of Plain Ware C were added, the frequency would have been raised to at least 79 percent in the two bottom levels of the sequence (fig. 4). Appendage Modes typical of the Burica Phase are Tripod Feet Mode al and the very shallow Ringstand Mode a. No other support or handles of any kind are associated with the pottery of the Burica Phase in the Chiriqui Gulf sites. Ranere (Appendix 2, p. 115) has nevertheless found additional modes associated with Isla Palenque Maroon Slipped in the districts of Punta Burica, San Felix, and Remedios. Stone arti? facts were absent from these areas. One of the eleven sherds in the bottom levd of Pit No. 3 at Site IS-3 has been tentatively identified as a Venado Beach type. It is a small rim sherd (pi. 14) that looks exactly like a complete vessel labeled Venado Beach in the Peabody Museum, Harvard University. Levels directly above this one produced trade pottery identified by Ladd as being Code, but their eroded condition makes it impossible to tell whether they are Early or Late Code. Associations with the Venado Beach sherd tend to favor an early date for the Burica Phase. San Lorenzo Phase The intermediate phase in the Gulf sequence is called San Lorenzo Phase, after the district in which the Site SL-1 (El Cangrejal) is located and where pottery types characteristic of the San Lorenzo Phase were found in highest frequency (fig. 6). The San Lorenzo Phase thus includes all of the levels of Site SL-1, plus the middle ten levels between 30 and 130 centimeters at Site IS-3 (Isla Palenque); it is poorly represented at Site IS-7 (Isla Villalba) and entirely absent from Site IS-11 (Las Secas). Pottery types restricted to the San Lorenzo Phase are Banco Red Line and Castrellon Red Slipped. Types that show their highest frequency in the San Lorenzo Phase, even though they are found in other phases also, are Arayo Polished Line, Caco Red Slipped, Cangrejal Red Line, Centeno Red Banded, Horconcitos Red Banded, Linarte Zoned Red Line. The Unclassified Plain Wares H, I, and J at Site IS-3 are most popular in the San Lorenzo Phase also. To date, no cemeteries with only San Lorenzo pot? tery types have been found, and some of the pottery types with their highest popularity in the San Lorenzo Phase, namely Arayo Polished Line, Banco Red Line, and Linarte Zoned Red Line are found in Classical Chiriqui graves (MacCurdy 1911; the Red Line Group). This would not appear to negate the value of establishing the intermediate San Lorenzo Phase. Future work on the mainland should prove if the San Lorenzo Phase has importance there from a chrono? logical standpoint, but on the islands of the Gulf of Chiriqui the pottery types with their highest frequency in the San Lorenzo Phase are definitely older and in lower strata beneath the Chiriqui Phase types. Chiriqui Phase This is the last and most recent occupation in the Chiriqui Gulf (fig. 42). It is characterized in the top levels by the appearance of three important pottery types: Tarrago Bisquit Ware, Villalba Red Streaked, and Cavada Applique and Red Banded. The first two occur in large percentages (20 to 30 percent) at Sites IS-7 and IS-11. Since these are two of the most popular wares in the "Classical Chiriqui Culture," corresponding, respectively, to Armadillo-Terra Cotta Ware and Fish-Tripod-Handled Ware in the nomen? clature of Osgood (1935 pp. 236-7), the phase was simply named Chiriqui. Some of the pottery types that characterize the San Lorenzo Phase continue into the Chiriqui Phase, but in very small percentages (fig. 42). The Chiriqui Phase is further characterized by the occurrence of the following Appendage Modes, some of which may have actually appeared in the very late part of the San Lorenzo Phase, achieving maximum representation in the Chiriqui Phase (fig. 43): Tripod Feet Modes e 1 to e4, f and h; Pedestal Base Modes c-d, i, m; Strapped Feet Mode b ; Ringstand Mode d; Strap Handle Mode d; Round Handles Modes a and b ; and Double Handle Mode c. PIT 1 PIT 2 1 > ? PIT 1 PIT 2 PIT 1 PIT 2 PIT 3 PIT 1 P IT 3 UJ r a A"' (T < Z> X QQ 0 . IS-3, PIT 3, 130-140 IS-7, PIT 2, 120-130 IS-3, PIT 3, 140-150 150-160 I 3i I I ? i o s I s ? - FIGURE 43.?Distribution of Appendage Modes in the San Lorenzo and Chiriqui Phases. Only presence or absence indicated. (IS-1 should read IS-11.) 68 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 THE CULTURAL INVENTORY AND WAY OF LIFE T h e limited extent of the excavations in the Gulf Islands, plus the perishability of materials, preclude detailed reconstruction of the social and material culture of the Gulf people throughout the years. I t is possible, however, to suggest a schematic outline of the main subsistence patterns in each of the phases, based on stone artifact distributions (fig. 44) and on the occurrence of mollusks (fig. 45) and bones (Appendix 1, table 7). Burica Phase The Burica Phase is surprisingly devoid of any cultural material except pottery. There are no stone artifacts, no remains of mollusks, fish, or mammals. This situation may be partly explained by the factor of preservation. Where shells are abundant in a midden, they contribute as much calcium as the soil can absorb so that other materials such as bone can be preserved. From a glance at figure 45 and Appendix 1, table 7, it is evident that the sites with the greater number of bone remains are also the ones with the greatest amount of shell concentration. There are no shells and consequentiy few bones at Sites IS-3 and IS-7 . What about the lack of stone artifacts? If we observe figure 44, we note that the absence of stone artifacts is a phenomenon of the lower levels of Sites I S - 3 and IS-7 , not only those that belong to the Burica Phase. Although the true explanation for this situation cannot be recognized from existing evidence, we can offer some suggestions. The people of the Burica Phase may have thrown their food refuse in areas other than those we excavated, or they may have eaten mollusks and fish at the beach and thrown their remains into the sea. If this is assumed to have been the case, it is more difficult to explain the absence of stone artifacts except to state simply that these people did not practice agriculture (though most people at the time did farm). Whatever nonagricultural tools were missing, one can blame on chance. A second possibility is that the people of the Burica Phase lived elsewhere, only visited the islands inter? mittently, and did not farm on them. At the islands they lived on marine fauna, traces of which have disappeared. A third explanation may simply be one of insufficient sample. The intensity of the occupation in the Burica Phase, judged crudely from the fact that the number of sherds in levels of a two by one meter strata cut of the Burica Phase never reached 200, was scanty as compared to subsequent phases. Nevertheless, the absence of stone artifacts at Site I S - 3 continues in all the levels up to 100 cm. (Appendix 1, table 6). In these levels sherds number in the five hundreds and yet stone ardfacts are missing. For this reason I prefer the explanation that the Burica Phase people, as well as the people of the early pa r t of the San Lorenzo Phase at I S - 3 , did not live permanent ly on the island. No remains of house structures were found in the Burica Phase or in any of the phases. This may be the fault of our small test excavations and the lack of preservation. We found no burials either. T h e people of the Burica Phase probably came to the islands from the mainland nor th and southwest of David, somewhere between Punta Burica and the Chiriqui Gulf Islands. This is the area where all of the Burica Phase sites have been found to date . None have been found east of David or of the Chiriqui Gulf. T h e ultimate origin of the Burica people may thus lie to the west in Costa Rica. As soon as they arrived on the islands these people apparently started trading with the Parita Bay area far to the east of the Gulf; presumably they also traded with people even farther away, as there is one sherd of Venado Beach in the lowermost level of Pit No. 3 at Site I S - 3 . None of the forms, decorative techniques, or paste composition of Isla Palenque Maroon Slipped are carried over into the types of the San Lorenzo and Chiriqui Phases. Apparendy, the Burica Phase people of Site IS-3 eventually abandoned the island, perhaps by the pressure of the people in the following phase. San Lorenzo Phase Cultural material other than pottery occurs at Isla Palenque (IS-3) in levels belonging to the San Lorenzo Phase, as well as in the bot tom levels of Site El Cangrejal (SL-1) , whose entire occupation belongs within the San Lorenzo Phase. If we recog? nize that the paucity of material makes all inferences highly tentative, we can hazard some interpretations. During the San Lorenzo Phase the inhabitants of Isla Palenque fished and hunted in the earlier part, but turned mostly to agriculture in the latter part . The kinds of fish that they caught have not been identified, but they are probably the same WHOLE VOLUME CHIRIQUI INHABITANTS 69 species existing in the area today and known locally by the following names: bonito, sierra, pez gallo, hurel, and mero. Lines and/or nets were weighed down by "sinkers" made of notched pebbles. T h e only mammal hunted was Odocoileus virginianus (Chiriqui White-tailed Deer), a large deer with branching anders that lives in the open savanna between the mountains and the Pacific side, all the way from Costa Rica to the Bayano River in eastern Panama. I t favors the thickets with secondary growth, rather than the deep forest (Gk)ldman, 1920). The same type of deer was also hunted by the inhabitants of El Cangrejal (SL-1) . I t is interesting to note that at both sites the only remains found are parts of the thigh and leg (Appendix 1, table 7). This may mean that hunters brought back to the village only the easily transportable parts of the animal which con? tain most of the good meat. O n a small island such as Palenque there is litde hope of finding deer a few months after the island has been settled by man. The people of IS -3 probably went hunting on Isla Brava where, even today, people from Boca Chica go to hunt . They may have hunted also in the vicinity of Site S L - 1 . Projectile points used in hunting are found for some reason only at Site SL-1 and not at I S - 3 . Only one type of shell, Melongena patula, which occurs in brackish waters, was found in Level 30-40 cm. of Site I S - 3 . At this time in their history the inhabitants did not rely solely on mollusks for food. This may be due to the fact that Isla Palenque, though small, offers good agricultural land. Further? more, the site is near the big island of Isla Brava where the inhabitants could have gone to cultivate their fields much in the same manner that the present inhabitants on the mainland of Boca Chica go to Isla Brava to farm. In contrast, the inhabitants of SL-1 used mollusks as a substantial par t of their diet. In figure 45 only the presence or absence of mollusks has been indicated; from field notes, one knows that they were abundant , especially in the top six levels of Pit No. 1. Incidentally, the inadequacy and arbitrariness of the sample from such small pits may be indicated from the fact that Pit No. 3 has such a small number of shells. Most of the species of mollusks, whether bivalves or univalves, live in mud flats and mangrove swamps. The only bivalves found at SL-1 are Anadara grandis, Anomalo- cardia subrugosa, and Lunarca bucaruana. Univalves are more numerous. Among them are Melongena patula and Strambus sp., two very large shells with plenty of meat. Tiny species such as Cerithidea sp., found in brackish water bays or entirely out of the water cling? ing to reeds and twigs, were gathered also. The species in < I a . g I t C H I P H A S E N ZO K o SA N < UJ K < IS-I1,PIT 1, 0 - 1 0 10- 20 2 0 - 30 3 0 - 40 4 0 - 50 5 0 - SO 6 0 - 70 70- 80 8 0 - 90 9 0 - 100 IS-7, PIT 2, u - 10 10- 20 2 0 - 30 3 0 - 40 4 0 - 50 9 0 - 60 6 0 - 70 7 0 - 60 lS-7, PIT 1, 0 - 1 0 10- 20 I S - 7 , P I T 2 , 8 0 - 9 0 9 0 - 100 IS-7, PIT 1, 2 0 - 30 3 0 - 40 4 0 - 50 IS-7,PIT 2, 100- no 110- 120 IS-3,PIT 3, 0 - 10 1 0 - 2 0 2 0 - 30 3 0 - 40 4 0 - 50 5 0 - 60 13-7, PIT 1 , 5 0 - 60 1S-3,P1T 3, 6 0 - 70 7 0 - 80 8 0 - 90 9 0 - 100 100 - 110 110- 120 120- 130 S L - l , P l T i a 2 , 0 - 10 10- 20 2 0 - 30 3 0 - 40 4 0 - 50 5 0 - 60 6 0 - 70 7 0 - 80 8 0 - 90 IS-3, PIT 3, 130- 140 IS-7,PIT 2, 120- 130 IS-3,PIT 3, 140- 150 150- 160 1 ' 1 1 1 1 ' 1 i 1 ' i i 1 1 1 1 1 z J 2 ^ S T O N E A R T I F A C T S ? ? ? 1 1 I I I I 11 1 1 1 11 U IP 1 1 i | i | 1 1 ' 1 ? i | i | " k k i> 11 ? I I I I 11 11 (1 I I I I i> 1 1 11 11 11 ? i i I I i * z (/) z 2 O. z z . ?1 ^ It O Z I ? ? 11 1 ? ? o < X t FIGURE 44.?Occurrence of stone artifacts in the San Lorenzo and Chiriqui Phases. Only presence or absence indicated. Thais kiosquiformis and Littorina sp. are found clinging to rocks between the mid-tide and high-tide line. The people living at El Cangrejal (Site SL-1) had rich mud flats and swamps to exploit. When the tide retreats, great expanses of mud are exposed below the bank on the edge of Site El Cangrejal (SL-1). The 70 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y VOLUME 8 G A S T R O P O D S Cerithidea sp. Conus sp. Fasciolaria princeps Fissurella virescens Latirus ceratus Littorina varia Littorina sp. Melongena patula Muricanthus radix Natica chemnitzii Merita scabricosta Oliva sp. Oliva porphyria Petaloconchus innumerabilis Purpura patula pansa Siphonaria gigas Strombus sp. Strombus galeatus Strombus granulatus Tegula pellis-serpentis Thais crassa Thais kiosquiformis Thais melones IS -3 Pit 1, Levels 1 2 3 4 5 6 7 IS-7 Pit 1, Levels 1 2 3 4 5 6 Pit 2, Levels 1 2 3 4 5 6 7 8 9 - ? X - X - PELECYPODS Anadara (Cunearca) Anadara grandis Anadara multicostata Anadara obesa Anadara tuberculosa Anomalocardia subrugosa Area pacijica Cardita laticostata Chama frondosa Dosinia dunkeri Lunarca bucaruana Ostrea sp. Ostrea fisher i Ostrea pinctada Pilar tortuosus Polymesoda maritima Protothaca asperina Spondylus princeps X X X FIGURE 45.?Distribution of mollusk remains by site. Only presence or absence indicated. WHOLE VOLUME CHIRIQUI INHABITANTS 71 IS-11 Pit 1, Levels 1 2 3 4 5 6 7 8 9 10 SL-1 Pit 1, Levels Pit 3, Levels 1 2 3 4 5 6 7 8 9 1 2 3 4 5 6 7 8 9 G A S T R O P O D S Cerithidea sp. Conus sp. Fasciotaria prirweps Fissurella virescens Latirus ceratus Littorina varia Littorina sp. Melongena patula Muricanthus radix Natica chemnitzii Nerita scabricosta Oliva sp. Oliva porphyria Petaloconchus innumerabilis Purpura patula pansa Siphonaria gigas Strombus sp. Strombus galeatus Strombus granulatus Tegula pellis-serpentis Thais crassa Thais kiosquiformis Thais melones PELECYPODS Anadara {Cunearca) Anadara grandis Anadara multicostata Anadara obesa Anadara tuberculosa Anomalocardia subrugosa Area pacifica Cardita laticostata Chama frondosa Dosinia dunkeri Lunarca bucaruana Ostrea sp. Ostrea fsheri Ostrea pinctada Pitar tortuosus Polymesoda maritima Protothaca asperina Spondylus princeps - X X X ? X ? X ? ? X - - - X ? X X FIGURE 45?Continued 72 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 people could gather mollusks here or go to the man? grove swamps nearby. Besides hunting, fishing, and collecting mollusks, the Indians of the San Lorenzo Phase practiced agri? culture. What species of plants they grew remains unknown. Most likely corn was one of them since manos and metates were found at the site. The Indians may have also gathered nuts, probably of the pejibaye (Guilema gasipaes) palm, which they cracked with the help of pounding-anvil stones. To clear the land, they used polished celts that were probably hafted unto strong handles. The people of both IS-3 and SL-1 made large quantities of pottery, which they polished with the aid of small, round pebbles (called by us pebble polishers); they chose smooth pebbles so as to compact the surface and leave few striation marks. Traces of any structures were absent; probably they were made of straw. We know nothing about the composition of their families and of the structure of their society. We know that they traded, directly by sea, or in? directly by land, with the Late Cocl6 people of the Parita Bay. Chiriqui Phase The same patterns of agriculture combined with hunting and fishing, which existed in the San Lorenzo Phase, continued into the Chiriqui Phase in the Gulf. Hunting and gathering mollusks were activities of secondary importance to the inhabitants of Isla Palenque; they lived mostly from farming, supple? mented with some fishing. Few bone remains of fish have been found; that they were caught is mainly an inference based on the presence of notched, flat cob? blestones that were supposedly used as weights. For farming they used celts; these may have served the dual purpose of felling trees and breaking the ground. Once the celts were useless in agricultural pursuits, they were employed as hammerstones, perhaps to shape other tools. The inhabitants of Site IS-7 occasionally supple? mented their diet with mollusks gathered in mud flats or mangrove swamps. In contrast, the inhabitants of Site IS-11, who lived very far out to sea, relied heavily on mollusks. The refuse piles they left behind are real shellmiddens. While an average of two to three pounds (1-1.5 kilos) of shell was found in each of the levels of Site IS-7, an average level of Site IS-11 yielded 30 to 40 lbs. (15-20 kilos) and on occa? sion up to 75 lbs. (35 kilos) of shell, not counting the enormous quantity of decomposed shell that was irrecoverable. The people living in these sites gathered different species of mollusks. Site IS-7 is located on a protected bay not far from the brackish waters of the mangrove swamps bordering the estuaries. Las Secas Islands, on the other hand, are located far out in the open sea. The species of shells found at each naturally reflect this environmental difference. Most of the mollusks that occur in Site IS-7, whether bivalves or univalves, are species that live in mud flats or mangrove swamps. Among the Pelocypods (bivalves) in these habitats are Anadara grandis, Anomalocardia subrugosa, Dosinia dunkeri, Pitar tortosa, and Polymesoda maritima. Other kinds of bivalves such as Lunarca bucaruana live intertidally on the reefs. Most of the gastropods (univalves) that we found at SL-1 are available at low tide on mud flats; among these is the large mollusk, Melongena patula, and other smaller species such as Muricanthus radix and Natica chemnitzii. The small species Littorina varia clings to rocks between mid-tide and high tide in the same way as Siphonaria gigas. The very large shell Strombus, with abundant food, is found just below the tide line. Only two of the eleven species of bivalves at Site IS-7 occur at Site IS-11: Anadara grandis and Ostrea sp. The rest of the species at IS-11 are found in completely marine, sandy, or rocky habitats. Several of the smaller species such as Area pacifica and Chama frondosa adhere to the underside of rocks in intertidal zones. Of the eight species of gastropods (univalves) at IS-7, only three occur at Las Secas. Most of the others that are typical of Las Secas are small species that live between rocks intertidally, such as Merita scabri? costa, Tegula pellis-serpentis, and Fissurella virescens. At Las Secas are found also the species Fasciolaria princeps, a very large shell (up to 20 cm. in length) found offshore, and the beautiful and colorful species Oliva porphyria, which lives in sandy beaches buried just be? low the surface. Siphonaria gigas is an interesting smaller species, which adheres to rocks and breathes air. One of the important shells not used for food is the Purpura patula pansa, found commonly on rocks in exposed locations. This mollusk has its mucous glands modified for the secretion of a purple fluid. This fluid was used by the Indians of western Central America to dye textiles (Keen, 1958, p. 376). The islands of Las Secas are surrounded by rocky cliffs and coral beds where, when the tide is low, many mollusks are found. The people who lived here gathered their mollusks around these rocks rather than wading in muddy waters or in mangrove swamps. In contrast to the setders of Isla Villalba who must have fished littie, those of Las Secas relied heavily on fish. Although the species that they caught have WHOLE VOLUME CHIRIQUI INHABITANTS 73 not been identified, there are numerous bones of large and smaller fish and even rays. These settlers also hunted; the inhabitants of Villalba apparently did not. The Indians caught iguanas and big lizards (Appen? dix 1, table 7) much in the fashion of the Guaymi Indians of today, who eat these reptiles. In addi? tion, the people of IS-11 hunted deer of two species. The smaller of the two species, Mazama americana, is commonly known by the name of "brocket" in English and "cabra del monte" in Spanish. I t prefers wooded habitats and lives in thickets in remote areas. I t is easily identified by its small size and unspiked antlers. Today it is found in the area of Boquete in Chiriqui and in the province of Bocas del Toro (Goldman, 1920). The larger species of deer hunted by the IS-11 inhabitants is the well-known species Odocoileus virginianus (Chiriqui White-tailed deer) . I t occurs in the open savanna regions in the Pacific coastal lands of the Is thmus; it is a widespread species occurring all the way to the United States. T h e only bones found in the refuse belonging to this species are fragments of the thigh and leg of the deer. It is possible that the people went away to hun t and only brought back the largest and most edible portions of the animal. In spite of the heavy reliance on hunting, fishing, and mollusks, the inhabitants of Las Secas also cultivated. O n e mano was found in the excavations (Appendix 1, table 6), plus many celts. Another type of artifact that is very intriguing is the sandstone saw, consisting of a flat abrader with beveled edges; it was probably used to shape hooks, made of shell, for fishing. Manos and metates used for grinding com, plus celts, were found in the upper levels belonging to the Chiriqui Phase at Site I S - 3 . The celts were used in felling trees or digging the ground. The people living on the island at the beginning of the Chiriqui Phase traded items with the Late Code dwellers of the Parita Bay region. At the end of the Chiriqui Phase, the trading was with the people of the Herrera Phase in the Azuero Peninsula (Parita Bay), who came later than the people of the Late Cocl6 occupation. During the Chiriqui Phase, Las Secas and Villalba may have been used as ceremonial gathering places, if one attributes some religious purpose to the carved stone basalt columns found on both islands. Those at Site IS-6 on Villalba were originally arranged in a circle on top of a walled mound with niches (Haber? land, 1960a). The most immediate source for the pottery types of the Chiriqui Phase that appear in the Gulf Islands is the mainland lowland area north and south of David. Thousands of graves of the "Chiriqui Culture" have been excavated here. They yielded vessels belonging to the four well-known Chiriqui wares, but in what percentages each ware was represented, and where, is not known. Only the two types mentioned appear in the Gulf in large numbers; the two others. Alligator Ware and Negative Ware, are only represented in very small quantity in the last part of the occupation (fig. 10). Perhaps this indicates that there are chrono? logical differences within the Chiriqui Culture; per? haps it is a matter of geographic distribution, Alligator and Negative Wares being more typical of the High? lands. These questions can only be settled by an archeological study of the Chiriqui Highlands. European Conquest and Assimilation of the Indians of Chiriqui Since the most recent aboriginal occupation in the archeological sequence, the Chiriqui Phase, lasts until the European Conquest, it is of interest to review here the chroniclers of the 16th and 17th cenmries for information on the Indians of Chiriqui. This task is beset with difficulties. Places mentioned in the chroni? cles are often impossible to identify, and references to the inhabitants are no more than inventories of gold contributed and souls converted. Many questions remain unanswered. For example, to what kind of cultural unit did the Spaniards refer by the term "provincia"? Was it a territory, one of several tribes, a linguistic group, a residence group, or a kinship unit? Interpretations to these and many other ques? tions may eventually be found in the yet unpublished manuscripts of this epoch. With the records now available, we can only hope to piece together a brief and sketchy historical outline in proper chronological order. EARLY EXPLORATIONS OF COASTAL CHIRIQUI When Pedrarias became governor of Castilla del Oro in 1514, he initiated armed explorations called entradas into the unknown western regions of the Isthmus. The first was by land. In 1515 Captain Gonzales de Badajoz, with a contingent of soldiers, marched from Darien as far west as the Azuero Peninsula, where he was defeated by the cacique Paris. The following year an overland expedition, under the Licenciado Caspar de Espinosa, was more successful in overruning the Gulf of Parita and the Azuero Peninsula. From here Espinosa dispatched a sea expedition under Bartolome Hurtado and Hernan Ponce to explore the coast west of the Azuero Penin? sula. Lothrop (1963, p. 6) says of this voyage: Their fleet consisted of three small dugout canoes from Chiru and two large ones with 18 thwarts (36 paddlers) from Guarari. Supposedly they ventured as far as the Gulf of Nicoya in north? western Costa Rica. With watercraft of this sort. Captains Hurtado and Ponce must have been forced to touch land at several points to get food and supplies. Apparently, they stopped at the island of Cebaco, where they were met by a very friendly chief, and also visited Isla Coiba, where they procured some gold from the in? habitants (Anderson, 1914, pp. 198-199). These two islands are off the south coast of Veraguas. They may have stopped also in the Gulf of Chiriqui. We know that they tried to land on the Golfo Dulce in Costa Rica, but were persuaded against it by a group of Indians called Chiuchires, who were armed to the teeth. The Chiuchires lived at or near the Diquis Delta (Lothrop, 1963, p. 6). The next entrada led by Espinosa in 1519 did not proceed beyond the Azuero Peninsula. His third venture the following year took him all the way by sea to the province of Chiriqui. At least four different accounts of this expedition exist: one is by Bartolom^ de Las Casas, another is by Andagoya, a third one is by Herrera, who plagiarized from the other two, and 75 76 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 a fourth is by the historian Fernandez de Oviedo y Vald^s. Only Herrera's account has been ignored in the following composite summary of events. The Licenciado Espinosa left the town of Panama by sea, commanding the two ships that had previously been built by Balboa to explore the coasts of the Pacific. He landed on Isla de Cebaco, whose inhabit? ants had been visited four years earlier by Bartolom^ Hurtado and Herndn Ponce. The Indians offered no resistance. Instead they cooperated by telling Espinosa of the great riches that the cacique Urraca had in store in his mountain habitat to the north. An attempt to steal Urraca's gold resulted in a resounding defeat for Espinosa's forces. His and his men's lives were saved only by the intervention of land forces under Pizarro, but the Spaniards were no match for this Indian chief. They were forced to flee in a great hurry to their ships. Espinosa and Pizarro then sailed together westward across the Gulf of Chiriqui to the province of Burica. Here they stopped at an island which they called Sancto Matias (probably Isla Burica off the peninsula of the same name). From this island they went to the mainland where many Indians came out to fight, but were soon defeated. Villages were raided and women and children were captured. Learning that a wealthy chief lived nearby, Espinosa sent Francisco Campanon with 50 men to raid him. After a four days' journey fr-om the Burica Coast into the mountains, they came upon the inhabitants of the Indian town: where they had a palisade made out of wood like a fortress, and the Spaniards entered it and killed many because they . . (the Indians) . . could not leave the inside of the palisade, stumbling one over the other, and the Spaniards had a good chance to use their force, their arms and their swords. (Acosta, p. 77). Espinosa then ordered one of his ships to sail on to the Gulf of Liicar in Nicaragua, while he returned by land to Panama. The other ship presumably followed him by sea because we next find Espinosa in the province of Huista in Chiriqui loading a ship with maize and sending it on to Panama. This ship may have been anchored near the Bay of Muertos where Isla Villalba (IS-7) is located. Of the province of Huista, Andagoya has the following words to say (Markham, 1865, pp. 24 and 25): The people of this province and of that of Burica were almost exacdy the same in the fashion of their clothes, and in their customs. The women wore a truss round their loins, as their clothing; and the men were naked. The country is fertile, with plentiful supplies of fish, and a great quantity of swine, which were caught with large nets of stuff like hemp, called by the Indians nequen, the meshes being a finger in breadth. These nets were fastened at the entrance of a wood where there was a herd of swine, who came against the nets and were unable to get through the meshes. Then the people called out, the nets fell over the swine, and they were killed with lances, so that none escaped, of those who fell into the nets. Leaving this province on our way to Panama by land, we arrived at a mountainous district, with a cold climate, where we found some forests of very beautiful oaks covered with acorns. There were three or four chiefs in this province, and their villages were well fortified with palisades made of very strong thorny plants, intertwined and forming a thick wall. Throughout these districts the Indians were seized and bound. From Burica to this province, which is called Tobreytrota, nearly every chief has a different language from the others. From this hilly country we turned to descend towards the sea, and came to the province of Nata, where the town was founded which is now called Nata. The next Spaniard to venture into the lands of what is today the province of Chiriqui was Gil Gonzales Divila (or de Avila), who sailed westward from the Pearl Islands in 1522 hoping to reach Nicaragua. Davila wrote a letter explaining his journey; the treasurer of his party, a man named Andres de Cerezada, left a detailed itinerary of the Indians met and of the gold gathered. The section of Cere- zeda's account which deals with the islands and main? land regions of western Panama is worth quoting in full (Peralta, 1883, pp. 27 and 28; my translation): He left the island of Las Perlas Tuesday January 21 of 1522, arriving at the island of Ceguaco which is more than 50 leguas away, baptizing the chief and 184 people, including those baptized on the way back; he (the chief) gave 1844 pesos and 7 tomines of gold. To this island the chief Guanant, who lives in the mainland, sent 86 pesos, 4 tomines of gold. The island of Madera is 15 leguas by sea from Ceguaco: there came the chiefs of the region, who are Tucug, Pera, Huysca, el Coao, Brocatebagia, Tacuria; 37 persons turned Christian, they gave 1095 pesos, 4 tomines of gold. The island of Cebo is 12 leguas by sea from the island of Madera; 6 persons were baptized, the chief gave 39 pesos, 4 tomines of gold. Chiriqui is 5 leguas of the island of Cebo on the mainland, and from here on the Captain proceeded by land. A chief came from the highlands and 8 people took communion; the cacique from the sierra gave 56 pesos of gold. The chief Copesiri is 6 leguas ahead; 44 persons were baptized, he gave 55 gold pesos, and the chiefs of Calaocasala, who came there, gave 174 pesos, and the chiefs of Barecla 84 pesos, and chief Cherique 26 pesos, which makes a total of 339 gold f>esos. The chief Charirabra is 3 leguas ahead: 64 persons were baptized, he gave 55 pesos, and other chiefs 35 pesos, making a total of 90 pesos. The chief Burica is 10 leguas ahead: there were baptized 47 persons, he gave 249 pesos and six tomines of gold, and Andr6s Niiio brought here 120 pesos given by a chief of the island of Quica, and 64 pesos given by the chief of the Island of Madera which are 433 pesos, 6 tomines of gold. Only one of the islands that Cerezeda mentions, Ceguaco, can be positively identified as the island of Cebaco in the Gulf of Montijo. His account is of much value because it proves that at least some of the Gulf of Chiriqui Islands and many points on the mainland coast of this province were populated at the time of the conquest by numerous Indian chiefdoms. WHOLE VOLUME E U R O P E A N C O N Q U E S T 77 T h e year after Ddvila's expedition, Pedrarias ordered Benito Hur tado to buUd a town in the province of Chiriqui. The name chosen for the town was Fonseca, and the place was near the present towns of San Lorenzo and Horconcitos (fig. 2), a few kilometers east of our Site El Cangrejal (SL-1). I t was built in the year 1523. I have translated the account of Hurtado 's experiences in this town as recorded by Las Casas (1951, vol. 3, pp. 399-400). When he (Hurtado) arrived he sent for the peoples of that land; his call was answered by the jieople of Chiriqui, and afterwards by a people called Bareclas, and then by those of the province of Burica and the people who lived on the Gulf of Osa (Golfo Dulce), all of it a heavily settied land that extends for about one hundred leagues. All of those people came without offering resistance because they were startled by the wars and the cruelties that they had heard of, or had suffered when, in the past years, Espinosa had traversed those lands. The Spaniards were two years in that town (Fonseca) with the Indians serving them, but they (the Indians), not being able to endure so much servitude and continuous tyranny, revolted and killed some of the Spaniards, who finally agreed to abandon their town. All that land, as well as thousands of leagues of land nearby, was at one time full of people but is today deserted and inhabited by wild beasts, mainly tigers. The discovery of Peru in the 1530s brought to a halt the first period of discovery and conquest in the Isthmus. For several decades no more "entradas" were made into the western provinces of Panama; instead, all able-bodied soldiers went to seek riches and adventure in Peru. From the above accounts of the first half of the 16th century we can make some general observations relevant to the archeological situation. None of the four islands visited in 1522 by Ddvila and recorded by Cerezada can be positively identified as one of our Chiriqui Gulf Island sites. Nevertheless, some of the conditions existing in them may have also applied to our islands. Each of the islands he mentioned was governed by one chief only. T o judge from the number of baptisms, 184 in one isle and 6 in another, the size of the population may have varied considerably. Not only the islands, but also points along the mainland coast of the Chiriqui Gulf were heavily inhabited. The Burica Peninsula was dotted with villages along the coast and in the interior. This fact differs with few finds of surface collections along the Burica coast belonging to our most recent Chiriqui Phase (see Ranere, Appendix 2, p . 113). The Indians of the Province of Huiste, in the vicinity of the Bay of Muertos where our site Isla Villalba (IS-7) is located, were said by Andagoya to be, "in the fashion of their clothes and in their customs," almost exactly like those of Burica. He does not mention whether or not they spoke the same language, but there apparendy was some sort of cultural continuity between the two areas. This fact may be significant since IS-7 , like Punta Burica, has an occupation belonging to our most recent occupation, the Chiriqui Phase. The area around our site of San Lorenzo (SL-1) may, however, have been uninhabited at this time since, in order to populate the town of Fonseca, Indians were sent for from as far away as the Golfo Dulce in Costa Rica. It is interesting to note that Site SL-1 lacks the most recent Chiriqui Phase. Concerning the seaworthiness of prehispanic Indian crafts, it is revealing to note that the Spaniards them? selves were able, in borrowed canoes, to cover enor? mous distances. Bartolome Hurtado and Hernan Ponce went from the Peninsula de Azuero as far west as the Gulf of Nicoya in Costa Rica. Francisco de Campafion, in a small canoe, navigated from the Gulf of Chiriqui to the town of Nata in the province of Code. It is not surprising, therefore, that the pre? hispanic Indians of our islands probably made voyages as far east as the Azuero Peninsula and as far west as Punta Burica?and probably much farther. 78 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 17TH-CENTURY EXPLORATIONS OF COASTAL CHIRIQUI A hundred years after the first Spanish contacts with groups on the mainland and coast of Chiriqui, the Audiencia de Panama commissioned a pilot called Diego Ruiz de Campos to carry out a detailed survey of ports in the Pacific. The Relacion which he wrote in 1631 is an important document since it contains detailed description of two of our Chiriqui Gulf sites; the first of these is Las Secas (Cuervo, 1891, p . 49) : Turning west four leagues to the sea (from the Rio San F61ix) there are the Secas islands which are nine in number and which though called Islas Secas (dry islands) all have fresh water and some have coconut palms. They are hilly and forested and in some parts they have some small beaches of sand where boats and ships can arrive. In all of them there is nothing important except water and lumber. T h e second one of the islands that Diego Ruiz de Campos talks about is Villalba, more commonly known as Isla Muertos (Cuervo, 1891, p . 50): On the eastern part of the river of Chiriqui there joins another river called the river of Chorcha which is 'caudaloso' and has much wood to construct ships and other 'baxeles' and on them have been built very big and good ships. At the mouth where the river Chorcha joins the Chiriqui river there is a small island called isla de Muertos which is heavily forested. T h e archeological analysis made in this study would favor the hypothesis that the occupation of Villalba (Isla Muertos) and Las Secas lasted until the European Conquest. I t would be logical, therefore, to expect that Diego Ruiz de Campos would have found these islands inhabited by Indians. W e forget, however, that more than 100 years had passed since the early Spanish explorers cruised around them in their efforts to explore lands farther to the west. T h e Indians of these islands could have been exterminated or taken into slavery; those that survived probably fled to the mainland. Living in such circumscribed surround? ings, they were an easy prey to conquerors. For this reason it is not surprising that the pilot de Campos found these islands uninhabi ted. In fact, he includes, in his report a description of the only 30 natives who survived after the conquest on the island of Taboga near Panama City (Cuervo, 1891, pp . 20 and 21). They represented a population which, a century before, had probably consisted of several hundred . MAINLAND INDIANS OF CHIRIQUI Indians living on the mainland managed to survive the first Spanish raids more successfully than did their neighbors on the islands. The task of colonizing these groups was left in the hands of Franciscans and Do? minicans, whose policy was to resettle them into towns in order to facilitate their conversion and con? trol. With this end in mind, the Spaniards established a number of towns on the savannas of southern Chiri? qui that served both as centers for encomenderos and as places from which to spread the faith. The first of these was Remedios, founded in 1589. It was followed two years later by the towns of Pedro de Montijo and Santiago de Alanje. In 1623 the Indian town of San Lorenzo was built on the shores of the Rio Fonseca, not too far from the spot where 100 years earlier stood the town of Fonseca. From Remedios and San Lorenzo the energetic Fathers proceeded to advance on the Guaymi Indians who lived in the highlands in the eastern part of Chiriqui and adjacent Veraguas. The best known of these missionaries was Fray Adrian de Santo Tomas. He lived with the Guaymi between 1622 and 1637 and left us a very complete account of their customs, which has been translated in a publication on the pottery of Veraguas (Lothrop, 1950, Appendix 11). T h e coastal savannas and the highlands nor th of the Gulf Islands were populated at the t ime of the Conquest by Indian groups who set themselves apar t from the Guaymi, who were their enemies. They were catechized by missionaries based in the town of Alanje. I shall discuss later how the distribution of these Indian groups conforms closely to the distribu? tion of the 'Classical Chiriqui ' ceramics, m a d e until the time of Conquest and used to define the most re? cent occupation in the Gulf. Here it suffices to point out that a knowledge of these Indians is useful in drawing analogies to the Chiriqui Phase occupation. Fortunately, there is a fairly detailed account of a visit to them made by a missionary called Fray Antonio de la Rocha sometime in the first three or four decades of the 1600s. I t was first published in 1681-82, but it was not until several years ago tha t a reprinted version brought it to the eyes of a larger public (de la Rocha, 1964, pp . 82-132). Unti l then WHOLE VOLUME EUROPEAN CONQUEST 79 it had been assumed that the Guaymi were the only inhabitants of Chiriqui in the 16th and 17th centuries, since they are the only ones that remain. In the follow? ing pages I have presented a summary of points in Fray Antonio's account, which are useful to us here. T h e two most important groups described by Fray Antonio de Rocha are the Doraces and the Zuries with whom he had longest contact. I have worked out their approximate homeland as being the area of what is today Caldera, Boquete, and the skirts of the Volcan Baru (fig. 2). Their homeland was high, cold, and thickly forested with abundant streams and rocky outcrops. They had large cultivated fields, on the savanas bordering the mountain slopes facing either ocean, where they grew maize. Other edible products were plantains (Spanish? introduction), cacao, and pejibaye {Guilielma gasipaes). They ground and processed maize into a mazamorra (thick gruel), which they diluted to make fermented chicha. There are passing references to stored maize from the year befof e, but this apparently referred to a small number of cobs stored by hanging in woven bags inside the houses and probably used for seed. There is no mention of manioc, either sweet or bitter. Other than maize, the most important food source was the pejibaye, a type of palm which yields clusters of a hard fruit with a rich pulp and a small nut. The Indians ate them cooked; they cracked and mashed the nuts, adding the pulp to their 'chicha' : 'Si la llevan en sus viajes, es cocida, o cruda para asar y con este sustento solo andan lucios y gordos como si comieran capones' (de la Rocha, 1964, p . 91). Unfortunately, data on agricultural practices is very scanty; there is no information on the ownership, location, size, and distance of their maize fields. All we know is that the men did the planting and that temporary huts were built near their maize fields in the agricultural season. Other subsistence activities mentioned by Fray Antonio were fishing, which was secondary, and hunting, which occasionally supplied them with a wild peccary. T h e settiement pattern was definitely dispersed: a cluster of houses here, another a mile apart. Houses were built of " t ierra ," secured on the outside by strong poles attached with vines and covered with palm leaves. The house dimensions were given by Fray Antonio as 30 or 40 feet long by 20 feet wide. They had two doors, one facing the path, the other facing the forest for quick retreat in case of enemies. Furniture was scanty: a metate for grinding, a ham? mock for sleep, two large vessels for cooking cacao or fermenting chicha, gourds as drinking cups, net bags for storage. Of these details, the most interesting are Fray Antonio de la Rocha's remarks on the social and political organization of the Doraces and Zuries. Since the purpose of the trip he describes was, pre? cisely, to negotiate for a place to build a church, he had to deal primarily with individuals who had authority in the different groups. Wha t we gather from Fray Antonio's scattered re? marks was that there were three types of people with authority: caciques (chiefs), cabras (warriors?) and elders. Tha t membership in all of these categories was ascribed and the people in it constituted a social group with rights and obligations in common, seems questionable. There were only one or two caciques per 'provincia' (territory). There was an older cacique ('cacique mayor*) and then, in some instances, a 'cacique menor,' which we can conclu? sively gather from several remarks was the 'nephew' of the first. The title to the chiefdom was not in? herited by a chief's son, but by his nephew (my em? phasis) : "This cacique Nori did not rule because here the sons do not inherit but the nephews do, and the latter give him a hand in the government since one will inherit h im" (de la Rocha, 1964, p. 116). It seems possible, therefore, to postulate that chiefly title was inherited matrilineally. Fray Antonio de la Rocha mentions at least one instance when succession between uncle and 'nephew' was a source of friction within a group. There were a number of neighboring Indian groups with whom the Doraces and the Zuries traded or fought. First, there were those Indians probably of mixed backgrounds, who were living in the town of Alanje. By the time Fray Antonio's account was written, the encomienda system had been abolished, but some Indians had stayed in town. T o the south? east of the Doraces and Zuries were the Guaymi, an enemy tribe with whom they fought for many wars. T o the north were the Northern Guaymi. Near the Bahia de Almirante in Bocas del Toro was a group called Duy, who had been friendly before but were now enemies. Toward the southern coast were the Dolegas, "who are the strongest of all because they live in a town together under one head and the others don't dare" (de la Rocha, 1964, p. 104). Near these were the Quer^balos and then the Bugabas. T h e location of these people was probably at or near the towns bearing their names today. I t was the custom of the Spaniards either to give a new town the name of a saint or to keep the aboriginal name of a group as the new town name. We have no way of guessing what were the cultural and linguistic links between these groups. All we know is that the Doraces and the Zuries spoke the same language and intermarried extensively. This did not deter them from being in a precarious state of friendliness, for they were constandy accusing each 80 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 Other of being abusive and of being traitors. "This is the best and enjoyable conversation which they have, saying bad things about each other, not only those of different nations, but those from one house about another, and this way they never live in peace, and in their gatherings when one says yes, the other says no, without more reason than passion or kinship" (de la Rocha, 1964, p. 115; my translation). A third type of activity which Fray Antonio describes is trade, which apprendy was open to anyone. Out of the mountains, they extracted several types of resins, Achiote (Bixa Orellana), and cacao; these, plus blankets which women wove out of cotton, were traded to the Spanish in the savanna town of Alanje in exchange for axes, machetes, beads, and dogs. They then returned to the highlands and had 'ferias' (ad hoc get together for the purpose of exchange) with their neighbors, the Dolegas, where a large amount of bargaining took place. Fray Antonio complains that they didn't gain anything in these, for they exchanged what they traded at Alanje for the same quantity of the original products. In summary, what emerges as a general picture of the Chiriqui Highland groups from Fray Antonio's account is a number of territorially defined groups (provincias), which were in either precarious alliance, or at war with each other. Neither the institution of chieftainship, nor war, nor trade was a powerful enough mechanism to fuse a number of these groups together. Within each of these groups there were people of different rank, but there is no evidence, except for the chieftainship, and this was weak, that status was inherited rather than based on achievement in war or in age. Absent were such features of class or occupational groups as have been ascribed to the Circum-Caribbean chiefdoms: multivillage confederations, permanent warriors, a strong class system, etc. The reason for their absence couldn't solely have been the Conquest, for though the Spaniards had been successful in killing many Indians and reducing others to serfdom in the coaistal savannas and in the islands, those of the highlands were still relatively untouched. INFERENCES Turning now to the problem of linking the last phase in the Chiriqui Gulf with the ethnohistorical groups described by Fray Antonio, I think we can assume that the people he mentions?the Zuries, Doraces, Dolegas, Bugabas, and Quer^balos?were the descendants of the makers of 'Classical Chiriqui' ceramics. The distribution of these groups in a roughly triangular area north and east of the town of David corresponds exactly to the distribution of 'Classical Chiriqui' pottery as plotted by Osgood (1935, fig. 1, p. 239). At least some of this pottery was made until the time of the Conquest (p. 50) and perhaps even later. If the Spaniards did not describe Indians making pottery, it was probably because pottery-making was hardly a new and novel art to the Spanish. At any rate, since the Chiriqui Phase in the Gulf is defined, precisely, by the appearance of types belonging to 'Classical Chiriqui' pottery, it is probable that it was the ancestors of one of Fray Antonio's groups who moved into the islands several centuries before the Conquest. Migrations of people to the islands, during and perhaps before the Chiriqui Phase, may have been due to a constant state of warfare and fission among mainland populations. There is no reason to believe that this was a condition brought on by the Spaniards. For one thing, the highland groups were relatively untouched when Fray Antonio visited them, and they were constantly being involved in raids and small- scale fighting. Also, all of the descriptions of Pana? manian groups from the Cocl^ and Azuero area, and from as far east as Darien, stress the bellicosity of the Panamanian Indians. Hostility was not only inter- group, but also within a single group; fission may have been caused occasionally by problems of suc? cession to political power. The shift in habitation from the mainland to the islands did not necessarily entail a whole new adapta? tion to a different environment. Before the Spanish established their towns on the coast, there must have been many Indians living near the coast and using the resources of the sea to supplement their diet. This was certainly so in Punta Burica, as mentioned by Cerezada (1883). Furthermore, a fairly large popu? lation could live comfortably on the islands by culti? vating additional maize fields on the larger islands nearby or on the mainland. Living on islands may have encouraged a more nucleated settlement pattern than Fray Antonio de- W H O L E VOLUME EUROPEAN CONQUEST 81 scribes for the highland Indians. He does mention, however, the case of the Dolegas, who lived in one single clustered town under one chief, but this seems to have been a rather atypical situation. What seems more certain is that each island group was unified under a single chief, whatever were the other ranks present in the society. This was precisely the situation found by Cerezada (1883) 100 years before, when he visited some other western Panamanian Islands. Trade between peoples on different islands, as well as with groups on the mainland, was probably exten? sive, as in Fray Antonio's days. Perhaps some of the pottery found in the Chiriqui Phase, whose proveni? ence was the Cocl^-Azuero area or Costa Rica, was traded directly by sea, using the islands as stopping places for 'ferias.' Many of the items which they traded may have just disappeared, judging from the items that were included in a man's household in Fray Antonio's time: cotton blankets, hammocks, and net bags. Of the other objects mentioned by the Father, large pots and metates, we have many examples in the excavations. 284-757 0 - 6 8 ? 7 Chronological Summary and Comparisons with Other Areas CHRONOLOGY OF CHIRIQUI During the 1961 project, we surveyed the coast from Punta Burica in Chiriqui to the western half of the Azuero Peninsula looking for preceramic sites like Cerro Mangote (McGimsey, 1956) or early ceramic sites like Monagrillo (Willey and McGimsey, 1954). We were unable to find any such early sites. Large shellmiddens were observed in Veraguas near the town of Mariato in Montijo Bay, but these have been tested subsequently by McGimsey (personal communication) and found to date after A.D. 1. Since there are no protected bay areas where mollusks can live in large numbers, it is therefore doubtful that early shell-gathering peoples inhabited the Chiri? qui and Veraguas coasts. In the survey we found no evidence of "stepping- stone" settlements, resulting from connections by sea between Middle and South America via the Chiriqui Gulf Our excavations were of limited scale and our archeological phases of such recent date, that nega? tive evidence of this sort does not rule out the possibil? ity that such sites do exist and may be found by future field research. Trade sherds from Parita Bay and Coc\6, found in sites on the Islands of the Gulf of Chiriqui, and Chiriqui sherds in the Diquis Delta of Costa Rica (Lothrop, 1963) indicate that the people of this general area carried on trade over fairly long distances; at least some of this trade must have been by sea. The oldest ceramic complexes found to date in Chiriqui are Scarified Ware and Aguas Buenas Com? plexes. Neither was present in our Chiriqui Gulf excavations. They are nonetheless relevant to our discussion since they establish the lower, or earlier, part of the sequence for the region upon which all three phases of the Gulf of Chiriqui must be super? imposed. Scarified Ware was first recognized as typologically very different from the other wares of the "Classical Chiriqui Culture" by Holmes (1888, pp. 87-90) and MacCurdy (1911, pp. 96-100). They did not pos? tulate, however, that it was older than the other pottery. In recent years it has been shown that Scarified Ware is probably an 'early' complex, older than the 'Classical Chiriqui' ceramics. Some of the work has been done by Haberland (1962), who renamed it La Concepcion complex, taking the name from a large cemetery of SccU"ified Ware, located about 35 kilo? meters west of David in the Solano Finca (fig. 46) at the entrance to the town of Concepcion. I have also excavated at Solano on two different occasions: first in 1959 in the cemetery, and then in 1961 as part of the fieldwork for this report. The area excavated in 1961 was a shallow undisturbed refuse deposit in back of the cemetery (material unpub? lished). The pottery in all of the levels was Scarified Ware; no other complexes were found stratified in our cuts. The shape of the Finca Solano graves has not been described in detail, but the pottery found in them has the following characteristics. All vessels are decorated by the alternation of areas covered by a dense, red slip and areas that are incised. The incisions are closely spaced and roughly parallel to each other; 83 84 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 ? = Modern towns * = Surface collections ? = Sites reported by other investigators FIGURE 46.?Map of the Gulf of Chiriqui area, showing selected modern town and villages, archeological sites on the mainland reported by other investigators, and surface sites in the Burica, San Felix, and Remedios districts. they were made with a pointed instrument when the clay was wet. On some vessels, the incisions are placed in chevron or palm-leaf designs; on others, they were cut into raised fillets, giving the vessel surface texture a sort of 'corrugated' look. The areas slipped in red are usually bands, although entire sections of the vessels, such as tripod legs, the area around the neck, the interior, or even most of the exterior, were also commonly slipped. A detailed study of the surface treatment has resulted in a division of La Concepcion complex into four groups (Haberland, 1962, pp. 384?386). Vessels could also be grouped into a larger number of forms: opened-mouth bowls with rounded bases; bowls with incurving rims resting on short, solid tripods in the shape of duck's feet; tall tripods with long, solid tubular legs; large jars; vessels with animal shapes of birds, turtles, mammals; bowls with flaring sides resting on ringstands, and so forth. All of these groupings have, to date, no chronological or dis? tributional significance. Other complexes showing close relationships with La Concepcion complex have in recent years been found very far east of David near the Veraguas border and in the province of Code. These are Pueblo Nuevo, Guacamayo, and El Limon. The first of these, Pueblo Nuevo, is near the town of the same name (fig. 46), located on the east side of the Tabasara River, which divides the provinces of Chiriqui and Veraguas. Graves here are said to be deep and have shafts (Lothrop, 1959, p. 88). Whether or not they contained Scarified Ware identical to that of La Concepcion is not clear, but they seemed to have contained specimens of Guacamayo Ware (see below) which resembles closely the Scarified Ware. The site of Guacamayo (Harte, 1958) is near Penonome in the central part of the province of Cocl6 (fig. 1). It is one of two sites with identical pottery, the second one recently being reported as El Limon (Stirling, 1964a). How close these sites are to each other is difficult to judge from the published descrip? tions. They are both cemeteries with shoe-shaped tombs that have long shafts and burial chambers to one side, closed off by stones. The typical vessel found in the Guacamayo and El Limon tombs is a tall vase with a flat bottom, a globular lower body, and a slightly constricted neck with flaring rim (Stirling, 1964a, plate 27, a-d). The WHOLE VOLUME SUMMARY 85 exact vessel form has not been found in La Concepcion at Finca Solano, but one with very similar shape has been reported from a private collection, known for certain to have come from here. Although fewer graves have been dug in Penonome than in Concepcion, there is enough material to notice that the similarity in paste, decorative techniques, and shape between the Cocl6 and the Chiriqui Scarified Wares is striking. There is little doubt in my mind that they are part of the same tradition. The Pueblo Nuevo material has been assigned an age of 2 0 4 5 ? 4 5 B.P. or 95 B.C. (Feriz, 1959, p. 186). This date has been corrected for the Suess effect to read 2 2 9 0 ? 4 5 B.P. or 340?45B.C. (Ladd, 1964, p. 12). Unfortunately, the initial charcoal sample was not from Pueblo Nuevo at all (Haberland, 1962, p. 388), but from another site in Chiriqui, as yet un? specified, said to have yielded pottery resembling that of Pueblo Nuevo. Clearly then, there are too many undemonstrated assertions about the similarity of all of these complexes for the single date to be readily accepted as clearly associated with the pottery complex under discussion. Nevertheless, assigning the Scarified Ware to a period between 300 B.C. and A.D. 300 (Baudez, 1963, p . 46) seems quite reasonable on the basis of shared similarities with the Zoned Bichrome Period of Costa Rica. Turning now to the area east of Concepcion, looking toward Costa Rica, we find another 'old' complex called Aguas Buenas, located in the highlands along the frontier. The type site is the Haciendas Aguas Buenas (fig. 46), just west of the border village of Canas Gordas in Costa Rica (Haberland, 1955 pp. 224?230). T h e site has refuse middens atop hills; these were not dug stratigraphically so that the temporal relationship between the types of pottery found here is unclear. I assume that what is called Aguas Buenas are two wares that closely resemble each other: a Red Ware and a Red Rimmed Ware. These were described on the basis of four nearly complete vessels and a number of sherds. In general, the characteristics can be summarized as: looped han? dles (V-shaped in cross section), slab legs, rich applique ornamentation, incising, and a dark, red slip applied to the whole vessel or to parts of it. Besides the two Red Wares described above, Haber? land (op. cit. p . 228) also found a few sherds of "Scarified Ware as published by Holmes." These were a part of an annular base and straight full feet. In addition, there was a ' rare ' Ware, one with geo? metrical paintings on a cream background. Like Scarified Ware, the Aguas Buenas complex is apparently also widespread. At least six sites, among them Barriles (fig. 46) on the outskirts of the Volcan Baru (Stirling, material unpublished), have been identified as Aguas Buenas (Haberland, 1960b, p. 13). How these differ from each other, or from the type site, is not known. The problem of comparisons will be made easier when the full description of the Aguas Buenas ma? terial is published. Here we are concerned mainly with its possible chronological placement. I t is the opinion of Haberland (1962, pp. 386-388), who is acquainted with both complexes, that Aguas Buenas and Scarified Ware are typologically very different, and that this difference also shows up in their different geographical distributions. Scarified Ware is found close to David and then farther east; Aguas Buenas is found fairly west of David in the area on both sides of the border. Since sherds of both com? plexes were found mixed at the type site of Aguas Buenas near Cafias Gordas, he placed them in the same period (Haberland, op. cit. p. 388). Coe. and Baudez (1961, p . 513) consider Aguas Buenas as slightly later in time than Scarified Ware, and the monochrome sherds to be "extremely similar to Early Polychrome monochromes of the Tamar indo zone." They believe the deposit was mixed. I agree with Coe and Baudez' judgment, and based on relationships with their well-worked-out sequence would place Aguas Buenas at A.D. 0-300. There may be, nevertheless, more typological continuities between Scarified Ware and Aguas Buenas than Haberland assumes. Some of the vessels seem to me very similar indeed. For this similarity to be translated into valuable comparisons in terms of space and time, more stratigraphic work is certainly necessary. Here I will hazard to predict that they will be found to be local variations, perhaps at differ? ent points in time, of a single, widespread 'early' tradition. The temporal relationships of Scarified Ware and the Aguas Buenas complex to the early Burica Phase in the Gulf sequence are worthy of comparison. The Burica Phase is the oldest of the three phases estab? lished for the Gulf of Chiriqui. Recently the distri? bution of this phase has been considerably expanded by finds from half a dozen sites on the mainland with secondary urn burials, where the only pottery is Isla Palenque Maroon Slipped (Miranda, P^rez, and de la Guardia, 1966; de la Guardia, 1966). Neither Scarified Ware nor what I understand as Aguas Buenas was represented in any of our cuts. Nevertheless, the Burica Phase may turn out in future excavations to be found stratified after Scarified Ware, directly or through the intermediary of the Aguas Buenas Phase. All one can do at present is make com? parisons. The technique of incising was not present in the collections from the Gulf Islands, but it was repre- 86 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 sented abundantly in the collections in the Burica Peninsula (Ranere, Appendix No. 2). The Burica incising is different, however, from Scarified Ware incisions. It occurs less frequently. The prominent characteristic in Scarified Ware, namely the alterna? tion of red slipped bands and areas incised with closely spaced parallel lines, is missing. Likewise, the type of burial shaft graves with Scarified Ware is very different from secondary urn burials in Burica. The continuity may be expressed better in the plain sherds. Scarified Ware is frequendy indistinguishable in paste composition and slip from Isla Palenque Maroon Slipped. Similarities of the Burica Phase to Aguas Buenas may be more marked, but detailed comparisons can? not be made until that complex has been published with detailed descriptions. No radiocarbon dates are available as yet to estab? lish the beginnings of the Burica Phase. I have never? theless placed them at A.D. 300-500. This estimate is based on the presence of Cocl6 sherds, probably Early Code, and a possible Venado Beach sherd in levels of this phase. Also, a tall pedestal base with cutout tri? angular designs has been found in the Burica Phase site of Tinajas (de la Guardia, 1966, fig. 13). It is identical to the one from Code, illustrated by Lothrop (1942, p. 167, fig. 337c) and assigned by him to his Early period. The exact duration of the Cocl^ tradition is not known, but it does not seem to have started earlier than A.D. 500. Ladd places Early Code in "the second half of the first millennium A.D." (Ladd, 1964, p. 222), together with Venado Beach which has two different dates: 1750?60B. or A.D. 200 and 1125 ?65 B.P. or A.D. 825 for two separate burials. Lothrop (1960, p. 96) notes that "No stylistic differences be? tween the burials were noted during excavation and there was no reason to expect much difference in age." Ladd favors the latter date. The terminal date of A.D. 800 for our Burica Phase is tentatively based on the radiocarbon date for the next phase. Because neither the San Lorenzo nor the Chiriqui Phase is represented alone in the surface collections from the Burica Peninsula, Ranere (Appendix 2, p. 116) is of the opinion that the two phases may be con? temporaneous in that district. This may well be the case. In the Gulf Islands, however, there does seem to be a chronological difference between the two phases in spite of overlap at the sites where both phases are present. The maximum frequencies of the San Lorenzo pottery types occur earlier in the stratigraphic columns than the appearance of the Tarrago Bisquit Ware and the Villalba Red Streaked types. In this connection it is interesting to note that a San Lorenzo Phase site has been found recendy at Batipa, on the mainland (fig. 46), near our site of El Cangrejal (SL- 1). Batipa also has the San Lorenzo Phase predomi- nantiy represented, with sherds from the next phase sparsely represented. This duplicates the situation in the Burica collections but not at El Cangrejal, where the San Lorenzo types occurred exclusively. The next occupation in the Chiriqui Gulf is the San Lorenzo Phase which I have tentatively placed between A.D. 800 and A.D. 1100. This estimate accords with one radiocarbon date for this phase of 930?100 B.P. or A.D. 1020 (M-1308), which comes from a charcoal sample from Site SL-1, Pit No. 3, Level 60-70 cm.; the entire occupation of this site belongs in the San Lorenzo Phase. An estimate for the end of this phase is around A.D. 1100, based on estimates for the beginning of the next phase. The beginnings of our Chiriqui Phase should be placed in the Gulf at around A.D. 1100. A carbon - 1 4 sample from Las Secas (IS-11, Pit No. 2, Level 40-50 cm.) yielded a date of 115 ?100 years ago (M-1309), but this date is clearly wrong since Las Secas Islands have not been occupied by man since the 17th century (p. 78). A terminal date for the Chiriqui Phase should be placed at the time of the Conquest (A.D. 1500) or slightiy later. A grave con? taining Alligator Ware has been found, together with iron tools (Stone, 1958, p. 48). Alligator Ware, though rare in our sites, does occur at Site IS-11 (fig. 10) toward the end of this phase. Furthermore, a Tripod Foot, Mode 12, which in Costa Rica has been found in association with an iron knife (op. cit., p. 50, fig. 7b), also appears at Site IS-11. The entire occupation of this site falls within the Chiriqui Phase. It seems quite possible that the Chiriqui Phase could have begun in the Chiriqui Highlands prior to its introduction into the Gulf at around A.D. 1100. Otherwise, 400 years, from A.D. 1100 to A.D. 1500, seems too short a time span for the development of the "Classical Chiriqui Culture." Future excavations in the highlands may push back the date for this phase and help subdivide it into subphases, but nothing at present modifies our estimate of the time span of the Chiriqui Phase in the Gulf. WHOLE VOLUME SUMMARY 87 RELATIONSHIPS WITH NEIGHBORING PANAMANIAN AREAS The geographical boundaries of the Chiriqui se? quence have been extended by Ranere's welcomed study (Appendix 2, p. 107) of the surface collections made in 1961 from 25 sites east and west of the Gulf Islands, in the districts of Punta Burica, San Felix, and Remedios (fig. 46). His study should be con? sulted in detail. Here we will give only a short sum? mary of his finds for the purpose of comparison. From the district of Punta Burica, Ranere (Appen? dix 2) analyzed a total of 826 sherds from 17 surface sites; he divided these sherds into types, appendage modes, and plastic motifs. In addition, he analyzed a total of 70 sherds from a one-meter excavation at site BA-8, which he found to be unstratified. One of the Punta Burica sites (BA-9) yielded two sherds belonging to the Scarified Ware and one sherd belong? ing to the Aguas Buenas complex. This same site (BA-9), plus nine other sites, also had the Burica Phase represented by large amounts of type Isla Palenque Maroon Slipped. The Burica Phase and the San Lorenzo Phase occurred together at three additional sites in Punta Burica. Two pottery types that do not occur in the Gulf Island sites were found here: Quebrada Baules Brushed Ware and Balsa Polished Ware. Only two of the Punta Burica sites had the Chiriqui Phase represented by a couple of sherds. Some of Ranere's data from Punta Burica raise interesting questions. The fact that he found all three "old" complexes?Scarified Ware, Aguas Buenas, and the Burica Phase?represented at the same site (BA-9) suggests that this is the area to search in the future for clarification of the geographical and temporal relationships between these complexes. I have interpreted all three complexes as being varia? tions, at different points in time, of a widespread tradition (p. 85), but it is possible, of course, that they are, rather, geographical variations at the same point in time. A few sherds found in the surface is too scanty an evidence, however, on which to base any premature statements in this direction. Another un? explained phenomenon is the paucity of evidence in this district for the Chiriqui Phase. Hundreds of vessels belonging to the "Classical" Chiriqui wares have been reported from cemeteries in this area. Turning now to the collections analyzed by Ranere (Appendix 2) from eight sites in the districts of San F^lix and Remedios (fig. 46), totaling a number of 2,623 sherds, we see that both the San Lorenzo and Chiriqui Phases are represented here, but that the Burica Phase is totally missing. Except for two sites, one of which (RE-3) may have only the San Lorenzo Phase represented, and another (RE-1) which may have only the Chiriqui Phase represented, the other six sites had sherds belonging to types found in both phases. Ranere (Appendix 2, p. 115) mentions that certain ceramic features such as plastic decorations are present in the San Felix and Remedios district but are missing from the Gulf Islands and the San Lorenzo district to the west. These features have in turn close counter? parts in the pottery from Veraguas. Contrary to what was thought 20 years ago (Lothrop, 1948, p. 143), the Chiriqui and Veraguas "cultures" were not that separate or distinct. Neighboring Chiriqui on the east is the province of Veraguas, which is known archeologically from Lothrop's study (1950) of grave vessels. He de? scribed these under one monochrome ware belonging to a single "culture," contemporaneous with the Chiriqui and Coc\6 "cultures." Typical forms of Veraguas pottery rest on strapped feet and tall pedestal bases; some are effigy vessels. Since these features also occur in the San Lorenzo Phase but are absent from the other two Gulf Phases, I have postulated a wave of influence from Veraguas into the Gulf at this time. If this hypothesis is borne out by future work in Veraguas, the beginnings of the "Veraguas culture" may be set at around A.D. 800. Older complexes from Veraguas, as yet undescribed, have been dated as early as A.D. 100 (McGimsey, personal communication). The end of the Veraguas "culture" may be placed at the time of the Conquest. Trade vessels from Veraguas have been found in the mainland graves of Chiriqui; the jaguar metates of both of these areas are virtually identical and Lothrop (1950, p. 85) believes that at least 56 percent of the metal products found in Chiriqui were imported from Veraguas. These associations assure the partial contemporaneity of both "cultures" (fig. 47). The connections between the Chiriqui Gulf and the Cocl6-Azuero area are noticeable as much in subtle aspects of vessel shape (p. 36) as in actual 88 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 TIME SCALE 1700 1500 1300 1100 9 0 0 700 5 0 0 300 100 AD BC 1 0 0 300 CHIRIQUI GULF ^ 1 1 CHIRIQ Ph / SAN U Ph Ul M S E DRENZO MSE BURICA PHASE CHIRIQUI MAINLAND A CLASSIC CHIRIQUI _ BURICA PHASE A . AGUAS BUENAS. (BARRILES) /K SCARIFIED (LA CONCEPCION) VERAGUAS CLASSIC VERAGUAS . A _ MARIATO PHASES 1 - i PUEBLO NUEVO PARITA BAY LA ARENA - EL TIGRE _ HERRERA PHASE LATE COCLE MACARACAS A . EARLY COCLE 1 SANTA MARIA COCLE _ HERRERA PHASE LATE COCLE A _ EARLY COCLE t SANTA MARIA .GUACAMAYO EL LIMON VENADO BEACH . VENADO BEACH FIGURE 47.?Chronological chart of regional archeological sequences in western and central Panama with approximate time periods indicated. The Burica Phase and Aguas Buenas estimates may be older by two or three hundred years than shown here, and the Herrera Phase of Parita Bay and Code should be placed at A.D. 1300. trade sherds found in levels of most of the Gulf sites (tables 1^). Levels belonging to the Burica Phase contain a few sherds, unfortunately too eroded to be assigned with certainty to Early or Late Cocl^, though there is no doubt that they are Code. Their early position in our sequence and their association with Venado Beach sherds strongly suggest, however, that they are Early Code. The Early Cocl^ Phase, defined by Lothrop (1942) at Sitio Conte, is also represented at sites in the Parita Bay and at Venado Beach in the Canal Zone. Ladd (1964, p. 222) places Early Code in the second half of the first millenium A.D. The associa? tions with our Burica Phase material would favor a date 300-500 A.D. and lasting undl A.D. 800. One sherd definitely identified as Late Code and 25 Macaracas sherds of the Higo and Pica Pica varieties (Ladd, 1964, pp. 97-120) were found in levels belonging to our San Lorenzo Phase. Late Code is the most recent of the two phases separated at Sitio Conte by Lothrop (1942) and also found at the Parita Bay sites. Macaracas is an Azuero poly? chrome type representing a local variant of Late Code polychromes that is also found at Sitio Conte (Ladd, 1964, pp. 95-120). Neither the Late Cod6 Phase nor Macaracas Polychrome Varieties have been assigned date estimates in the literature; cross- datings with the San Lorenzo Phase and the begin? nings of the Chiriqui Phase would, therefore, favor a range of years for Late Cocl^ and Macaracas be? tween A.D. 800 and A.D. 1200 or 1300. Following the first few levels of the Chiriqui Phase, there is a hiatus in trade material in levels correspond- WHOLE VOLUME SUMMARY 89 ing to the middle part of this phase. Then in the top levels, representing the terminal part of this phase, are found trade sherds belonging to the Herrera Phase, Azuero ceramic group in Parita Bay, thought by Ladd (1964, p . 223) to postdate Late Code . These relationships help estimate a date for the Her? rera Phase between A.D. 1200 or 1300 to A.D. 1500. What has been listed in tables 1 and 4 as Azuero belongs to the ceramic tradition from this area, but the phase. Late Code or Herrera, is not known with certainty. It is interesting to note that Ladd's chron? ological arrangement of the Parita Bay Phases and my arrangement of the Gulf sequence, done inde? pendently from each other before the results of the carbon -14 tests were in, agree almost perfectly. T h e only problem is a puzzling Late Code sherd found in Level 70-80 cm. of Site IS -11 , together with Herrera Phase sherds. This is probably a case of mixture since one isolated sherd does not serve to invalidate his or my sequence. C H R O N O L O G I C A L L I N K S W I T H O T H E R A R E A S In an article which he wrote a few years ago, Baudez (1963) nicely summarized known archeological phases in lower Central America into a framework of six periods. He obtained the information on Chiriqui by correspondence in 1962 from McGimsey, who had made a preliminary inspection of the collections while in the field (McGimsey, 1961). I have suggested below some revisions in the names used for the Chiri? qui Phases?in their description and in their chrono? logical placement. These are the inevitable changes that come with a full analysis of the material and in no way should be interpreted as detracting from the value of McGimsey's apt field observations and Baudez' valuable synthesis. The first two periods in Baudez' article need not concern us here since Period I is Preceramic and Period I I precedes any complexes yet found in Chiri? qui. We will start the discussion with Period I I I , manifestations of which appear in Chiriqui, though not in the Gulf. In Period I I I , dating from 300 B.C. to A.D. 300, Baudez (op. cit. p. 46) has placed the following phases: Scarified Guacamayo Phase, Vera? guas Province; Scarified Phase, Chiriqui Province; Early Diquis Phase (pottery from the deepest levels in Lothrop's excavation in the Diquis region in south? western Costa R ica ) ; Zoned Bichrome Period, north? western Costa Rica ; Aviles and San Jorge Phases, Isthmus of Rivas. The placement of the Scarified Phase in this period seems quite reasonable. The similarities' between Scarified Ware and some of the pottery types of the Zoned Bichrome Period in northwestern Costa Rica have been stated by Coe and Baudez (1961, pp. 512 and 513) as follows: In the Chiriqui area of Costa Rica and Panama, the group dubbed "Scarified Ware" by Holmes (1888) is quite similar to some of our zoned bichrome varieties, especially to Bocana Zoned Incised. Here again zones of parallel incised lines are opposed to red-slipped areas; as one moves south on this early horizon, it is interesting to note that the incised areas increase in size at the expense of the color areas. Baudez has been kind enough to show me his mate? rial from the Valley of Tempisque in the Province of Guanacaste, Costa Rica. There is little doubt in my mind that his Variete Non Peignee of the Bocana Incise Bichrome in the Catalina Phase of the Zoned Bichrome Period (Baudez, 1967, pi. 19 C, E, I, J ) is similar to the Scarified Ware, especially to the Zone Scarified variety of Haberland (1962, pp. 384-385). The decoration in both varieties consists of slipped surfaces alternating with zones of individually made, irregularly spaced parallel incisions. In spite of the above similarities. Scarified Ware lacks the two colors, the dentate rocker stamping, the fork punctations, and the use of multiple brush to produce black lines, all characteristic of the Zoned Bichrome Period. Turning now to the Diquis Delta, we find a Fugitive Red Ware and Brown Wares in the Early Mono? chrome Horizon, the oldest phase here (Lothrop, 1963). Only the Brown Wares have been described in full (op. cit. pp. 45-65) ; they occur from top to bottom of the Diquis refuse. The technique of Parallel Line incising as applied to the large jars of the Brown Ware is "definitely older than any painted technique except negative painting" (op. cit. p. 55). This technique consists of "incised bands of parallel vertical lines on the shoulders, interspaced with incised conical lugs which project about 2 inches (5 centimeters)" (op. cit., p. 57). The same kind of decoration also occurs in slightly smaller jugs. It is not associated with any slips or paints, the name Brown Ware being based on the natural color of the clay itself. The resemblances between the pottery of the Early Diquis Phase and the Scarified VVare of Panama are 90 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 based solely on the technique of parallel-line incision which they share. The shapes, the presence of lugs, the absence of a slip, the large size, characteristics found in the Brown Ware jugs of the Diquis Delta, are all missing from the Chiriqui Scarified Ware. Lothrop (op. cit. p. I l l ) does not draw any connections between the two. I am of the opinion that the similarities of Early Diquis Monochrome Horizon typified by the Fugitive Red and Brown Wares are with our Burica Phase. The Pacific coast of Nicaragua forms one arche? ological area with northwest Costa Rica; together they form the Greater Nicoya subarea. I do not see any close similarities between the pottery types of the Aviles and San Jorge Phases (Norweb, 1961, p. 12) and the Scarified Wares of Chiriqui. In short, during Period III of Baudez, the closest relationships of the Scarified Ware of Chiriqui are with types of the Zoned Bichrome Period of northwest Costa Rica; the most distant are with the Aviles and San Jorge Phases of Nicaragua. In Period IV, dating from A.D. 300 to A.D. 500, Baudez (1963, p. 47) has placed, among other phases, the so-called Thin Red Phase of Chiriqui. The pottery type referred to is the thin variant of our type, Isla Palenque Maroon Slipped. It was provisionally identified in the field by McGimsey who tentatively named the phase after the type. Here, I prefer a geographical name for the Burica Phase and tend to span the bulk of its occupation in the Gulf between Period IV and Period V, together with Early Cocl^ and Venado Beach. In Period V, dating between A.D. 500 to A.D. 800 Baudez (1963, op. cit. p. 48) includes the following phases: Early Code Phase, Parita Bay; El Hatillo Phase, Parita Bay; Alvina Phase, Parita Bay; Venado Beach Phase, Canal Zone; Fine Red Line Phase, Chiriqui Province; Early Polychrome Phase, north? west Costa Rica; Curridabat Phase, Highland Costa Rica. From this period the Fine Red Line Phase of Chiriqui, which is our San Lorenzo Phase, should be removed, renamed, and placed in the subsequent period. Instead, the Burica Phase should be included here also; in terms of trade wares, it is contempo? raneous with Venado Beach and Early Cocl^ in spite of not having the bold geometric multicolor painting that is common to other phases of this period. It may be added that the Curridabat Phase of Highland Costa Rica does not have this painting either, nor does the bulk of the Early Diquis Monochrome ma? terial. Baudez (op. cit. p. 49) includes the following phases in Period VI, dating between 800 A.D. and the Con? quest: Late Code and La Mula Phases, Parita Bay; Classic Veraguas Phase, Veraguas Province; Gross Red Line and Armadillo Phases, Chiriqui Province; Classic Boruca Phase, southwestern Costa Rica; Cartago Phase, central Costa Rica; Middle and Late Polychrome Phases, northwest Costa Rica; White Polychrome Phase, southwestern Nicaragua. It now seems clear from Ladd's subsequent study (1964, pp. 51-66) that El Hatillo Phase, placed by Baudez in Period V, refers not to a phase but a pottery type with several varieties belonging to the Herrera Phase, which comes after Late Cocl6 and, therefore, should be put in Period VI. As for the Classic Veraguas Phase, its beginnings may be contempo? raneous with our San Lorenzo Phase. The Gross Red Line and Armadillo Phases corre? spond to my Chiriqui Phase. Gross Red Line is the type now called Villalba Red Streaked and Armadillo refers to my Tarrago Bisquit Ware. The beginnings of the Chiriqui Phase are contemporaneous with Late Coc\6, but the trade wares in the latter part of this phase belong to the Herrera Phase of Parita Bay. In terms of the Panamanian Phases, I would there? fore divide Baudez' Period VI into two others: Period Via, which would include Late Cocl^, the San Lorenzo Phase, and the beginnings of the Chiriqui Phase and would last between A.D. 800 and 1200; Period VIb, which would include the latter part of the Chiriqui Phase and the Herrera Phase and would last between A.D. 1200 and 1500 or slightly later. WHOLE VOLUME SUMMARY 91 FINAL REMARKS Taking an overview now, we can say with some con? fidence that Chiriqui archeology in the last 10 years has begun to emerge out of the purely "flat" historical stage where it has languished since the time of the grand pioneers. Holmes (1888) and MacCurdy (1911), who wrote their descriptions of Chiriqui ceramics more than 50 years ago. In fact, the history of Chiriqui can now be extended to the period between 300 B.C. and A.D. 300, roughly corresponding to the Late Formative and Proto-Classic in Mesoamerica, when there was in the western par t of the Isthmus a wide? spread ceramic horizon represented by Scarified Ware, or variations thereof, found in a number of interrelated sites: La Concepcion, Pueblo Nuevo, El Limon, Gua? camayo, and, perhaps, Aguas Buenas. Most of these complexes remain ill-defined, based as they are solely on funerary finds whose associations and interrela? tionships are far from clear, with virtually nothing being known of the culture that produced them. No matter how these complexes are defined and redefined in the future by further investigation, it seems certain that they will be found to precede by many hundreds of years the better known "Classical" Chiriqui ceramics. I t is precisely this chronological gap between the "ear ly" complexes and "Classical" Chiriqui Culture that the excavations in the Gulf can now adequately fill. The oldest phase here, the Burica Phase, is prob? ably a continuation of the old complexes at a later point in time (p. 85). Around A.D. 800, a second tradition, typified by the San Lorenzo Phase which emphasized linear decoration by painting rather than a slip and incisions, appeared in the Gulf in a restricted area around the Estero de Horconcitos and the Bahia de Muertos. It was influenced by the Veraguas style in the neighboring area to the east and may have even borrowed shapes and decorative techniques, without ever adopting polychrome painting, from as far away as the Cocle-Azuero area. The transition from the San Lorenzo Phase to the subsequent Chiri? qui Phase was a gradual one, arguing for some con? tinuity in the population. By at least A.D. 1100 in the islands, and perhaps much earlier on the mainland, the third tradition, "Classical Chiriqui," had attained a distinctiveness and a coherence that only the Spanish conquest could disrupt. Sameness in ceramic style should not, however, imply political cohesion or cul? tural or linguistic unity. The various Indian groups that the Spaniards found in Chiriqui showed signifi? cant variations in all three respects. Although many questions about Chiriqui archeology remain unanswered by our study, the limited excava? tions in the Chiriqui Gulf have nevertheless given us the first basic time sequence on which we shall be able to elaborate in the future a fuller picture of the culture- history of this important part of the New World. Literature Cited ANDERSON, G . L . G . 1914. Old P a n a m a and Gastilla del Oro . Boston, The Page Go. AcosTA, JOAQUIN 1848. Gompendio historico del descubrimiento y colonizaci6n de la Nueva Granada en el siglo decimo sexto. Paris. BAUDEZ, GLAUDE F . 1963. Gultural development in lower Gentral America. In Aboriginal cultural development in Lat in America : An interpretative review. Meggers and Evans, editors, Smith? sonian Misc. GolL, vol. 146, no. 1, publ. 4517, pp . 45-54, Washington, D.G. 1967. Recherches archeologiques dans la Vallee du Tempisque, Guanacaste, Gosta Rica. T ravaux et Memoires de ITnstitut des Hautes Etudes de I 'Amerique Latine, no. 18, Universite de Paris. BISHOP, WILLIAM H . 1961. A summary of field notes and personal impressions of the geology of a portion of south? western Panama. [Unpublished manuscript on file at Museum, Univ. of Arkansas, Fayetteville, Ark.] GEREZADA, ANDRES DE 1883. I t inerario y cuentas de Gil Gonzalez Davila por el Tesorero Andres de Gerezada. In Gosta Rica, Nicaragua y Panama. D. Manuel de Peralta, ed. Madrid-Paris , 1883. GoE, M I C H A E L D . and BALTDEZ, GLAUDE F . 1961. T h e Zoned Bichrome Period in northwestern Gosta Rica. American Antiquity, vol. 26. pp. 505-515. GuERVo, ANTONIO B . , editor. 1891. Goleccion de documentos ineditos sobre la geogrcifia y la historia de Golombia, tomo I I Bogota; relaci6n sobre la costa Panamena en el M a r del Sur, Ano de 1631 por Diego Ruiz de Gampos. DE LA GUARDIA, R O B E R T O 1966. El Sitio Tinajas. Boletin del Museo Ghiricano, no. 3, pp. 7-13, David, Panama. DE LA R O C H A , F R A Y ANTONIO 1964. Tesoros Verdaderos de las Yndias (1682). Seccion Documental , editor: Rodrigo Mir6 , H o m b r e y Gultura, tomo I, no. 3, Diciembre, pp. 82-132, Panama. EKHOLM, GORDON and EVANS, GLIFFORD 1962. T h e interrelationships of New World cultures: A coordinated research program of the Institute of Andean Research. Akten des 34. Internationalen Amerikanistenkongresses, Wien, 18-25 Ju l i 1960, pp . 253-278. F E R I Z , H A N S 1959. Zwischen Peru und Mexico. Koninklijk Instituut voor de Tropen, no. 134. Amsterdam. GOLDMAN, EDWARD A. 1920. M a m m a l s of Panama . Smithsonian Misc. GolL, vol. 69, no. 5, Washington, D.G. HABERLAND, W O L F G A N G 1955. Prel iminary report on the Aguas Buenas complex, Gosta Rica. Ethnos, no. 4, pp. 224 - 230. Stockholm. 93 94 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 1958. A re-appraisal of Ghiriquian pottery types. Actas, 33rd Int . Gong, of Americanists, vol. 2, pp . 339-346. San Jos6. 1960a. Villalba: A preliminary report. Panama Archeologist, vol. 3, pp . 7 -21 . Balboa, G. Z . 1960b. Cien Anos de Arqueologia en Panama. Loteria, reprint no. 12, Panama , R .P . 1962. T h e scarified ware and the early cultures of Chiriqui ( P a n a m a ) : Akten des 34. Inter- nat ionalen Amerikanistenkongresses, pp. 381-389. Vienna. H A R T E , H . 1958. Guacamaya Ind ian culture. Panama . HARTMAN, C A R L V. 1907. Archeological researches on the Pacific coast of Gosta Rica. Memoirs of the Carnegie Museum, vol. I l l , no. 1, Pittsburgh. HOLMES, WILLIAM H . 1888. Ancient art of the Province of Chiriqui, Golombia. Bureau of Amer ican Ethnology, 6th Ann. Rept . Washington, Smithsonian Institution. K E E N , A. M Y R A 1958. Sea shells of tropical west America. Stanford, Stanford Univ . Press. LADD, J O H N 1964. Archeological investigations in the Parita and Santa Mar ia Zones of P a n a m a . Bureau of American Ethnology, Bull. 193, Smithsonian Insti tution. L A S CASAS, BARTOLOME 1951. Historia de las Indias. Fondo de Gultura Economica 1-111, Mexico-Buenos Aires. LINARES, O L G A 1964. Archeological investigations in the Gulf of Chiriqui, Panama . [Unpublished Doctora l Dissertation, Harvard Univ., Cambridge, Mass.] LINARES DE SAPIR, O L G A 1966. La cronologia arqueologica del Golfo de Chiriqui, P a n a m a . X X X V I Gongreso In ter - nacional de Americanistas, vol. 1, pags. 405-414. Sevilla, Espaiia. L O T H R O P , SAMUEL K . 1919. The discovery of gold in the graves of Chiriqui, Panama . M u s e u m of Amer ican Indians, Heye Foundation, Indian Notes and Monographs , vol. V I , pp . 27-36. New York. 1926. Pottery of Costa Rica and Nicaragua, 2 vols. Heye Foundat ion Contribs. , vol. V I I I , New York. 1942. Code , An archaeological study of central Panama. Memoirs of the Peabody Museum, vol. 8, pt. 2, Harvard Univ., Cambridge, Mass. 1948. T h e archeology of Panama. In Handbook of South American Indians , Bureau of American Ethnology, Bull. 143, vol. 4, pp. 143-167. Washington, D .C. 1950. Archaeology of southern Veraguas, Panama. Memoirs, Peabody Museum, vol. 9, no. 3. Harvard Univ., Cambridge. 1959. A re-appraisal of Is thmian archaeology. Amerikanistsche Miszellen, Mit te i lungen Aus dem Museum fiir Volkerkunde, vol. 25, pp. 87 -91 . H a m b u r g . 1960. C-14 Dates for Venado Beach, Canal Zone. P a n a m a Archaeologist, vol. 3. 1963. Archaeology of the Diquis Delta, Costa Rica. Papers of the Peabody M u s e u m of Archaeology and Ethnology, vol. LI , Harvard Univ. M A C C U R D Y , G E O R G E G . 1911. A study of Ghiriquian antiquities. Connecticut Academy of Arts and Sciences, Memoi r 3, New Haven. M A R K H A M , CLEMENTS 1865. Transla tor and editor of: Narrat ive of the proceedings of Pedrarias Davila in Provinces of Tierra Firme or Castilla de Oro and of wha t happened in the discovery of the South Sea and the coasts of Peru and Nicaragua by Pascual de Andagoya. T h e Hakluyt Society, ser. 1, no. 34, London. M C G I M S E Y , C H A R L E S R . I l l 1956. Cerro Mango te : A preceramic site in Panama. American Antiqui ty, vol. 22, p p . 151-161. 1961. Interrelationships of New World Cultures Project G : Pacific coast of P a n a m a . [Mimeo? graphed report to the Institute of Andean Research Inc. , on the first season of field- work in Panama. ] M I R A N D A , M . , F*EREZ, J . and DE LA GUARDIA, R . 1966. El Sitio Dolega. Boletin del Museo Chiricano, no. 2, David, P a n a m a . WHOLE VOLUME LITERATURE CITED 95 NoRWEB, A L B E R T H . 1961. The archaeology of the greater Nicoya Subarea. [Unpublished manuscript on file in Peabody Museum, Harvard Univ., Cambridge, Mass.] OSGOOD, CORNELIUS 1935. The archaeological problem in Chiriqui. American Anthropologist, vol. 37, pp. 234-243. PERALTA, MANUAL M . DE 1883. Costarica, Nicaragua y Panama en el Siglo XVI: Su Historia y sus Limites. Madrid- Paris: pp. 5-26 entitled. El Gapitin Gil Gonzalez Davila a S. M. el Emperador Carlos v . , Rey de Espaiia, sobre su expedicion a Nicaragua, Santo Domingo, 6 de Marzo, 1524. STIRLING, MATTHEW and STIRLING, MARION 1964a. El Limon, an early tomb site in Gocl6 Province, Panama. Bureau of American Eth? nology Bull. 191, pp. 247-254, Washington, D.C. 1964b. Archeological notes on Almirante Bay, Bocas del Toro, Panama. Bureau of American Ethnology Bull. 191, pp. 255-284, Washington, D.C. STONE, DORIS 1958. Introduction to the archaeology of Gosta Rica. Museo Nacional, San Jose, Gosta Rica. 1963. Cult traits in southeastern Gosta Rica and their significance. American Antiquity, vol. 28, pp. 339-359. WILLEY, G . R . and MCGIMSEY, G. R . 1954. The Monagrillo culture of Panama. Papers of the Peabody Museum of Archaeology and Ethnology, Harvard Univ., vol. 49, no. 2, Cambridge, Mass. Appendix 1 TABLES 1-7 Levels (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 130-140 140-150 150-160 TOTAL Plain Ware Variety C No. % - - - - - - - - 19 1.75 17 3.70 75 7.81 82 18.10 63 44.06 82 46.07 - - TABLE 1.?Frequency of pottery types from Pit No. Plain Ware Variety D No. % 1, 545 74.10 921 72.01 324 64.67 563 61.13 590 50.38 341 36.08 583 64.49 705 67.72 664 66.60 618 57.06 320 69.57 573 59.69 186 41.06 16 11.19 6 3.37 - - Plain Ware Variety H No. % - - - - - 6 0.66 10 .96 8 .80 29 2.68 19 4.13 46 4.79 46 10.15 5 3.50 3 1.69 - - - Plain Ware Variety I No. % 55 2.64 71 5.55 52 10.38 122 13.25 297 25.36 278 29.42 124 13.72 99 9.51 124 12.44 177 16.34 6 1.30 18 1.88 11 2.43 - - ? - - 3, Site Isla Palenque (IS-3) Plain Ware Vziriety J No. % 52 2.49 2 .16 - - - - - - - - - - - - - ? ? - Arayo Polished Line No. % 1 0.08 - 2 .22 1 .09 1 .11 6 .66 7 .67 34 3.41 45 4.16 22 4.78 15 1.56 6 1.32 - - ? ? - Banco Red Line No. % 1 0.08 - - 1 .09 25 2.65 11 1.22 - - - - - - - - ~ ~ - Caco Red No. 13 16 5 27 13 4 6 30 32 28 25 30 25 4 1 1 - Slipped % 0.62 1.25 1.00 2.93 1.11 .42 .66 2.88 3.21 2.59 5.43 3.13 5.52 2.80 .56 10.00 - 284-757 0-68?8 97 98 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY TABLE 1.?Frequency of pottery types from Pit No. 3, Site Isla Palenque {IS-3)?Continued VOLUME 8 Levels (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 130-140 140-150 150-160 T O T A L 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 130-140 140-150 150-160 T O T A L Cangrejal Red Line No. % 10 0 .48 10 .78 4 .80 11 1.19 11 .94 4 .42 12 1.33 6 .58 4 .40 6 .55 5 1.09 17 1.77 6 1.32 1 .70 - Tarrag6 Bisquit Ware No. % 136 6.52 38 2 .97 17 3.39 - Castrell6n Red Slipped No. % 9 0 .77 3 .32 3 .29 1 .10 5 .46 1 .22 2 .21 1 .22 - Villalba fled Streaked No. % 11 0 .53 6 .47 - Cavada Applique and Red Banded No. % 3 0 .14 2 .16 3 .60 - Zapote Red Banded No. % 2 0 .16 2 .40 2 .22 10 .85 2 .21 3 .33 3 .29 3 .30 6 .55 5 1.09 19 1.98 8 1.77 1 .70 - Centeno Red Banded No. % 3 0 .33 4 .38 14 1.40 24 2.22 4 .87 15 1.56 4 .88 - Modes No. % 227 10.89 165 12.90 61 12.18 124 13.46 177 15.12 106 11.22 117 12.94 133 12.78 88 8 .83 102 9.42 27 5.87 109 11.35 42 9.27 12 8.39 15 8 .43 - Horconcitos Red Banded No. % 8 0 .38 3 .23 2 .40 8 .87 9 .77 5 .53 1 .11 22 2.11 14 1.40 13 1.20 5 1.09 3 .31 - Isla Palenque Maroon Slipped No. % 7 0 .34 6 .47 2 .40 2 .22 3 .26 4 .42 1 .11 8 .77 6 .60 9 .83 4 .87 30 3 .13 36 7.95 41 28 .67 72 40 .45 9 81 .82 - Linarte Zoned Red Line No. % 15 0 .72 35 2 .74 29 5 .79 59 6.41 46 3 .93 147 15.56 29 3.21 5 .48 - Pan de Azucar Red Line No. % 1 0 .05 1 .11 2 .17 2 .21 2 .22 6 .58 5 .50 2 .18 7 .73 - Trade Pottery Venado Beach? 1 - Cocl6 (Early?) 1 - Cocle (Late) 1 - M a c a r a c a s (Higo or Pica Pica) 2 23 - Azuero 2 - Grand Total Per Level 2 ,085 1,279 501 921 1, 171 945 904 1,041 997 1,083 460 960 453 143 178 11 13, 132 W H O L E VOLUME APPENDIX 1 99 TABLE 2.?Frequency of pottery types from Pit No. 1 and Pit No. 3, Site El Cangrejal {SL-1) Levels (cm.) PIT N O . 1 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 T O T A L P I T N O . 3 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 TOTAL Levels (cm ) P I T N O . 1 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 TOTAL P I T N O . 3 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 TOTAL Plain Ware No. % 186 86.51 910 83.26 401 78.47 357 76.77 216 87.10 296 86.55 129 81.65 90 81.82 20 90.91 - 1,258 74.44 1,239 74.41 168 46.15 473 80.03 161 78.54 297 73.51 97 64.38 98 69.01 30 73.17 - Horconcitos Red Banded No. % _ - 6 0 .55 6 1.17 9 1.94 2 .81 4 1.17 - - - - 4 .24 9 .54 7 1.93 6 1.02 - 2 .50 1 .68 2 1.41 1 2 .44 - Arayo Polished Line No. % 5 2 .33 37 3.39 21 4 .11 22 4 .73 - 7 2 .05 1 .63 - - - 79 4 .67 120 7.21 29 7.97 17 2 .88 5 2 .44 12 2.97 9 6.16 5 3.52 3 7.32 - Linarte Zoned Red Line No. % 2 0 .93 8 .73 2 .39 2 .43 - - - - - - 2 .12 4 .24 15 4 .12 3 .51 - - - - - - Caco Red Slipped No. % 2 0 .93 33 3.02 24 4 .70 30 6.45 8 3.23 13 3.80 2 1.27 9 8.18 - - 83 4.91 62 3.72 24 6.59 17 2 .88 12 5.85 23 5.69 11 7.53 11 7.75 - - Pan de Azucar Red Line No. % 4 1.86 2 .18 5 .98 4 .86 - 2 .58 7 4 .43 1 .91 - - 9 .53 4 .24 3 .82 3 .51 1 .49 2 .50 - 1 .70 - - Cangrejal Red Line No. % 2 0 .93 11 1.01 4 .78 5 1.08 3 1.21 3 .88 2 1.27 1 .91 - - 25 1.48 18 1.08 8 2 .20 9 1.52 3 1.46 7 1.73 4 2 .74 4 2.82 1 2 .44 - Zapote Red Betnded No. % - 9 0.82 1 .20 2 .43 1 .40 2 .58 1 .63 - - - 7 .41 17 1.02 2 .55 9 1.52 2 .98 1 .23 5 3.42 2 1.41 - - - Castrell6n Red Slipped No. % _ _ 1 0.09 2 .39 - _ - - - _ _ - - 7 .41 2 .12 1 .27 1 .17 5 2 .44 - - - - - Modes No. % 14 6.51 76 6.95 45 8.81 33 7.10 18 7.26 13 3.80 13 8.23 9 8 .18 2 9.09 - 216 12.78 188 11.29 107 29.39 53 8.97 16 7.80 54 13.37 19 13.01 19 13.38 6 14.63 - Centeno Red Banded No. _ - - _ - 2 2 _ - - _ 1 - - - 5 - - - - Trade Pottery? Code (Late?) - - - 1 - - 1 - - - - 1 - - - 1 - - - - % _ - - _ - 0.58 1.27 - - - _ .06 - - - 1.24 - - - - Grand Total Per Level 215 1,093 511 465 248 342 158 110 22 3, 164 1,690 1,665 364 591 205 404 146 142 41 5,248 100 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY TABLE 3.?Frequency of pottery types from Pit No. 1 and Pit No. 2, Site Villalba {IS-7) VOLUME 8 Levels (cm.) PIT NO. 1 0-10 10-20 20-30 30-40 40-50 50-60 60-70 TOTAL PIT NO. 2 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 TOTAL Plain Ware No. 182 238 120 464 447 49 7 - 518 251 357 745 365 120 146 141 104 60 259 125 96 - % 50 28 57.21 44. 12 58.81 64.97 79.03 77.78 - 56.00 47.00 49.65 47.73 42.25 36.36 40 67 47. 16 53.33 49.59 63.02 64.43 53.33 - Arayo Polished Line No. % _ _ 1 0 24 1 .37 6 .76 4 .58 - - - _ _ - 1 . 14 6 .39 3 .35 2 .61 1 .28 1 .33 1 .51 - 6 1.46 3 1.55 2 1. 11 - Caco Red Slipped No. % _ _ 1 0 24 4 1.47 8 1.01 5 .73 - - - _ _ 1 . 19 - 2 . 13 - 3 .91 2 .56 - - - 1 .24 5 2.58 2 1. 11 - Cavada Applique and Red Banded No. % _ _ _ - _ - - - - - _ - - - - 1 O 12 - 1 .28 1 .33 - - 1 .24 - - - Centeno Red Bjuided No. % 3 0.83 6 1.44 2 .74 1 . 13 - 1 1.61 - - - - - - - - - - - - - - 1 .52 - -1 Horconcitos Red Banded No. 2 1 4 4 1 - - - - - 1 2 - - - 1 1 - - 1 - - % 0 55 . 2 4 1.47 . 51 . 15 - - - - - . 14 . 13 - - - . 3 3 .51 - - . 52 - - Isla Palenque Maroon Slipped No. % - 6 1.44 11 4.04 27 3.42 11 1.60 1 1.61 - - 5 .54 5 .94 5 .70 22 1.41 14 1.62 8 2.42 3 .84 2 .67 1 .51 1 .83 8 1.95 4 2.06 60 33.33 - Lincirte Zoned Red Line No. % 1 0 28 1 .24 4 1.47 2 .25 1 . 15 3 4.84 - - 1 11 - 5 .70 7 .45 3 .35 2 .61 - 2 .67 1 .51 - 6 1.46 4 2.06 2 1. 11 - Levels (cm.) PIT NO. 1 0-10 10-20 20-30 30-40 40-50 50-60 60-70 TOTAL PIT NO. 2 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 TOTAL Pan de Azucar Red Line No. % - - - 2 0 25 - - - - 1 11 - 1 . 14 1 .06 5 . 58 - - - - 1 .83 2 .49 2 1.03 - - TarTag6 Bisquit Ware No. % 26 7. 18 45 lO 82 21 7.72 39 4.94 21 3.05 - - - 156 16.86 101 18.91 133 18.50 308 19.73 133 15.39 53 16.06 64 17.83 72 24.08 17 8.72 21 17.36 12 2.92 - - - Villalba Red Streaked No. % 113 31.22 84 20. 10 50 18.38 136 17.24 103 14.97 - - - 212 22.92 154 28.84 182 25.31 383 24. 54 276 31.94 127 38.48 102 28.41 67 22.41 51 26. 15 22 18.18 39 9.49 7 3.61 2 1. 11 - Zapote Red Banded No. % 1 0 28 - 2 . 74 7 .89 5 .73 - - - 3 .32 3 .56 1 . 14 2 . 13 3 .35 - 1 .28 - - - 1 .24 1 .52 - - Red on No. - - - 8 21 1 - - 1 - 1 7 1 1 1 1 2 1 3 1 - White % - - - 1.01 3.05 1.61 - - 11 - 14 45 . 12 . 3 0 . 2 8 . 3 3 1.03 . 8 3 . 7 3 . 52 - - Modes No. 34 33 51 85 69 7 2 - 28 19 32 76 60 14 38 11 16 15 72 38 10 - % 9.39 7.93 18.75 10 77 1O03 11.29 22.22 - 3.03 3.56 4.45 4.87 6.94 4.24 10.58 3.68 8.21 12.40 17.52 20.65 5.56 - Trade Cocl6 (Early?) - - - - - - - - _ - - - - - - - - - - - 6 - Pottery Cocl6 (Late?) _ - 2 - - - - - _ - _ - - - - - 1 - 1 2 - - Grand Total Per Level 362 416 272 789 688 62 9 2,598 925 534 719 1,561 864 330 359 299 195 121 411 194 180 6,692 WHOLE VOLUME A P P E N D I X 1 T A B L E 4.?Frequency of pottery types from Pit No. 1, Site Las Secas (IS-ll) 101 Levels (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 T O T A L 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 T O T A L 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 T O T A L Plain Ware Variety A No. % _ _ - - - - - 2 0 65 5 4. 10 7 7.22 16 36.36 - Caco Red Slipped No. % 3 0.29 - 2 . 5 4 2 . 3 5 - 2 .42 - - 1 1.03 - - Red on White No. % - 2 0 54 2 . 3 5 1 . 2 0 - - - - - - Plain Ware Variety B No. % 268 26.30 84 23.46 80 21.62 76 13.43 82 16.63 79 16.53 105 34.31 48 39.34 35 36.08 15 3 4 09 - Cavada Ap? plique and Red Banded No. % - - - - 1 0.21 3 . 9 8 4 3.28 1 1.03 1 2.27 - Red Slip? ped Bisquit Ware No. % 1 0. 10 - 2 . 5 4 2 . 3 5 5 1.01 2 42 1 . 33 - - - - Plain Ware Variety E No. % 28 2 .75 16 4.47 16 4 32 16 2 .83 35 7. 10 27 5.65 5 1.63 - - - - Chocolate Incised No. % 1 O 10 1 . 2 8 15 4. 10 6 1.06 17 3.50 10 2.09 - - - - - Tarrago Bisquit Ware No. % 492 48.28 158 44. 13 111 30.00 271 47.88 197 39.96 196 41.00 71 23.20 15 12.30 21 21.65 4 9.09 - Plain Ware Variety F No. % 71 6.97 38 lO 61 39 10.54 83 14 66 64 12.98 59 12. 34 16 5.23 - - - - Horconcitos Red Banded No. % 3 0 29 - - 2 .35 2 .41 1 .21 - - - - - Villalba Red Streaked No. 37 8 13 3 2 5 3 11 2 1 - % 3.63 2 .23 3.51 . 5 3 .41 1.05 . 9 8 9.02 2 .06 2.27 - Plain Ware Variety G No. % - - - 20 4 .06 4 . 8 4 19 6.21 - - - - Linarte Zoned Red Line No. % 2 O 2 0 3 . 8 4 - 2 . 3 5 2 41 - 4 1.31 3 2 .46 3 3.09 - - Zapote Red Banded No. % _ _ - 1 0 27 2 . 35 - 1 .21 1 . 3 3 2 1.64 - 2 4 55 - Alligator Ware No. % - - - - - 1 0 33 - - - - Pan de Azucar Red Line No. % _ - - - - - 1 0 .33 - 1 1.03 - - Modes No. ? 106 10 40 11 46 12 70 12 54 10 87 18 68 22 29 23 25 25 5 11 - ^ 40 17 43 37 95 20 22 77 77 36 - Arayo Polished Line No. % - - - 1 0 .20 1 . 21 3 . 9 8 - 1 1.03 - - Polished Black Plain No. % _ _ 6 1.68 20 5.41 9 1.59 - - - - - - - Black on Red Negative No. % 1 0 .28 2 . 5 4 - - - - - - - - Beige Plain No. % _ _ 1 0 28 21 5.68 20 3 .53 11 2 .23 1 .21 - - - - - Trade Pottery Azuero 1 - - - - - - 1 - - - Azuero? 6 2 - - - 2 - 2 - - - Cocl6 (Late) _ - - - - - - 1 - - - H errera _ - - - - - 3 1 - - - Grand Total Per Level 1,019 358 370 566 493 478 306 122 97 44 3,853 102 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 T A B L E 5.?Frequency of miscellaneous sherds, necks, supports, handles, and rims at Sites IS-3, IS-7, IS-11, and SL-1 (pit numbers in parentheses) Levels (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 130-140 140-150 150-160 TOTAL 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 130-140 140-150 150-160 TOTAL Miscellancovis Sherds IS-3 SL-1 (3) (1) 11 3 - - 11 9 10 11 1 5 - 12 - - 5 - 78 SL-1 (3) 5 5 5 3 - 4 1 1 1 - - - - - - - 25 IS-11 IS-7 (1) (1) 11 2 1 4 4 6 1 3 2 2 1 2 - 11 1 - - - - - - 42 9 IS-7 (2) 2 1 3 2 2 - 1 1 2 - 2 1 3 - - - 20 Necks IS-3 SL-1 (3) (1) 7 1 10 2 4 3 4 2 3 2 2 9 4 5 2 2 - - 48 14 Sites SL-1 IS--11 ,IS-7 (3) (1) (1) 9 8 4 1 - - - - - - - - - - - - 22 Handles 3 1 2 7 3 - 4 - - - - - - - - - 4 16 IS-7 (2) 2 - 1 4 3 - 1 - 1 2 2 3 - - - - 19 Supports IS-3 SL-1 (3) (1) 17 1 19 10 4 10 21 7 27 3 11 1 13 1 10 7 1 9 6 10 4 4 1 - 163 34 SL-1 IS-11 IS -7 (3) (1) (U 24 20 22 10 8 6 8 1 7 6 10 3 12 1 8 4 - - - - - - - - 65 85 1 1 3 3 1 - - - - - - - - - - - 9 IS-7 (2) 2 1 1 3 5 - 1 - - - 1 1 - - - ? 15 Rims Sites IS-3 (3) 38 38 15 30 26 19 16 13 10 3 3 - - - - - 211 SL-1 (1) 5 5 9 4 - 1 - 1 - - - - - - - - 25 SL-1 (3) 13 12 11 1 - 3 1 1 - - - - - - - - 42 IS-11 (1) 14 4 6 9 6 10 3 6 2 2 62 IS-7 (1) 7 9 10 19 13 2 - - - - - - - 60 IS-7 (2) 2 4 7 17 22 1 5 - 3 4 21 9 - - - - 95 IS-3 (3) 147 104 40 70 109 65 75 97 68 76 14 82 36 6 9 - 998 SL-1 (1) 8 51 22 22 15 11 12 8 1 - - - - - - - 150 SL-1 IS-11 IS-7 (3) 165 141 79 48 15 41 14 16 5 - - - - - - - 524 (1) (1) 61 23 24 21 30 32 47 55 38 50 65 4 47 2 15 8 2 - - - - - - 337 187 [S-7 (2) 20 13 20 50 28 13 30 10 10 9 46 24 7 - - - 280 WHOLE VOLUME A P P E N D I X 1 T A B L E 6.?Frequency of stone artifacts at Sites IS-3, IS-7, IS-11, and SL-1 103 Excavated Levels Site IS -3 IS-7 Pit 3 1 2 Level (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-86 80-90 90-100 100-110 110-120 120-130 130-140 140-150 150-160 160-170 0-10 10-20 20-30 30-40 40-50 50-60 60-70 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 Pound- Projec- Pecking ing Pebble Sand- Scrap- Net or Sharp- tile Flakes stones anvil Manos Metates Celts polishers stone ers line ening points stones saws weights stones - - - - 1 - 2 - 1 - - - - - - - 2 1 1 - - - 8 1 - 1 - 2 - - - - - - - 1 i _ _ _ _ _ _ 4 1 - - - 2 - 2 - - - - 2 : ! ! ! ! ! i ! i ! 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 I I I C M ? I I I I C M I I I 1 1 1 1 1 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 104 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY TABLE 6.?Frequency of stone artifacts at Sites IS-3, IS-7, IS-11, and SL-1?Continued VOLUME 8 Excavated Levels Site IS-11 SL-1 Pit 1 1 3 Level (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 Pound- Projec- Pecking ing Pebble Sand- Scrap- Net or Sharp- tile Flakes stones anvil Manos Metates Celts polishers stone ers hne enmg points stones saws weights stones l l l l l l l l l l l l l l l l l l l l l l 1 ? ? 1 1 1 1 1 1 1 1 - 1 1 1 1 1 1 1 1 1 1 CO 1 ? ? C M 1 I ? ? 1 1 l l l l l l l l l l l 1 ? 1 1 1 1 1 1 1 ? 1 1 1 1 1 C M 1 1 1 1 1 1 CM 1 1 ? 1 ? 1 C M 1 1 1 l l l l l l l l l l l l l l l l l l l l l l 1 1 1 1 1 1 1 ? ? 1 1 1 ? 1 1 1 1 1 . 4^ 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ? 1 1 1 1 1 1 1 N O 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 ? ? 1 1 1 1 ? 1 1 1 1 ? 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 t 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 _ _ 1 ! ! _ _ _ _ _ _ _ WHOLE VOLUME APPENDIX 1 TABLE 7.?Animal bones from Sites IS?3, IS-11, and SI-1 105 Excavated Levels Site IS -3 IS-11 SL-1 Pit 3 1 1 Level (cm.) 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 100-110 110-120 120-130 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 0-10 10-20 20-30 30-40 40-50 50-60 60-70 Fish Uniden? tified species X X X X X X X X X X X X X X X X Reptile Large Basilicus Iguana lizard - - - 1 jaw 1 jaw fgt. 1 jaw fgt. 1 jaw fgt. - - - Mammal Mazama Odocoileus Uniden- Homo sapiens americana virginianus tified scrap - - - 1 - - - 1 - - 3 phalange - - - - 1 - - 1 calcaneum - 1 fifth right - - metatEirsal 1 femur shaft - 1 left radius 1 molar - 1 left femur shaft - - 1 left tibia proxi- - mal fgt. 1 left astragalus 1 right humerus distal end - - 1 metatarsal - distal fgt. 1 right tibia dis? tal end - ? 1 metacarpal - - - 1 ulna - - - 1 humerus - 1 tibia Appendix 2 A N A L Y S I S O F P O T T E R Y S U R F A C E C O L L E C T I O N S F R O M T H E P A C I F I C DISTRICTS O F P U N T A B U R I C A , S A N F ? L I X , A N D R E M E D I O S I N C H I R I Q U I , P A N A M A Anthony J. Ranere This particular study reports on the 17 surface col? lections and one small test excavation made in the Punta Burica District, the 4 surface collections made in the San Felix District, and the 3 surface collections made in the Remedios District by members of the field project in 1961. All three districts lie in the Chiriqui Province, but they are not contiguous (fig. 46). In fact, Punta Burica District is separated from the other two districts by an air distance of approxi? mately 120 kUometers. The aims of this paper are: (1) to determine the distribution of pottery types in the Punta Burica, San Felix, and Remedios districts of Chiriqui; (2) to describe those types which have not appeared in the l i terature; (3) to modify those types which have been described on the basis of additional data from the collections; and (4) to establish a rough chronology for the districts under consideration. Dr. Olga Linares de Sapir has made available the collections used in this study and has supervised my work for several years. In the process, she has taught me all I know about ceramic analysis and has prodded me along when the work on this paper faltered. My debt to her is clearly a great one. I am further in? debted to Dr. Betty J . Meggers and Dr. Clifford Evans for their patience and assistance while this paper was being prepared for publication. My own ineptitude in drawing was kept from being exposed by Charles Schurcliff, who made the drawings in figures 48-55. The present contribution is a revised version of a research paper presented to the Department of Anthropology, Harvard College, in partial fulfill? ment of the requirements for the degree of Bachelor of Arts, with honors in the subject of Anthropology, in 1964. Natural Setting The Burica Peninsula, where all but one of the Punta Burica sites are located, juts southward approxi? mately 20 kilometers in the Pacific, forming the western boundary of the Gulf of Chiriqui (fig. 46). The width of the peninsula varies from 11.7 kilometers at the latitude of Puerto Armuelles to less than 5 kilometers near its seaward extremity. A ridge runs completely down the length of the landform slighdy off-center to the west, corresponding to the political border between Panama and Costa Rica. This ridge caps a mountainous and rugged terrain. Terraces can be seen throughout the eastern region, which is the only one that concerns us here. They are not continuous, ranging from 92 to 182 meters in length, but they do seem to occur at constant eleva? tions along the coast. These terraces represent elevated marine-cut benches. Bishop (1961) states that the surface rocks consist of soft gray to light green clay, a very fine-grained calcareous sandstone and siltstone, conglomerates, and thin lentils of lignite. He further classifies the soils as residual, that is, formed by weathering of underlying strata, or as resulting from deflation from adjacent areas. The terraces, with well- developed soils and proximity to water, provide good habitation areas. 107 108 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 The eastern coast of the peninsula is characterized by a wide sand beach composed principally of angular to subangular grains of quartz and a wide variety of igneous minerals including magnetite. A number of small streams which drain the eastern half of the peninsula contribute much of the materials for the formation of the beach, while marine erosion con? tributes the remainder. This section of the coast, being somewhat sheltered, is not subject to the violent surf characteristic of other Chiriqui Gulf coastal regions and therefore is more favorable for water travel and exploitation of the water resources. The San Felix and Remedios districts are contigu? ous and share the same geographic and geologic features. The shoreline of this combined region consists of a wide sand beach broken occasionally by rock outcroppings which extend to the sea. The many streams which flow through the region provide the beach sand which consists of subangular quartz grains and a number of igneous minerals. A plain covered with marine clay extends inland from the beach for a distance of eight miles, covering the extremities of the older igneous rock strata (Bishop, 1961). The plain is separated from the beach by a low escarpment. A number of small estuaries extend back into the plains, producing a rather varied coast line. The Sites The following descriptions were taken from the Site Survey Sheets of the 1961 expedition. Since I have not visited any of the sites myself, I can add nothing more to the site descriptions. For location of the sites, see figure 46. Punta Burica District BA-5: The site is located 2.5 kilometers southwest of the coast at Puerto Armuelles and 2.25 kilometers west of the coast at Corotii. It is a habitation site on the top of a level plain cut off on two sides by the Corotvi River. The elevation of the site is 70-80 meters and it covers approximately one acre. The site is located in a cleared but not cultivated field with a few scattered trees. The soil is composed of clay material with quartz grains. Three benches are developed in material adjacent to the stream valleys. There was a coverage of humus which varied in thickness from 0.7 to 2.0 meters. One group of sherds was found in a bank which had been eroded by the river. One group was found 50 cm. from the bank top, while a few were found 200 cm. from the top. Occupation of the site appeared to be scattered. BA-6 (El Rompio): A cemetery site 2.5 kilometers from El Rompio town proper and 3 kilometers from the town of Palo Blanco at a spot 2 kilometers inland. The nearest fresh-water source is the Rio Palo Blanco, 1 kilometer away. On the site are closely placed holes which may or may not have been graves, 1.5-2.0 meters long and 1.0-1.5 meters deep. They were apparently oval in shape, covered with a dome? like layer of small boulders. Site elevation is 10-20 meters; extent of the site area is given a minimum of 200-300 square meters. The ground is covered with virgin timber and moderate undergrowth and is not cultivated. The soil is extremely sandy. Potsherds were taken from the backfill of previous grave excavations. The site has been extensively potted and is in danger of total destruction because of this. BA-7: A probable habitation site located 1.5 kilo? meters west of Charco Azul proper on the eastern side of rolling mountains where the elevation is 120- 150 meters. This is a region of rolling steep-sided hills covered with grasslands and a few scattered trees. Formerly the site was a forested area, but it was cleared for a pasture about seven years ago. The land is deeply gullied. The soil is a thick, clay sort, slightly silty, and more weathered than beach ma? terial. It also contains more clay and humus than beach soil, being older geologically. Potsherds were collected from the surface and ex? cavated from a test pit dug on the deeply gullied side of a slope near the top of the hills. All sherds were found in the top 50 cm., below which was sterile soil. Occupation of the site appeared to be scattered. BA-8 (Quebrada Baules): A habitation site lo? cated less than half a kilometer west of the coast between two streams; the northern is called Quebrada de Balsa (70 meters distant), and the southern is called Quebrada Mamey (150 meters distant). This is a flatland region about 25-30 meters in elevation. The site itself is on a rise one meter above the flatland. The extent of the habitation zone is unknown, but exceeds 100 meters in length. The area was formerly planted with maize, but is now covered with platanillo and guarumo. The major soil constituent is clay- sized material with a high percentage of fine, angular quartz sand and humus. Potsherds were collected from the surface and from the hole of a previous excavation. A small test exca? vation 80 X 90 cm. was undertziken with the following results: Level 0-10 cm., 8 sherds. Level 10-25 cm., 6 sherds. Level 25-35 cm., 3 sherds. Level 3 5 ^ 5 cm., 6 sherds. Level 45-60 cm., 13 sherds. Level 60-70 cm., 7 sherds. WHOLE VOLUME APPENDIX 2 109 Level 70-80 cm., 20 sherds. Level 80-90 cm., 4 sherds. Between 85-90 cm. there is a slight change in the soU, which becomes more sandy and lighter in color due to redistribution of magnetite. Level 90-100 cm., 2 sherds. These sherds might have been intrusive. D a m p sand is first encountered at this level. Level 100-110 cm., sterile. BA-8a (Quebrada Baules): A continuation of BA-8 about 70 meters to the northwest. This places BA-8a 6 meters south of Quebrada Balsa. I t is on the side of a raised mound about 1.0-1.5 meters in height. Potsherds were taken from the backfill of two previous excavations (potholes) 2 meters apart . BA-9 : A habitation site 19 kilometers south of Puerto Armuelles on the first point north of Charco Balsa and 50-100 meters south of Quebrada Limones. I t is located on the slope of a hill immediately above the beach and to one side of a present-day dwelling. The land is planted in maize and supports a few scattered trees. T h e soil is residual, consisting of fine sandstone with a high percentage of clay. Potsherds were found on the slope of a hill planted in maize. T h e slope is 20 degrees, the grade is 30 percent, and the elevation in 40-60 meters. The site appears to be worth excavating. BA-10: A habitation site some 100 meters west of the coast and 20 meters north of Quebrada Charco Azul. T h e area of the site is flat and 50 meters in elevation; there are plateaus and hills to the north and west. Maize and rice were once planted in this area, bu t it is now a pasture for cattle. The soil is principally clay. Potsherds were found in an eroded hillside ravine in the first 65 cm. of ravine facing. Occupation of the site appeared to be scattered. B A - 1 1 : A habitation site due west of the coast from La Quebrada de Charco Azul in the Quebrada de Manzanillo. This is a deep (about 10 meters) stream channel that runs in a big " S " from the coast. The sides are densely overgrown with bijao and chi- chica plants. Soil is residually derived from highly argillaceous rock. There is also some clay shale, which is quite contorted. Potsherds are in the east-west bend of the stream in a gully that runs north-south about midway up the banks of the gully. BA-12: A habitation site 1.5 kilometers from Mel- licita and about 1 kilometer from the coast. It extends for approximately 150 meters along a branch of the Quebrada de Mellicita. Vegetation takes the form of tall grass, scattered trees, and brush. The soil is sandy, but compact. Potsherds were found in the banks on either side of the branch. The banks rise steeply for about 8 meters in this area. All sherds came from 0-40 cm. below the surface. BA-13: A habitation site approximately 100 meters west of the coast on a high plateau 50 meters from the Quebrada de Felix Sanchez. The plateau was culti? vated until a few years ago, but it is now open grass? land. The soil is a highly silty to fine sand clay, locally derived from residual weathering. Potsherds were collected from the surface of the site. BA-14: A habitation site 150-200 meters west of the coast, 75 meters north of the Corotu River, and 125-150 meters south of Quebradas Grandes. The site is on a plateau overlooking the sea. The area was once under slash-and-burn cultivation and is now cleared as far back as the river. There is some minor gullying. The soil is residually derived, silty, and clayey. Potsherds were collected from the plateau area and from the slope of the hill below. BA-15: A habitation site 100 meters south of Cana Blanca and 330 meters inland on the first plateau from the sea; area is 10 hectares. Vegetation is grassland with a few scattered trees. The soil is residually derived, sUty, and clayey. Potsherds were collected from the surface of an eroded gully about 2 meters below the crest of the plateau. BA-16: A habitation site 750 meters west of the coast and 500 meters south of Quebrada Calabazo. This region is flat, some 40-50 meters in elevation and about a hectare in area. The land had just been burned for planting at the time it was visited. Potsherds were found along the eroded edges of the plateau. San Felix District SF-1 (La Merejilda): A habitation site to the south of Las Lajas and to the east of Cerro El Jobo. The Rio San Felix is 300 meters to the east. The site is located between two quebradas. The area is now a pasture for cattle, which are driven through the erod? ing bank where the sherds are found. The soil is light colored and chalky. Potsherds were collected from the eroded bank to a depth of at least 15 cm. and for a length of more than 200 meters in an east-west direction. Potsherds were also found to a depth of 15 cm. on small raised mounds adjacent to the eroded bank. Although this is not an impressive site, the concentration of sherds may warrant eventual excavation. SF-2 (Cerro El Jobo ) : A habitation site directly south of Las Lajas, three quarters of the way to the beach on a large hill north of the Quebrada del no SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 Gato. T h e area has been burned recently for culti? vation of maize and root crops. The red clay soil is compact and ferruginous. Potsherds were found on the surface of the hill on both the east and west faces, from 25 meters from the bottom to the top of the hUl. Sherds were cJso found adhering to the roots of large trees. Occupation was relatively scattered, although concentration of sherds may justify eventual excavation. SF-3 (El Escudo): A habitation site in the town of Santa Cruz, west of the fence which separates the two houses from the property called El Escudo. At the time it was visited, the property was undergoing plowing in order to convert it into a pasture. The soil is of a compact red ferruginous clay type. A quebrada runs some 100 meters west of the site. Scattered potsherds were collected on the surface of what is today the refuse area of the town of Santa Cruz. A tractor working there, at the time of the visit, was uncovering many sherds as it plowed the ground. Sherds were collected over an area of at least 175 square meters. SF-4 (El Zapote) : A habitation site southeast of Las Lajas in the direction of the Playa del Jobo . I t is located next to the Quebrada Mamey. The land is a pasture for, cattle. The soil is black humus to a depth of at least 15 cm. Potsherds were found to a depth of 15 cm., although no sherds were visible on the surface. Remedios District R E - 1 (El Arbol Pan ) : A habitation site and possibly a cemetery located several kilometers directly west of Cerro San Cristobal, between Que? brada del Espave and Quebrada del Arbol Pan. The site is on a slight rise and extends about 250 meters northwest-southeast by 50 meters in the opposite direction. The region has been plowed in order to plant pangola grass. The soil is a rich dark brown clay, which is quite compact. Potsherds were found on both raised and low areas, as well as in the hole of a previous excavation. The concentration of sherds warrants eventual excavation. R E - 2 : A habitation site approximately 2 kilo? meters east of the old bridge at the Salado River along the old highway to Panama City. The site is 100 meters south of the road. The area is cleared with some scattered brush. The soil is a red ferruginous clay. Potsherds were found at specific spots, not scattered in the usual manner . Sherds were excavated to a depth of 10 cm. in a spot which might have been a house location. The soil in the excavation contained a good deal of carbon. R E - 3 (Potrero Grajales): A habitat ion site south? east of the town Remedios and 300 meters east of the Quebrada del Serracin on the side of Cerro la Gari ta , the largest hill in the vicinity. T h e land is presently cleared for pasture. T h e soil is a dark brown compact clay. Potsherds were scattered in small numbers over an area 100 meters square. A number of sherds were found from 0-20 cm. below the surface where the cattie had possibly trampled them. O n the Cerro la Gari ta was the 16th-century town of Remedios which was burned by pirates. Pottery Type Descript ions Of the pottery types encountered in the surface col? lections from the Punta Burica, San F^lix and Reme? dios Districts, all but two are described by Dr . Linares de Sapir in the main body of this volume. T h e descrip? tions of the two new types encountered in these collec? tions are recorded. Balsa Polished PASTE : Method of manufacture: Coiling. Coiling junctions are invisible on the surface, but in a very few instances can be seen in cross section. Temper: Small rounded red (hematite) and black (mag? netite?) particles. These inclusions are distributed homo? geneously throughout the paste and are characteris? tically less than 1 mm. in diameter, although the hematite grains reach a diameter of 4 mm. on rare occasions. Texture: Uniform, fine, and compact. Breaks along straight lines. Color: OfF-white to buff-orange and more rarely, light gray. Occasionally there is an oxidation band of light orange extending inward 1-3 mm. from either side or just from the outside of the sherd. Less frequently a gray core 2-3 mm. thick is found. Some fire clouding is noticeable, but not to any appreciable amount. SURFACE: Color: Dirty white to buff, with some gray areas due to fire clouding. Treatment: The exterior is characteristically very highly polished. The interior is sometimes lighdy polished and sometimes smoothed, both being equal in incidence. Hardness: Fairly hard; 3.5-4. FORM: Body wall thickness: 5-13 mm. Rim and Lips: Group 1 (fig. 48a): Slightly outflaring rim which is an extension of the body wall. Diameter of the rim is 17 cm. Group 2 (fig. 486): Everted rim slightly thickened on the inside, having an angular profile. Rim diameter is 14 cm. W H O L E VOLUME APPENDIX 2 111 I 1 I I f 0 I 2 3 4 CM FIGURE 48.?Rim profiles of Balsa Polished, groups 1-3, Burica Peninsula. Group 3 (fig. 48c): Extension of the body wall. The r im has a grooved lip like that of a sugar bowl. Stria? tions in the groove indicate that the vessel probably had a top. R i m diameter is 22 cm. Bases and Supports: None in the sherd sample. Appendages: Flat strap handles, about the thickness of the body wall and relatively small. Reconstruction: Open bowls are the only form clearly identifiable from the sample. DECORATION: A uniform red strip 1-2 mm. wide is painted along the lip of the rims and extended completely around the vessel mouth. T h e rest of the exterior is polished, but not painted. T h e color of the paint is a rich red. In one instance there is evidence for some sort of ap? plique decoration. This takes the form of a broken circular knob 6 mm. in diameter, extending outward from the body wall 2.5 cm. below the lip. CHRONOLOGICAL POSITION : This type occurs in association with both Isla Palenque Maroon Slipped and Ca? strellon Red Slipped Wares, the first diagnostic of the Burica Phase, and the second diagnostic of the following San Lorenzo Phase in the central Chiriqui Gulf sequence. GEOGRAPHICAL DISTRIBUTION: Balsa Polished is found only in collections from sites BA-8, BA-8a, and BA-13, all located near the tip of the Burica Peninsula. COMPARATIVE MATERIAL: This type may have some affinities with the Red Lined Ware of Holmes (1888, pp. 109-111) and M a c C u r d y (1911, pp. 92-95), but no direct connection can be ascertained. Balsa Polished was not described by Linares de Sapir in this volume, but a number of Red Lined Wares were. Balsa Polished may simply be another one of the Red Lined Wares. Q u e b r a d a Baules Brushed PASTE : Method of manufacture: Probably coiling, although coil junct ions are completely obliterated. Temper: Small angular quar tz particles with minor inclusions of rounded, red particles (hematite) and other minerals. Texture: Coarse and somewhat friable. Air pockets are frequent. Fractures irregularly. Color: Buff to reddish-brown. Sometimes there is a gray to black zone extending in from the interior surface, caused by incomplete oxidation on firing. SURFACE: Color: Dirty buff to dirty brown. Occasional gray to black areas due to fire clouding. Treatment: Traces of coiling are removed, and the exterior body wall is roughly scraped. T h e surface appears to have been roughed with either cording, or a stiff brush, or a scraping tool. Rough parallel stria? tions, generally 2-4 mm. apart , are visible on a number of sherds. T h e r im and neck of these vessels are not roughened, but instead are lightly polished. T h e surface of the body sherds is differentially eroded, giving a heavily pitted effect. T h e interior surface is smoothed but not polished. Hardness: Very soft; 2. F O R M : Body wall thickness: 5-15 mm. Rims and Lips: Group 1 (fig. 4:9a-f): Everted rims apparently be? longing to large globular jars with short necks. Rims are all exteriorally thickened by addition of coils, whose junctions are sometimes not entirely obliter? ated. T h e contour of the lip is angular, tending toward being rectangular. T h e angle formed with the neck ranges from 60 to 90 degrees. R i m diameters vary from 29 to 45 cm. Group 2 (fig. 49^): Vertical to slightly everted r im thickened by what appears to be a single coil added W ^^ Wf W I I I I I 0 I 2 3 4 CM FIGURE 49.?Rim profiles of pottery type Quebrada Baules Brushed, groups 1 and 2, Burica Peninsula. 112 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 to the exterior. The lip comes to a point, forming three sides of a diamond from 0 to 20 degrees off of vertical. Bases and Supports: None in the sample. Reconstruction: Large globular jars with short necks and large openings are the only type represented. DECORATION : See surface treatment. CHRONOLOGICAL POSITION: This type is associated with Isla Palenque Maroon Slipped, a diagnostic type of the Burica Phase in the central Chiriqui Gulf sequence. It is also associated with Castrellon Red Slipped, which is diagnostic of the San Lorenzo Phase in the same Gulf sequence. GEOGRAPHICAL DISTRIBUTION: Quebrada Baules Brushed is found only in the BA-8, BA-8a, and BA-13 sites on the Burica Peninsula. These sites are near the southern tip of the peninsula. COMPARATIVE MATERIAL: None. G e o g r a p h i c a l D i s t r ibu t ion of Po t te ry T y p e s and M o d e s Punta Burica District The oldest materials among the collections in? cluded in the analysis come from the district of Punta Burica. At site BA-9, there were evidences for the occurrence of La Concepcion complex and/or the Aguas Buenas complex. A three-toed tripod foot (fig. 50) in the collection from this site is duplicated on a vessel of the Scarified Ware group in the Pea? body Museum, Harvard University. Moreover, the same three-toed tripod foot on Scarified Ware vessels is pictured by Holmes (1888, fig. 118) and by Mac? Curdy (1911, pi. X X V I , f). Since this tripod foot ? J 0 I 2 CM FIGURE 50.?Three-Toed Tripod Foot belonging to Scarified Ware of La Concepcion complex from Site BA-9 in the Burica Peninsula. does not occur in any of the other pottery types found in Chiriqui or adjacent areas, the find, even though only a single sherd, can be taken as incontro? vertible evidence of the presence of La Concepcion complex in the BA-9 surface collection. I t is interest? ing to note that Linares de Sapir found no trace of this complex in her collections. The same site (BA-9) also produced an applique figure in the form of a bird decorating one of the r im sherds (fig. 51). This figure closely resembles one pictured by Haber land (1955, p . 227, fig. 5) and assigned by him to the Aguas Buenas complex. Although applique figures are common in a number of Chiriqui pottery types and in types found in adjacent areas, I have not seen one similar to this particular applique bird in any of the literature. Exacdy what the occurrence of these two sherds means in terms of the chronological sequence is difficult to ascertain. The predominant pottery type in the site collection is Isla Palenque Maroon Slipped, characteristic of the Burica Phase. T h e alternative interpretations we have to choose from in order to explain the presence of the three-toed foot and applique bird figure are as follows: (1) O n e may dismiss the evidence for both La Concepcion and Aguas Buenas complexes as being too tenuous and assign the site to the Burica Phase. (2) O n e may admit the evidence and label BA-9 a two- or three- phase site, depending on the relationship between La Concepcion, Aguas Buenas, and the Burica Phase (Is Aguas Buenas a contemporary of La Concepcion complex or of the Burica Phase, or does it fall within the two?). (3) One may dismiss the evidence for the Aguas Buenas complex, move u p the date for La Concepcion complex, and label BA-9 a two-compo? nent site. (4) O n e may admit the evidence and either lower the initial date given for the Burica Phase, or raise the terminal date for La Concepcion and/or Aguas Buenas complex, making them overlap or at least appear shordy after one another. I am inclined to accept the fact tha t the three-toed foot belongs in La Concepcion complex and, with some reservations, I am inclined also to accept the fact that the applique figure belongs to the Aguas Buenas complex. Fur thermore, I suspect tha t the two complexes are largely contemporaneous and have a terminal date around A.D. 300. All three pottery types found at BA-9 have their origins nor th of the site location; La Concepcion is associated with the lowlands and foothills of the central Chiriqui Province (Haberland, 1962), Aguas Buenas is associated with the highland area on the Panami-Cos ta Rica border (Haberland, 1955), and Isla Palenque Maroon Slipped may have had its prototype somewhere in W H O L E VOLUME APPENDIX 2 113 J 0 I 2 CM FIGURE 51.?Applique figurine from Site BA-9, Burica Peninsula, resembling figurine from Aguas Buenas complex in Costa Rica. the Diquis area, though this inference is highly tentative at present. If we postulate a general south? ward or southeastward migration of people (or diffusion of pottery types), then, on purely distribu? tional grounds. La Concepcion complex would arrive on the Burica Peninsula before or at the same time as the Aguas Buenas complex. The Isla Palenque Maroon Slipped prototype, being farther away, would arrive later. Excavation is needed to confirm or deny any of these alternatives. The major pottery type at BA-9 is Isla Palenque Maroon Slipped, the principal trait of the Burica Phase. Starting with BA-14 and working northward up the east coast of the Burica Peninsula to the location of BA-5 (fig. 46), there are eight other sites in which the dominant or sole type is Isla Palenque Maroon Slipped: BA-14, B A - 1 6 , BA-15, BA-12, BA-11, BA-10, and BA-7. Furthermore, this type occurs at BA-8 and BA-13, although not as the main type. The presence of a single pottery type would seem to indicate a concentrated population on the east coast of the peninsula during the Burica Phase. At BA-8, BA-8a, and BA-13, pottery types occur which place part of their occupation in the San Lorenzo Phase. Castrellon Red Slipped, diagnostic of the San Lorenzo Phase, occurs in the BA-8 and BA-8a collections. Also represented at BA-8 are two additional San Lorenzo Phase types, three sherds belonging to Arayo Polished Line, and four sherds belonging to the general category of Red Line Ware (Banco Red Line?). All three sites also contain two pottery types not previously described in the litera? ture, Balsa Polished and Quebrada Baiiles Brushed. The types found at BA-8 and BA-8a (which is 200 meters from BA-8) are basically the same, with the absence of Isla Palenque Maroon Slipped at BA-8a being the exception. This might be explained by the hypothesis that BA-8 was the original habitation area and that sometime after the advent of the San Lorenzo Phase, this area was expanded, or reoccupied, to include the BA-8a area. O n the other hand, absence of this type could simply be the result of an inadequate sample. There are five trade sherds in the BA-8 collection, one of which has been tentatively identified as Code , though its fragmentary condition makes it impossible to assign it to Early or Late Cocl^ (fig. 52a). The only other identifiable sherd closely resembles pottery from the Nicoya area in Costa Rica. Lothrop (1963, p . 89) illustrates several trade sherds from Nicoya that employ bar and circle motifs similar to the sherd from Punta Burica (fig. 52b). Lothrop dates these trade sherds in the Middle Polychrome period, which Baudez and Coe have assigned to the period between A.D. 750-1100. This date agrees with the Burica and San Lorenzo Phase dates assigned by Linares de Sapir and also to the date of the Code Polychrome (fig. 47). BA-13 has been classified as a two-phase site, even though it contains no sherds considered diagnostic of the San Lorenzo Phase by Linares de Sapir. The Burica Phase is represented here by Isla Palenque Maroon Slipped sherds. The sample is small, only eight sherds, and it does contain Balsa Polished and Quebrada Baules Brushed, both of which are associ? ated with diagnostic San Lorenzo Phase pottery types at sites BA-8 and BA-8a. The decrease in the number of sites on the Burica Peninsula, from 13 in the Burica Phase to 3 in the San Lorenzo Phase, suggests a general abandonment of the coastal area at the end of the Burica Phase. The reasons for this abandonment are not known, but it may have been due to renewed pressures from outside areas, represented by new types of pottery. From the archeological evidence, it would seem that abandonment of the Burica Peninsula was all but completed in the Chiriqui Phase. Only two sherds characteristic of this phase are found in the collections from the peninsula, one at BA-5 and the other at BA-10, both of which are on the northern half of the peninsula. The two sherds belonged to the readily identifiable Tarrago Bisquit Ware. The Spanish chroniclers (Markham, 1865, pp. 24? 5) present a conflicting opinion. According to their accounts, the Burica Peninsula was well peopled at the time of the Spanish arrival. Perhaps the Indians 284-757 0-68?9 114 SMITHSONL\N CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 0 I 2 CM FIGURE 52.?Polychrome trade pottery at Site BA-8. a, Code (early or late?), b, Middle Polychrome period, Greater Nicoya area, Costa Rica. of the Chiriqui Phase moved away from the coastal areas of the peninsula and back into the hilly areas which were not explored by the 1961 archeological survey. Or perhaps the survey simply failed to locate the sites in use at the time of the Chiriqui Phase. Site BA-6 is the only single Chiriqui Phase site in the Punta Burica district. It differs from the other sites by not being a habitation site, but rather a cemetery, and by not being located on the Burica Peninsula proper, but to the east of it (fig. 46). The sherds collected from this site came from the backfill of pothunters who have all but destroyed the site. The pottery types found at BA-6 are all diagnostic of the Chiriqui Phase. The dominant type was Tar ? rago Bisquit Ware ; 10 globular vessels of this type were partially reconstructed. Villalba Red Streaked was also present; the sherds appeared to belong to shallow bowls with tripod feet. Two of these tripod feet were included in the sam,ple. In addition, the site yielded sherds from a small round-bottomed globular jar, which belong in the Negative Painted Ware category. Four sherds of Chocolate Ware, frag? ments of a shallow bowl decorated with incision, complete the sample from BA-6. San Felix and Remedios Districts We can now skip across the central Chiriqui Gulf area into the San Felix and Remedios districts, approximately 120 kilometers to the east (fig. 46). These two districts are geographically contiguous and share a number of pottery types and modes. The Burica Phase is not represented, thus the first evidence of occupation belongs to the San Lorenzo Phase. R E - 3 is the only site that might be assigned to the San Lorenzo Phase as a single phase site, but evidence supporting this assignment is tenuous. The only decorated type is Caco Red Slipped, which is char? acteristic of the San Lorenzo Phase. R E - 3 produced, however, a round handle which proved to be re? stricted to the Chiriqui Phase in the central Chiriqui Gulf area. Three strap handles, common to both phases, also occurred at this site. Sites S F - 1 , SF-2 , and SF-3 have been placed in both the San Lorenzo and Chiriqui Phases. T h e sample from SF-4 was insufficient to assign it to any phase, but it probably belongs here also. All three sites contain Caco Red Slipped, diagnostic of the San Lorenzo Phase. In addition, Congrejal Red Line occurs at sites SF-1 and SF-2 , while Castrellon Red Slipped occurs at S F - 1 . Both types belong principally to the San Lorenzo Phase. Tar rago Bisquit Ware, diagnostic of the Chiriqui Phase, is found at all three sites, while the SF-1 collection contains a double handle (fig. 53), diagnostic of the same phase. There is only one site, R E - 1 , which may have had an occupation restricted to the Chiriqui Phase. The site contained the very diagnostic trait from the Chiriqui Phase, the double handle. T h e surface sample contained only one unidentifiable painted sherd; the rest were plain sherds, characteristically thick. A collection taken from the back dirt of a pothole in the site area, however, yielded two sherds belonging to Congrejal Red Line, a San Lorenzo Phase time marker. This may indicate that a San Lorenzo occupation underlies the Chiriqui Phase occupation represented by the surface collections. Slash punctation and anthropomorphic serpent designs were the only plastic decorations present. The following modes occurred in the two districts of San Felix and Remedios (table 4), which cannot be assigned to any one phase: Applique r idging; applique ridge notching; applique raised knobs; fine-line incis- 2 CM FIGURE 53.?Double Handle Mode from Site SF-1, San Felix district; a Chiriqui Phase marker. WHOLE VOLUME APPENDIX 2 115 FIGURE 54.?Decorative Modes from the districts of San Felix and Remedios. a-b. Applique ridging, c-d. Applique ridge notching. e-f. Applique raised knobs, g-h. Fine line incising, i-k, Broad line incising, l-m. Slash punctation. n-o. Shell edge stamping. ing; broadline incising; slash punctations; and shell- edge stamping (fig. 54). T h e Cultural Sequence Summary The cultural sequence established and defined by Linares de Sapir for the central Chiriqui Gulf area can be applied to the districts of Punta Burica, San Felix, and Remedios with considerable success. There are regional variations in the definition of archeological phases, however, which make it necessary to broaden the descriptive content of these temporal units to make them apply accurately to the entire Chiriqui Gulf area. The initial Burica Phase is the one most easily defined in the cultural sequence. In order that the description of this phase encompass the Punta Burica district, it needs to be expanded in the following instances. The tripod foot (Mode al) does not occur in the Pun ta Burica district, but a different one does (fig. 55a) and should be added to the description. The ringstand (fig. 55^) found in our collections is like Ringstand Mode a in the Gulf of Chiriqui collec? tions (p. 50). In addition, seven strap handle fragments and one round handle are definitely classified as Isla Palenque Maroon Slipped in the Punta Burica collec? tions (fig. 55 a-c). The occurrence of these handles is rare and should be included in the phase description (and pottery type description) as such. Incision, a plastic decoration which occurs in the central Chiriqui Gulf area as well as on the Burica Peninsula, should also be included. (In this particular case, where a single pottery type is diagnostic of a phase, any changes in phase description are simply a matter of change in the pottery type description of Isla Palenque Maroon Slipped. This will not be the case, however, in the succeeding two phases.) The San Lorenzo Phase is a bit more complicated to modify. First, it will probably be necessary to include Balsa Polished and Quebrada Baules Brushed as diagnostic of this phase, although additional re? search is needed to substantiate this. It is difficult to separate the San Lorenzo Phase from the Chiriqui Phase in the San Felix and Remedios districts on the basis of surface collections. It appears, however, that the San Lorenzo Phase in these districts had round handles which do not occur at this time period in the central Gulf area. On the other hand, my collections lacked tripod feet, pedestal bases, and a number of pottery types considered diagnostic of this phase in that area. The following decoration modes (fig. 54a-o) should be noted as occurring (although they cannot be considered diagnostic, since they seem to last into the Chiriqui Phase): Applique ridging; 116 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY VOLUME 8 c ^ FIGURE 55.?Appendage Modes in sites of the Burica Peninsula associated with Isla Palenque Maroon Slipped, Burica Phase. a-c, Looped handles, d. Tripod foot, e, Ringstand. applique ridge notching, applique raised knobs; fine- line incising; broad-line incising; slash punctat ion; scoring; and shell-edge stamping. Shell-edge stamp? ing probably stands the best chance of eventually being proven diagnostic of this phase. T h e Chiriqui Phase description does not apply completely to the San Felix and Remedios districts. The two diagnostic types, Villalba Red Streaked and Cavada Applique and Red Banded, are absent from the two districts, while Remedios even lacks Tarrago Bisquit Ware. Also absent from these districts during the Chiriqui Phase are pedestal bases and tripod feet, both of which are found in the central Gulf area. The principal diagnostic trait for the Chiriqui Phase in the San Felix and Remedios districts is the double handle. Strap handles and horizontal handles occur in this phase, but cannot be considered diagnostic. Conjectures Although it is difficult to speak conclusively about chronological sequence relying primarily on data from surface collections, particularly those collections which are not the product of r andom sampling, the da ta from San F d i x and Remedios bring u p an interesting point. T h e occurrence of pottery types of the San Lorenzo and Chiriqui Phases always in association (in other words, the lack of single occupation sites of either phase), particularly in the San Felix district, suggests that the division made by Linares de Sapir between these two phases is even less marked here than in the Gulf of Chiriqui. Excavation in these districts, or even more surface collections for analysis, could provide the answers to the question as to whether or not the San Lorenzo and the Chiriqui Phases are chronologically distinct as in the central Chiriqui Gulf area, or whether they are in the San Felix and Remedios districts largely contemporaneous. I t may be the case that only in the San Felix-Remedios area are the two phases synchronous, and that they are somewhat distinct in the central Chiriqui Gulf area. At the beginning of the San Lorenzo Phase, the island sites in the Gulf were inhabited by people coming in from the Chiriqui mainland, probably replacing the people of the Burica Phase, or at least replacing their cultural assemblage. These new in? habitants introduced pottery types decorated with red bands and red lines, which spread in both directions, for we find Castrellon Red Slipped and Arayo Polished Line pottery types on the Burica Peninsula and Caco Red Slipped, Castrellon Red Slipped, and Cangrejal Red Line in San Felix and Remedios. Then, sometime after A.D. 1000, another group of people seem to have moved out into the coastal areas of the central Chiriqui Gulf and onto the islands; or what is more likely, there was a diffusion of two pottery types in this direction, namely Tar rago Bisquit Ware and Villalba Red Streaked. T h e three districts examined in this report were actually peripheral to this pattern of communication between the mainland Chiriqui and the island people. T h e Tar rago Bisquit Ware, being a popular type, diffused across the dis? trict of Alanje to the northern par t of Punta Burica. I t also diffused eastward to San Felix, bu t not as far as Remedios. Villalba Red Streaked did not even diffuse as far as San Felix. At this time, both San Felix and Remedios were apparently west of the main communication channel. At the same time as Chiriqui Phase pottery types were diffusing into the central Chiriqui Gulf area, another cluster of traits was moving eastward into the Chiriqui Gulf area. This diffusion is perhaps most easily traced by use of the double handle. There is no reference to double handles in either Holmes (1888) or MacCurdy (1911). Double handles are pictured by Lothrop (1950, p . 20, fig. 11) and assigned to the provenience of Bubi, which is in the WHOLE VOLUME APPENDIX 2 117 southwest coastal area of Veraguas, not too far from Remedios. Linares de Sapir (p. 55) notes a double handle exactly like the ones found in the central Chiriqui Gulf area (which are like the ones found in Remedios) in a collection made by M. W. Stirling in 1951 from a locality 25 miles northwest of Panama City. The direction of the diffusion was westward from this point to the islands of the Chiriqui Gulf. Double handles are not known from the Burica Peninsula. Another trait found in the San Felix and Remedios districts, which seldom occurs in the districts to the west, is plastic decoration. Such decoration had a wide distribution in Central America to be sure, but on Veraguas exists almost to the exclusion of other decorative techniques. Lothrop (1950) pictures a number of plastic decorative modes that are similar and sometimes identical to those occurring in the two districts of San Felix and Remedios. Ladd (1964, plate 14) illustrates plastic decorations from the Parita Bay area, some identical to those found in these two districts (slash punctation, applique ridge notching, fine-line incising, lip grooving, shell-edge stamping, scoring, broad-line incising, and applique ridging). The San Felix and Remedios districts did not form, during the Chiriqui Phase, an integral part of the "Chiriqui Classical Culture" sphere. They were in a transitional zone between the two well-defined late styles of Chiriqui and Veraguas, sharing character? istics with both. This variation in local styles accords well with the large number of Indian groups (provincias) found by the Spaniards within what is today the Province of Chiriqui. The area around San Felix and Remedios was probably occupied by the Guaymi, as opposed to the "Classical Chiriqui" area of the Doraces, Zuries, etc. (Linares de Sapir, p. 79). As a final summary, it may be said that the analysis of the surface collections from Punta Burica, San Felix, and Remedios has served to expand and clarify the phases of the Central Gulf sequence and has also contributed important information on the limits and distributions of each phase. Interesting suggestions have been brought forth by the appearance of some traits that were absent from the Central Gulf area and by the lack of well-documented single-phase sites for either of the two last phases. What is needed now are substantial excavations placed at selected sites to secure the chronology for these districts and to tie them in with neighboring regions. The first step has been taken in this analysis. Pottery types Isla Palenque Maroon Slipped, thin variety Isla Palenque Maroon Slipped, thick variety Quebrada Batiles Brushed Balsa Polished Castrellon Red Slipped Tarrago Bisquit Ware Villalba Red Streaked Negative Painted Ware Chocolate Incised Scarified Ware Polychrome Trade Ware Decorated, unclassified Plain, unclassified T O T A L TABLE Total 78 434 115 50 15 32 28 2 4 2 5 21 40 826 1.?Distribution of pottery types 16 6 33 - - - 3 42 15 14 1 5 1 33 - - - 1 1 3 39 13 - 3 2 3 - - 8 12 - 36 - - 8 5 49 in the 11 2 20 - - - - 22 distric of Punta Burica BA?Sites 10 1 18 - - 1 - - 20 9 26 117 - - 2 5 3 152 8a ' - - 9 2 3 - 3 17 8a - - 34 6 2 - 3 45 8' - 7 19 16 5 - 4 52 8 3 22 51 23 5 8 8 125 7a - 19 - 1 20 7' 9 42 - 5 56 7 14 52 - 2 68 6 - - 30 28 2 4 - - 64 5 11 31 1 - - - 1 44 118 S M I T H S O N I A N C O N T R I B U T I O N S TO A N T H R O P O L O G Y TABLE 2.?Distribution of pottery types in the districts of San Felix and Remedios V O L U M E 8 Pottery types Castrellon Red Slipped Congrejal Red Line Caco Red Slipped (thin) Caco Red Slipped (thick) Tarrago Bisquit Ware Villalba Red Streaked Decorated, unclassified Plain Ware, unclassified Plain Ware, thick Polychrome Trade Ware TOTAL T O T A L 2 10 8 508 53 64 26 1,695 251 6 2 ,623 R E - 3 - - 192 - - - 375 13 1 581 R E - 2 - - - - - - 121 - - 121 R E - 1 2 - - - - 3 58 35 - 98 Habitation sites a R E - 1 - - - - - - 57 170 - 227 S F ^ - - - - - - 20 - - 20 SF-3 - - 21 11 - 6 402 14 - 454 SF-2 2 - 289 21 - 5 359 7 2 685 SF-1 2 6 8 6 21 64 12 303 12 3 437 TABLE 3.?Distribution of appendage modes (handles, supports) in the districts of Punta Burica, San Felix, and Remedios Appendage Modes Handles Double Horizontal Strap Round Bases and Supports Flat, thickened Ringstand Pedestal Looped Feet Tripod Feet Appendages, misc. T O T A L T O T A L 145 4 2 115 24 22 2 14 1? 2? 3 3 170 R E - 3 4 _ - 3 1 - _ - - - - - 4 R E - l a 3 _ - 2 1 - _ - - - - - 3 R E - 1 45 3 2 32 8 4 _ 2 - 2? - - 49 SF-3 27 _ - 21 6 9 _ 8 1? - - - 36 SF-2 18 _ - 17 1 3 1 2 - - - - 21 Sites SF-1 38 1 - 31 6 - _ - - - - - 38 BA-14 - _ - - - 1 _ 1 - - - 1 2 BA-12 - _ - - - 1 _ - - - 1 - 1 BA-11 2 _ - 2 - - _ - - - - - 2 BA-9 7 _ - 7 - 2 _ 1 - _ 1 2 11 BA-7a 1 _ - - 1 1 1 - _ _ - - 2 BA-7 - _ - - 1 - _ _ 1 - 1 WHOLE VOLUME APPENDIX 2 119 Motifs TABLE 4.?Distribution of plastic decorative motifs in the districts of Punta Burica, San Felix, and Remedios Applique Ridging Applique Ridge Notching Applique Ridge Scalloping Applique Raised Knobs Anthropomorphic Applique Fine-line Incising Broad-line Incising Engraving Slash Punctation Scoring Circular Depressions Shell-Edge Stamping T O T A L T O T A L 21 10 10 18 6 7 3 1 2 96 Sites R E - R E - R E - S F - S F - S F - S F - BA- BA- BA- BA- BA- BA- 3 2 1 4 3 2 1 14 12 11 9 7 5 10 1 5 1 6 13 1 3 15 1 4 _ 1 1 - 51 10 TABLE 5. ?Number of sherds in each pottery type of the test excavation at site BA-8 in the Punta Burica district Pottery types Level (cm.) Isla Palenque Maroon Slipped Balsa Polished Castrellon Red Slipped Red Lined Quebrada Baules Brushed Incised Plain, unclassified TOTAL TOTAL 14 22 13 1 13 1 6 70 90-100 1 - - - - 1 2 80-90 1 1 2 - - - 1 5 70-80 4 6 1 - 8 - 1 20 60-70 2 2 1 - 2 - - 7 45-60 3 4 2 - 2 - 2 13 3 5 ^ 5 1 2 2 - 1 - - 6 25-35 2 - - - - 1 3 10-25 1 3 1 - 1 - 6 0-10 3 3 2 - - - - 8 PLATE 1 Typical estuaries south of David, showing water level at low tide, a. Note absence of turbulence and mangrove vegetation on mud embankments, b. Temporary encampment of fishermen marked by coconut grove on the background; hidden from view by tall secondary growth. 284-757 0-68?10 m^ PLATE 2 Typical mangrove vegetation on western shores of the Estero de Horconcitos. a, Closeup of areal roots; note area to left cut down by coastal inhabitants for firewood, b, Bank eroded by action of tides. '-^ ~ -^S^- I I I I I I 0 I 2 3 4 5 CM ^'^tM'-^*-^^ PLATE 11 Type sherds diagnostic of San Lorenzo and Chiriqui Phases, a-e, Horconcitos Red Banded, diagnostic o San Lorenzo Phase. f-j, Cavada Applique and Red Banded, diagnostic of Chiriqui Phase. J I L 0 2 3 4 5 CM :^- .^^T PLATE 12 Type sherds of Linarte Zoned Red Line, San Lorenzo Phase. '^^ ..^ "? % m I j _ J I I 0 1 2 3 4 5 CM PLATE 13 Type sherds diagnostic of the Chiriqui Phase, a-i, Tarrago Bisquit Ware, j-m, Villalba Red Streaked. PLATE 14 Trade pottery from other Panamanian regions and sherds of Classical Chiriqui pottei-y. a. Possibly Venado Beach, Canal Zone, b-g, Code polychromes, probably Late Code, h-j, Macaracas type, Higo variety, k. Alligator Ware, Classical Chiriqui Culture. /, Negative, Black-on-Red, Classical Chiriqui Culture. I 1 I I I I I 0 I 2 3 4 5 CM PLATE 15 Tripod Feet Modes of San Lorenzo and Chiriqui Phases, a-e, San Lorenzo Phase, a. Mode cl . b. Mode a5. c. Mode a2. d, Mode b2. e. Mode b3. / , g, Chiriqui Phase. / , Mode e3. g. Mode h. 3 CM PLATE 16 Local and Trade Tripod Feet Modes in Chiriqui Phase, a. Mode h. b, Mode 11. c, Mode f3, probably trade from Diquis, Costa Rica, d. Mode ii , belongs to Chocolate Incised Ware, trade pottery from Diquis, Costa Rica (?). ^m?'' I I I I 1 J 0 5 CM a 1 1 ? 1 ; 0 4CM b-h ? h PLATE 17 Strapped Feet and Pedestal Base Modes and Ringstands, diagnostic of San Lorenzo and Chiriqui Phases, a, Strapped Feet, Mode a San Lorenzo Phase, b, Pedestal Base, Mode f, Chiriqui Phase, c. Pedestal Base, Mode a, San Lorenzo and Chiriqui Phases, d Pedestal Base, Mode i, San Lorenzo Phase, e. Pedestal Base, Mode g, San Lorenzo Phase, top view, f-h, Ringstand, Mode b two interior and one exterior views, San Lorenzo and Chiriqui Phases. L I I I I ? 0 I 2 3 4 5 CM PLATE 18 Handle Modes of the San Lorenzo and Chiriqui Phases, a-c, Strap Handles, d-e, Ladle Handles, f-h, Round Handles, i, Horizon? tal Handle, j-l, Double Handles. i^^'Z- P'''"^ *' '^A ?'/A7?^' MM 0 "^" '^5^ I L__l I I I 0 5CM 0 5 CM a-c, e-f, h-q PLATE 19 Stone artifacts of San Lorenzo and Chiriqui Phases, a, Pounding-anvil stone, b-l, Celts, m-q, Line or net weights. "n XS-7 ^. XS-'f .1 U-ST V A ^ W.ii^ ??J^ -f") m /S- / / ' S' ^ I ' I I 4 CM ' ' I ' ? ? 0 5CM j - q PLATE 20 0 5 CM s-u I I I I I I 0 5CM r Stone artifacts of San Lorenzo and Chiriqui Phases, a-e. Projectile points. / , Flake, g-i. Pebble polishers, j , Pecking stone, k-q. Sandstone saws, r, Metate leg. s. Saddle-shaped metate. t-u, Manos. U.S. GOVERNMENT PRINTING OFFICE 1968 0 ? 284-757