SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY ? NUMBER 2 0 The Walakpa Site, Alaska ITS PLACE IN THE BIRNIRK AND THULE CULTURES Dennis J. Stanford SMITHSONIAN INSTITUTION PRESS CITY OF WASHINGTON 1976 A B S T R A C T Stanford, Dennis J. T h e Walakpa Site, Alaska: Its Place in the Birnirk and T h u l e Cultures. Smithsonian Contributions to Anthropology, number 20, 226 pages, 29 figures, 119 plates, 10 tables, 1976.?An archeological survey directed by the author near Point Barrow, Alaska, during the summer of 1968 resulted in the discovery of Walakpa, a deeply stratified coastal Eskimo site. It contained over 20 occupation levels, showing the development of Eskimo culture from Birnirk to Thule , as well as earlier and later Eskimo occupation levels. On the basis of excavations at Walakpa during 1968 and 1969, previous estimates of Birnirk and Thu le origins are reexamined and a new i n t e r p r e t tion of the genesis of this Eskimo culture proposed. Specifically, on the basis of the Point Barrow excavations, this monograph (1) examines the development of the Birnirk and Thu le Eskimo cultures and (2) provides definitions for the horizon markers for each of the Eskimo stages repre? sented at Walakpa. As a result of this detailed study, the author concludes that (1) Birnirk developed out of Old Bering Sea; (2) Birnirk can be divided into three phases: Early, A.D. 500-700, Middle, A.D. 700-800, and Late, A.D. 800-900; (3) Thule Eskimo culture developed directly out of Birnirk; and (4) the development from Birnirk to Thu le took place because of over-utilization of seals as the primary food resource and a change to a warmer climatic regime that further depleated the already weak seal resource, resulting in an increased use of whales for food and an expansion of hunt ing territories to the east. OFFICIAL PUBLICATION DATE is handstamped in a limited number of initial copies and is recorded in the Institution's annual report, Smithsonian Year. Library of Congress Cataloging in Publication Data Stanford, Dennis J. The Walakpa site, Alaska. (Smithsonian contributions to anthropology ; no. 20) Bibliography: p. 1. Walakpa site, Alaska. I. Title. II. Series. GN1.S54 no. 20 E99.E7 979.8'7 74-32434 For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 - Price $4.30 (paper cover) Stock Number 047-000-00330-0 Contents Page PREFACE xi INTRODUCTION 1 Origins of Thu le Eskimo Culture 1 Geophysical Conditions 3 Geology 3 Physiography and Ecology 3 Low T u n d r a Ecosystem 4 Marine Ecosystem 5 Pond and River Ecosystems 7 Prehistoric Climates and Climatic Change 7 Modern Climatic Conditions 8 SITE DESCRIPTION 10 Stratigraphy 10 A Area 12 B Area _ 13 Archeological Excavations of the Walakpa Site 14 1968 Excavation 14 1969 Excavation 15 Additional Excavations in the Walakpa Bay Area 16 ANALYSIS AND CLASSIFICATION OF ARTIFACTS 18 Hunt ing and Fishing Equipment 18 Harpoon Gear 18 Harpoon Heads 18 Harpoon End Blades 23 Harpoon Foreshafts 24 Harpoon Shafts 25 Harpoon Ice Picks 25 Seal Poke Gear 26 Float Bars 26 Seal Float Mouthpieces 26 Seal Float Mouthpiece Stoppers 26 Seal Float Plugs 27 Detachable Lance Points 27 Ice Hunt ing Gear 27 Wound Plugs 27 Wound Pins 27 Ice Scoops 27 Two-Piece Ice Scoops 28 Seal Drag Handles 28 Seal Scratchers 28 Seal Rattles 29 Nets and Net Weights 29 Atlatl Equipment 29 Atlatl Dart Boards 29 Atlatl Hooks 29 Dart Shafts and Butt Pieces 29 Harpoon Socket Pieces 30 iii IV S M I T H S O N I A N CONTRIBUTIONS T O A N T H R O P O L O G Y N U M B E R 2 0 Page Inflation Nozzles 30 Bird Dart Points and Side Prongs 30 Harpoon Dart Heads 31 Bow and Arrow Equipment 31 Bows 31 Marlin Spikes 31 Sinew Twisters 31 Arrowshafts 32 Feather-Cutting Boards 33 Antler Arrowheads 33 Antler Arrowhead Preforms 37 Chipped Flint Arrowheads 37 Bola Weights 38 Miscellaneous Hunt ing Gear 39 Snow-Probe Ferrules 39 Ice Staff Rings 39 Snow Shovels 39 Snow Knives 39 Wooden Hunt ing Hats 39 Snow Goggles 40 Fish Arrow Prongs 41 Fishing Poles 41 Transportat ion Equipment 41 Boat Parts 41 Boat Paddles 41 Paddle Tips 41 Umiak Parts 43 Kayak Ribs 43 Kayak Cockpit Frames 43 Sled Parts 43 Antler Arches 43 Sled Shoes 44 Harness Swivels 44 Miscellaneous Boat or Sled Parts 44 Manufacture, Maintenance, and Processing of Equipment 44 Men's Knives 44 Crooked Knives 44 Knives with End Blade Slots 45 Composite Knives 46 Knife Blades 46 Whetstones 48 Flint Flakers and Hammerstones 48 Flaker Points 48 Hammerstones 49 Bone Hammerheads 49 Engraving Tools 49 Bow Drills 49 Firemaking Equipment 50 Drill Bearings 50 Fire Drill Platforms 50 Fire Drill Spindles 50 Adzes 50 Adz Handles 50 Adz Blades 50 CONTENTS Page Sleeves 51 Whale Rib Tools 51 Mattock Blades 51 Whale Rib Ice Picks 51 Whale Rib Mattocks or Pick Handles 52 Wedges 52 Ulus 52 Ulu Handles 52 Ulu Blades 53 Scrapers 53 Planoconvex End Scrapers 53 Pebble End Scrapers 53 Side Scrapers 54 Scraper Handles 55 Split Pebble Scrapers 55 Two-handed Scrapers 55 Scapula Scrapers 55 Miscellaneous Metapodial Scrapers 55 Humerus Scraper 56 Miscellaneous Bone Scrapers 56 Hide Pegs 56 Ladles and Spoons 56 Marrow Extractors 56 Wooden Trays 57 Pottery Lamps and Cooking Vessels 57 Baleen Buckets 57 Wooden Bucket Bottoms 57 Complete Baleen Buckets 58 Bone Crushers 58 Awls and Bodkins 58 Gauged Drills 59 Stone Drill Bits 59 Awl Handles 60 Bone Needles 60 Community Activity Artifacts and Equipment 60 Perforated Teeth 60 Beads 60 Labrets 61 Drums 61 Toys 62 Cup and Pin Game 62 Inserted Seal Vertebrae 62 Dolls 62 Toy Kayaks 62 Gaming Pieces 63 Toy Bows 63 Charms and Amulets 63 Bear Jaws 64 Miscellaneous and Unidentified Artifacts 64 Belt Toggles 64 Ivory Chains 65 Unidentified Ivory Artifacts 65 Miscellaneous Coal 65 Unidentified Bone Artifacts 65 Vi SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 Page Unidentified Wooden Artifacts 66 Unidentified Baleen Artifacts 67 Miscellaneous Stone Artifacts 67 FAUNAL ANALYSIS 69 Introduction 69 Procedure 69 Measurements 72 Analysis by Archeological Level and Cultural Units 75 Seals 76 Caribou 81 Whale 86 Polar Bears 86 Dog, Fox, and Wolf 86 Other Mammals and Birds 87 ARCHEOLOGICAL ANALYSIS OF THE WALAKPA CULTURAL UNITS 88 Historic Point Barrow Eskimo 88 Archeological Considerations 89 T h e Chronology of the Walakpa Site 90 Cultural Analysis 90 Birnirk 90 Early Thu le 92 Late Thu le 93 COMPARISON OF WALAKPA AND OTHER BIRNIRK AND T H U L E SITES 96 Northern Alaska 97 Western Alaska 102 St. Lawrence Island 103 Siberia 104 Canadian Arctic 104 Greenland 106 Summary 107 Early Birnirk 107 Middle Birnirk 107 Late Birnirk 107 Early Thu le 109 Late Thu le 109 DEVELOPMENT OF BIRNIRK AND T H U L E CULTURES IN NORTHERN ALASKA I l l LITERATURE CITED 115 PLATES 119 Tables 1. Climatic data from Point Barrow 9 2. Hunt ing and fishing equipment 42 3. Transportat ion equipment 44 4. Manufacture, maintenance, and processing equipment 61 5. Community activity artifacts and equipment 65 6. Distribution of bones of various species of animals 70 7. Distribution of bones of the various species of birds 72 8. Distribution of selected bone elements for polar bear, bearded seal, and walrus . . . . 87 9. Summary of the occurrence of various Birnirk and Early T h u l e harpoon types . . . . 94 10. Summary of artifacts and faunal remains 99 CONTENTS Vll Illustrations FIGURES Page 1. Summary of climatic data for the Arctic 8 2. Temperature-wind-chill index 8 3. Eskimo archeological sites in the vicinity of Point Barrow 10 4. Topographic map of Walakpa site 11 5. Schematic profile showing correlation of the stratigraphy of the A and B areas . . . 12 6. A area stratigraphy 13 7. B area stratigraphy 14 8. Contour map of the A and B areas 16 9. Seal humerus showing the distal condylar width and articular length 74 10. Seal femur showing distal condylar width and notch length 74 11. Seal tibia showing diaphyseal length 74 12. Seal bones from Walakpa site 77 13. Seal bones from Walakpa site 78 14. Seal bones from Walakpa site 79 15. Seal bones from Walakpa site 80 16. Seal bones from Walakpa site 81 17. Caribou bones from Walakpa site 82 18. Caribou bones from Walakpa site 83 19. Caribou bones from Walakpa site 84 20. Caribou bones from Walakpa site 85 21. Caribou bones from Walakpa site 86 22. Seasonal options and their general geographic location 88 23. B-9 occupation floor showing logs 91 24. Plan of house floor from B-8 occupation level 92 25. B-6 occupation floor 93 26. A-5 occupation floor showing possible tent structure 93 27. Occurrence of antler arrowhead ownership marks 93 28. Polar projection of the locations of Birnirk, and Early and Late Thule sites 98 29. Development of harpoon heads and antler arrowhead types 108 PLATES 1. Walakpa site. 2. Walakpa site. 3. Walakpa site A area. 4. Early Thule levels in B area. 5. Birnirk Occupation Level B-8. 6. Birnirk artifacts from the mound A test trench. 7. Birnirk artifacts from the mound B test trench. 8. Birnirk artifacts from the mound B test trench. 9. Birnirk artifacts from the mound B test trench. 10. Birnirk artifacts from the mound B test trench. 11. Birnirk artifacts from the mound B test trench. 12. Birnirk artifacts from the mound B test trench. 13. Small bow from the mound B Birnirk level test trench. 14. Birnirk artifacts from the mound B test trench. 15. Birnirk artifacts, level B-10. 16. Birnirk artifacts, level B-10. 17. Birnirk artifacts, level B-10. 18. Birnirk seal rattle, level B-10. Viii SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 19. Birnirk artifacts, level B-10. 20. Wooden hunting hat, level B-10. 21. Birnirk artifacts, level B-10. 21. Birnirk artifacts, level B-10. 22. Birnirk artifacts, level B-10. 23. Birnirk artifacts, level B-10. 24. Birnirk artifacts, level B-10. 25. Birnirk artifacts, level B-10. 26. Birnirk artifacts, level B-10. 27. Birnirk artifacts, level B-10. 28. Birnirk artifacts, level B-10. 29. Birnirk artifacts, level B-10. 30. Birnirk artifacts, level B-10. 31. Birnirk artifacts, level B-10. 32. Wooden meat tray, level B-10. 33. Birnirk artifacts, level B-10. 34. Birnirk artifacts, level B-10. 35. Drum rim, level B-10. 36. Birnirk artifacts, level B-10. 37. Birnirk artifacts, level B-10. 38. Birnirk artifacts, level B-9. 39. Birnirk artifacts, level B-9. 40. Birnirk artifacts, level B-9. 41. Birnirk artifacts, level B-9. 42. Birnirk artifacts, level B-9. 43. Birnirk artifacts, level B-9. 44. Birnirk artifacts, level B-9. 45. Birnirk artifacts, level B-9. 46. Birnirk artifacts, level B-9. 47. Birnirk artifacts, level B-9. 48. Birnirk artifacts, level B-9. 49. Birnirk artifacts, level B-8. 50. Birnirk artifacts, level B-8. 51. Birnirk artifacts, level B-8. 52. Birnirk artifacts, level B-8. 53. Birnirk artifacts, level B-8. 54. Birnirk artifacts, level B-8. 55. Baleen bucket, level B-8. 56. Polar bear mandible tied with Baleen cord, level B-8. 57. Birnirk artifacts, level B-8. 58. Early Thule artifacts from the mound A test trench. 59. Early Thule artifacts from the mound B test trench. 60. Early Thule polar bear carvings from the mound B test trench. 61. Early Thule artifacts from the mound B test trench. 62. Early Thule artifacts from the mound B test trench. 63. Early Thule artifacts, level B-7. 64. Early Thule artifacts, level B-7. 65. Early Thule artifacts, level B-7. 66. Early Thule artifacts, level B-7. 67. Early Thule artifacts, level B-7. 68. Early Thule artifacts, level B-6. 69. Early Thule artifacts, level B-6. 70. Early Thule artifacts, level B-6. 71. Early Thule artifacts, level B-6. CONTENTS 1 X 72. Early Thule artifacts, level B-6. 73. Early Thule artifacts, level B-6. 74. Early Thule artifacts, level B-5. 75. Early Thule artifacts, level B-4. 76. Early Thule artifacts, level B-4. 77. Early Thule artifacts, level B-3. 78. Early Thule artifacts, level B-3. 79. Early Thule artifacts, level B-3. 80. Early Thule artifacts, level B-2. 81. Late Thule artifacts from the area A test trench. 82. Late Thule artifacts from the area A test trench. 83. Late Thule artifacts from the area A test trench. 84. Late Thule artifacts, level B-l. 85. Late Thule artifacts, level A-1. 86. Late Thule artifacts, level A-1. 87. Late Thule artifacts, level A-2. 88. Late Thule artifacts, level A-2. 89. Late Thule artifacts, level A-2. 90. Late Thule artifacts, level A-2. 91. Late Thule artifacts, level A-2. 92. Late Thule artifacts, level A-2. 93. Late Thule artifacts, level A-3. 94. Late Thule artifacts, level A-3. 95. Late Thule artifacts, level A-3. 96. Late Thule artifacts, level A-3. 97. Late Thule artifacts, level A-4. 98. Late Thule artifacts, level A-4. 99. Late Thule artifacts, level A-4. 100. Late Thule artifacts, level A-4. 101. Late Thule artifacts, level A-5. 102. Late Thule artifacts, level A-5. 103. Late Thule artifacts, level A-5. 104. Late Thule artifacts, level A-5. 105. Late Thule artifacts, level A-5. 106. Late Thule artifacts, level A-5. 107. Late Thule artifacts, level A-5. 108. Late Thule artifacts, level A-6. 109. Late Thule artifacts, level A-6. 110. Late Thule artifacts, level A-6. 111. Late Thule artifacts, level A-6. 112. Late Thule artifacts, level A-6. 113. Late Thule artifacts, level A-6. 114. Late Thule artifacts, level A-6. 115. Late Thule artifacts, level A-7. 116. Late Thule artifacts, level A-7. 117. Late Thule artifacts, level A-7. 118. Late Thule artifacts, level A-8. 119. Late Thule artifacts, level A-9. Preface During the summer of 1967, Professor John M. Campbell, at that time Chairman of the Department of Anthropology at the University of New Mexico, conducted field recon? naissance in the areas of Anaktuvuk Pass and Chandler Lake, Central Brooks Range, Alaska. Before returning to New Mexico he spent several days at Point Barrow. There he purchased a sizeable collection of Eskimo antifacts from the Birnirk and Utkiavik sites, most of which were obtained from Mr. Charles Brower of Barrow. Dr. Campbell was impressed by the possible antiquity of the large midden deposits at Utkiavik and both he and Dr. Max Brewer, Director of the Naval Arctic Research Laboratory, were disturbed over the rapid destruction of the site from relic collectors and wave erosion. It was their opinion that Utkiavik should be excavated at the earliest opportunity, lest the site and its archeological record be permanently lost. Upon Dr. Campbell's return to Albuquerque, he urged me to go to Point Barrow the following summer for the purpose of conducting salvage-archeological operations and with the scientific aim of possibly discovering new evidence relative to the controversial question of the origin of Eastern Thule culture. It was agreed that I would lead a team to Point Barrow and work on these problems with Charles W. Amsden as a co-leader. Mary M. Rush ton was to work on problems relat? ing to temporal changes of tool kits and corresponding changes in the local ecology. Mary Rushton intended to work on seriation of harpoon-head types with the idea of comparing her conclusions with Ford's (1959) harpoon seriation. With the Barrow Town Council approval, we began excavating Utkiavik in June 1968. The crew consisted of Peter Eidenbach, Marc Stevens, Mrs. Natalie Pattison, Simeon Kunaknana, and James Itta. The remaining Utkiavik middens were mapped with 2-meter grids and test trenches for profile controls were started. After one week our excavations were discontinued at the request of the town council. Apparently several prominent members of the community were concerned that our excavations would dis? rupt their relic trade. After many frustrating attempts to convince them of the urgency of our work and its scientific significance, we temporarily abandoned the project. Being in Alaska with a team ready to conduct field work, we were not willing, however, to abandon our major theoretical problems and so decided to examine the possibilities of excavating other sites in the Barrow area. With Robert Jack Smith, a civil engineer from Rawlins, Wyoming, who joined our expedition to help with the surveying, I tested the site of Nuwuk located at the tip of Point Barrow. After making several test trenches we decided that this site did not contain sufficient time depth to be useful in working on the problem of Thule origins. All artifacts collected were of historic Nuwuk age. Kenneth Toovak, maintenance supervisor of Naval Arctic Research Laboratory, thought that we should examine the house mounds at Walakpa Bay, which, according to local legend, are the most ancient in the area. Walakpa Bay, is sometimes known as "monument" as there is a monument on the North shore erected in honor of Will Rogers and Wiley Post, who crashed on that spot in 1935 (Plate Id). After consulting Dr. Max Brewer, an expedition was launched to examine Walakpa, to explore the coast for other archeological sites, and to investigate additional house mounds reported earlier by Ford (1959:18). Our first stop was at the former site of Nunavak, located two miles (3.3 km) south? west of Barrow village, where we learned that all traces of the former village had been destroyed by gravel operations of the Burgess Construction Company. All that remains of this former site are several burial mounds, which had been excavated by Fred Hopsen in x i Xii SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 1929 (Ford, 1959:21). Due to an impending snow storm and my impatience to see Walakpa, we decided to take the inland trail and go directly to the bay rather than search? ing for sites between Nunavak and Walakpa. When we arrived at Walakpa, the crew was divided for a more efficient approach to the various problems before us. Amsden, Eidenbach, and Stevens were to explore the area around the main mounds and judge the size and extent of the occupation, while Mary Rushton, Sherry Heizer, and I made a preliminary test trench into the large trash midden. Our work was rewarded by finding evidence that both Birnirk and Thule cultures were represented at Walakpa. We decided that the Walakpa midden was indeed a site with considerable antiquity and possibly contained evidence bearing on the origin of Eastern Thule culture. It was due to these several circumstances, therefore, that the excavations at Walakpa were undertaken in the summer of 1968 and continued during the summer of 1969. The work of those two summers was supported by grants-in-aid to Dr. Campbell, University of New Mexico, from the Arctic Institute of North America, and the Office of Naval Research, Department of Navy, as well as from the New York Explorers Club. As anyone knows who has worked in the Arctic, and especially at Point Barrow, field conditions are less than desirable. One soon learns to expect a continual series of disasters, mostly propagated by adverse weather conditions. It is therefore important that special acknowledgments be made to the field crews who assisted me during both seasons. These people, who without pay, suffered snowstorms, subzero temperatures, frostbite, perma? frost, mosquitos, and other deprivations, include (in 1968) Charles Amsden, Mary Rush- ton, Marc Stevens, Natalie Pattison, Peter Eidenbach, Walter Akpik, Nate Elavgak, James Itta, and Simeon Kunaknana; (in 1969) James Duguid, Peter Eidenbach, T . Weber Grieser, Natalie Pattison, Susan Kaplan, Edwin Samuelson, Steve Beckerman, and Walter Akpik. Without their generous and industrious support, it would not have been possible to conduct this field excavation. Laboratory work, by sheer numbers of artifacts and faunal remains, seemed to be an insurmountable task. Over 60,000 bones and 7000 artifacts had to be cleaned, preserved, identified, and cataloged. Through the untiring work of Natalie Pattison, Susan Kaplan, Dan Witter, and Alison Witter, with a total cash budget of less than $500, this work was accomplished in less than two years. Laboratory space was provided by the University of New Mexico. Unfortunately, during the course of the study, due to administrative problems, the laboratory had to be moved no less than four times. During these moves, a number of the artifacts were lost or misplaced. Fortunately most of the descriptive work was completed before these trage? dies occurred. It is for this reason, that all of the artifacts described in the text are not illustrated. Those artifacts which survived are now housed at the Maxwell Museum of Anthropology at the University of New Mexico, Albuquerque, New Mexico. I am indebted to a great many people for their advice, encouragement, and assistance throughout the production of this monograph. These people include John M. Campbell of the University of New Mexico, William Taylor of the National Museum of Canada, Robert McGhee of the National Museum of Canada, Douglas Anderson of Brown Uni? versity, Wilbert Carter of Tufts University, and Kiyushi Yamaura, a predoctoral fellow at the Smithsonian Institution in 1973-1974 from the University of Tokyo, who were most generous with their time, collections, and ideas. T o Drs. Betty Meggers, Clifford Evans, Henry Collins, William Fitzhugh of the Smithsonian Institution, William Taylor of National Museum of Canada, and John M. Campbell of the University of New Mexico, all of whom suffered through various versions of this manuscript, I owe special thanks. I have incorporated most of their suggestions to make this a more readable, better organized and more accurate scientific monograph. Special acknowledgment is also due to those people whose technical knowledge and efforts were required. These people include Mr. Robert Jack Smith of Rawlins, Wyom? ing, who surveyed and mapped the Point Barrow sites, Steve Beckerman who was the site photographer, Dick Dunatchik of the University of New Mexico and Victor Krantz of P R E F A C E X l l l the Smithsonian Institution who photographed the specimens. The scientific illustra? tions were done by Edward Schumacher, George Robert Lewis, and Marcia Bakry of the Department of Anthropology, Smithsonian Institution, who converted crudely done sketches, maps, and charts into magnificent scientific illustrations as well as Paula Card- well for her efficient typing of a difficult manuscript. Special appreciation is due to editor Joan Horn of the Smithsonian Institution Press for her ability to bring metamorphic change to a monograph so that it is now easily read? able and consistent and whose editorial guidance on my first major publication should make her and my job simpler next time. T o Al Ruffin, Series Managing Editor, and Vern Shaffer, Series Production Manager, of the Smithsonian Institution Press, I owe a word of thanks for their advice and assistance at various stages of production of this monograph. A note of thanks to scientific collaborators from other fields includes Dr. James Duguid of Princeton University, who worked on the geology of the Walakpa Bay area; Dr. Loren Potter, who worked on the ecology of Walakpa Bay, Dan Witter of the Uni? versity of New Mexico, who identified the faunal remains, and Dr. Clayton Ray of the Smithsonian Institution, whose work on the paleontology provided important informa? tion that led to the conclusions based on archeozoological interpretations. James Toler of Albuquerque Technical Vocational Institute, Don Ortner and Neal Roth of the Smithsonian Institution provided computer assistance for handling the bulk of the raw data. Finally, to my wife Jeanne, who for two summers was an archeological widow and subsequently lived for three years with all of the artifacts and faunal remains from the Walakpa site scattered throughout her living room, dining room, and bedroom, I owe my deepest debt of gratitude. THE WALAKPA SITE, ALASKA Dennis J. Stanford Introduction O R I G I N S OF T H U L E ESKIMO C U L T U R E The antiquity of the Eskimo occupation of the Point Barrow area of Alaska was first noted by Lieutenant P. H. Ray (1885). During the Interna? tional Polar Expedition to Point Barrow, Alaska (1881-1883), Ray observed many ancient villages along the coast and in the interior (Ray, 1885:37). The site of Perigniak (Birnirk) was observed, and its antiquity was apparent as "it stands in the midst of a marsh . . . and it is their custom to select the high, dry points of land . . . for their villages . . . and the fact of our finding a pair of wooden goggles twenty-six feet below the surface . . . points con? clusively to the great lapse of time since these shores were first peopled" (Ray, 1885:37). Early excavations conducted at Point Barrow (Stefansson, 1914; Mason, 1930) also pointed to an occupation of considerable antiquity in the area. Wissler (1916), in considering Stefansson's collection from Point Barrow pointed out the similarity between harpoon heads found at Birnirk and those reported by Bogoras from ancient sites in Siberia (in Wissler, 1916:439). Wissler also noted a greater similarity among the harpoon heads from old sites in Siberia, Alaska, and the Hudson Bay district than among those found in ethnographic collections from these areas (Wissler, 1916:440). Five archeologists have contributed to our under? standing of the significance of Wissler's observation and specifically to the origins and spread of Eskimo culture in Northern Alaska, Canada, and Greenland. These archeologists are Mathiassen (1927b; 1930a), Collins (1933, 1935, 1937b,c; 1940; 1951), Ford (1959), Taylor (1963), and McGhee (1970). Their research and conclusions are the most relevant works on the subject of Birnirk and Thule origins and must neces- Dennis J. Stanford, Department of Anthropology, National Museum of Natural History, Smithsonian Institution, Washington, D.C. 20560. sarily be reviewed before attempting further discus? sion of late Eskimo cultural developments. The first systematic archeological studies in the Arctic were conducted between 1921 and 1924 by Therkel Mathiassen of the fifth Thule expedition, which was organized by Knud Rasmussen. On the basis of his excavations in the Central Canadian Arctic, he defined the Thule culture (Mathiassen, 1927b). The Thule Eskimos, according to Mathiassen, were those people who represented a cultural horizon older than that of the modern Central Eskimos (Mathiassen 1927b: 182). Mathiassen described Thule culture as dependent upon marine resources, pri? marily whale, seal, and walrus. He further defined 152 elements as representative of Thule material culture; and contended that wherever all of these elements appear in archeological context, Thule cul? ture is necessarily represented. By plotting the occur? rence of these artifacts from the various Arctic collections that he was able to study, Mathiassen con? cluded that Thule, with its highly evolved tool kit, must have originated in an Arctic area where whales and other large marine mammals were abundant. Further, due to the presence of such items as the women's boat and conical-shaped tents, the place of origin must have been an area where wood was present. He believed this to have been in Alaska (Mathiassen, 1927b: 182). Mathiassen (1927b: 183) pointed out that there are many traits of the Thule culture, the prototypes of which are known only from the Alaskan regions. Examples are ornamental harpoon side blades, the winged needle case, the Y-shaped ornamental design, barbed antler arrow points and round earthenware cooking pots. But, having made this formal bow to what was indeed compelling evidence, he never re? ferred to it again. On the contrary, when he discussed SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 the chronological relationships between Birnirk and Thule harpoon heads, he argued that the Thule harpoon heads were older, because they were simpler in form and had a wider geographical distribution in the Arctic (Mathiassen, 1930a,b). Birnirk harpoon heads, he thought, were a later, local type influenced by the Old Bering Sea culture. The first papers written by Henry Collins on Eskimo archeology were, in considerable measure, directed to putting Thule culture into a chronologi? cal perspective. Reporting on his work of 1928, Collins (1930:149) indicated that early Alaskan finds from St. Lawrence Island exerted a strong influence to the east as evidenced in Mathiassen's Thule col? lections, and that the Thule culture was clearly derived from Alaska, from where it seems to have spread at a relatively late period. In considering Mathiassen's claim for the antiquity of Thule, based on its simple form and geographical distribution, Collins (1929) showed that these axioms could not be applied in the Arctic. The complex prehistoric art and harpoon-head sequence, which Collins worked out for Alaska, pointed to the opposite conclusion?that the Birnirk harpoon heads were older than the Thule harpoon heads (Collins, 1929: 45). According to Collins, the results of the excavations at Kurigitavik at Cape Prince of Wales, Alaska (Jen- ness 1928; Collins, 1937a, 1940) provided confirma? tion of the Birnirk-Thule sequence in Western Alaska. The presence of Birnirk, Western Thule, Punuk, and modern Western Eskimo traits at Kuri? gitavik, which are not found in the Canadian Thule culture, indicated to Collins that the eastward mi? gration of Thule culture must have originated from some point farther east, such as Point Barrow (Col? lins, 1940:562). Another interpretation of the development of Es? kimo culture has been offered by James A. Ford (1959). After careful analysis of the artifacts from the Birnirk site, Ford (1959:232) seriated harpoon head types from throughout the Arctic. In his analy? sis he considered five significant types of harpoon heads: (1) those with side blades or barbs at right angles to the line hole; (2) those with side blades or barbs in the same plane as the line hole; (3) those with a slit for an end blade in the same plane as the line hole; (4) those with a slit for an end blade at right angles to the line hole; and (5) those that have closed sockets. These harpoon types were seriated in terms of their individual traits, proposed temporal position, and their areal distributions. Ford also constructed trait- comparison lists of artifacts shared by Birnirk and other Eskimo cultures. He discussed these traits in terms of their earliest occurrences and their spread through time and space. The harpoon seriation along with the trait comparison led Ford to construct a complicated system of cultural relationships and de? velopment indicating that Birnirk culture was de? rived from Okvik and Old Bering Sea cultures, with influences from Ipiutak and early Punuk. Canadian Thule did not stem directly from Birnirk, but that it was derived from the "Nunagiak stage" (Ford, 1959:241). The Nunagiak stage, named for the Nunagiak site located near Wainwright, Alaska, was thought by Ford to correspond with the end of the Birnirk period, dating around A.D. 1200. This would imply that the Thule culture of Canada could not be any older than A.D. 1200. William E. Taylor, Jr. (1963) has opposed Ford's views on the origin of Thule culture. Taylor's opinion was based on data that were unavailable to Ford at the time of his study. These data are pri? marily (1) collections of Canadian harpoon heads of the "Birnirk" types?Natchuk and Sicco from Devon Island, the Inman River, Booth Islands, and Lady Franklin Point, Victoria Island; and (2) radio? carbon dates of Thule occupations at Inugsuk, which give ages of A.D. 1010 ? 120 and A.D. 1240+120. There? fore, Taylor maintained that the trait comparisons made by Ford between Birnirk, Nunagiak, and Thule are inadequate to support Ford's conclusions. On the basis of this new evidence noted, Taylor (1963:461) concluded that a) Birnirk culture extended as far east as Cape Parry prior to A.D. 900, and that this eastward spread of Birnirk con? stituted a first stage in the eastward Thule migration; b) Canadian Thule developed not in Alaska, but from this eastern extension of Birnirk; c) Thule culture generally developed from Birnirk along the arctic coast between Capes Prince of Wales and Parry; d) an early Thule stage must separate Birnirk from Nunagiak; e) Nunagiak, or something closely akin, existed as far east as Lady Franklin Point on Victoria Island; f) finally, the existence on the Beaufort Sea Coast of a Proto- Thule stage, between Birnirk and Nunagiak, allows one to suggest a beginning for the latter, not around A.D. 900 but about A.D. 1100 or 1200. In a recent paper, Robert McGhee (1970) has indicated that Thule probably developed out of Birnirk in Northern Alaska and spread eastward across the Canadian Arctic to Greenland. McGhee (1970:173) stated that the uniformity of early Thule culture and the close-clustering of C-14 dates sug? gest this expansion took the form of a single rapid migration. It was McGhee's opinion that the transi? tion from Birnirk to Thule took place due to the effects of the climatic changes proposed by Bryson and Wendland (1967:280). These climatic changes INTRODUCTION are (1) the Sub-Atlantic (550 B.C.-A.D. 400), (2) the Scandic (A.D. 400-900), (3) the Neo-Atlantic (A.D. 900-1200), (4) the Pacific (A.D. 1200-1550), (5) the Neo-Boreal (A.D. 1500-1850), and (6) the Recent (A.D. 1850-1960) episodes. McGhee (1970:175) has outlined the major effects of these episodes in the Arctic. He saw the Scandic as a transition between the cooler Sub-Atlantic, and the warmer Neo-Atlantic. He stated that the Scandic was probably similar to the climatic conditions that are found today in the Arctic. He further thought that the Neo-Atlantic was apparently warmer and ice conditions would be significantly different from cooler periods. Drift ice would be reduced, the pack ice would retreat from the coast, open water would last longer, and new ice would form later and be thinner. The result would be an increase in whales, walruses, and bearded seals, with a reduction in ringed seals. This would necessitate cultural adjustments from spring ice lead hunting to summer whaling on open water. McGhee (1970:178) believed that Thule culture developed during the Neo-Atlantic period and spread eastward into the Western Canadian Arctic. This spread was thought to have consisted of movements of small groups that were large enough, however, to hunt whales. They were not associated with large settlements and moved several times during the year to wherever subsistence was assured by either the successful killing of a whale or by other means. GEOPHYSICAL CONDITIONS Geology The Quaternary geology of the Point Barrow area has been well defined (Black, 1957). The Gubik formation comprises the entire Quaternary unit throughout the Arctic coastal plain of Northern Alaska, and it mantles a Cretaceous unit that is com? posed of horizontal, bedded sandstone and shale. These sandstone and shale deposits are frequently exposed and are a potential source of raw materials for the manufacture of artifacts (Paige, Foran, and Gilluly, 1925). The Gubik is comprised of lenses and admixtures of silt and fine-grained sand, which are predominantly quartz and chert. The sediments are primarily marine, but are partially fluvial, lacustrine, and eolian in origin (Black, 1957:1701). The Gubik is made up of three lithologic units: (1) Skull Cliff; (2) Meade River; and (3) Barrow. The Skull Cliff unit is generally widespread throughout the Alaskan coastal plain. It is the oldest unit, dating to 38,000 B.P. (W-380) in the Barrow area (Coulter, Hussey, and O'Sullivan, 1960). Its maximum thickness is about 6 meters (20 ft), consisting of poorly sorted clays and gravels, which are both marine and glacial in origin. The intermediate stage, the Meade River unit, has been frequently eroded and at Walakpa it does not exist, as the Barrow unit unconformably overlies the Skull Cliff unit. In other localities the Meade River unit reaches a maximum thickness of nearly 60 meters (200 ft) and consists of light-colored, well-sorted marine sand, grading upward into loess in the south? ern and southeastern areas of the coastal plain. The Barrow unit, the youngest, dates possibly as late as 2000 B.P. or later and rarely exceeds 6 meters (20 ft) in thickness. It consists of poor to well-sorted mixtures of clay, silt, sand, and gravel. In many areas ice constitutes more than one-half of its structure. Organic matter is abundant in the upper part of the unit. The sediments are generally marine, but are locally fluvial and lacustrine in origin (Black, 1957: 1701). At Walakpa, the Gubik formation comprises only two lithologic units?Skull Cliff and the Barrow unit. The Barrow unit is of prime importance to the Walakpa studies. Its upper portion consists of several feet of eolian sand. It is in this sand that the first occupation at Walakpa, Walakpa Phase Denbigh, is found. Above the Walakpa phase, the Barrow unit continues to be eolian in origin for 5-10 centi? meters. Above this, it grades into an organic matter, mixed with gravels and sand (James O. Duguid, pers. coram.). Physiography and Ecology The coastal plain consists of numerous meandering streams, lakes, marshes, and varying types of patterned ground. It is nearly flat except for occasional low re? lief hills and ridges. The plain is treeless, with a ground cover of grasses, sedges, willows, and mosses. The northern coast of Alaska, although harsh, possesses unique qualities, which have great potential for economic exploitation by nonmechanized, non- industrial cultures. Four recognizable ecosystems exist. The low tundra ecosystem and the marine eco? system are of the first order of importance in terms of SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 economic potential. The ponds and river ecosystems are of the second order of importance. These ecosystems are so situated that they could all be successfully exploited from a central geo? graphical area. Moreover, each could act as a buffer for the others, so that during certain seasons when the food supply of one ecosystem is at a low point, another system could be exploited. An important point is that these ecosystems have seasonal and long- term cycles of production so that hunters could, with a high degree of success, predict a successful pattern of exploitation. Low TUNDRA ECOSYSTEM The low tundra ecosystem is part of the Arctic coastal plain physiographic province (Spencer, 1959:9). This ecosystem, and more specifically, the interrelationships of the use of the low tundra and the marine ecosystem, is of prime importance to this study. The Walakpa midden is located at the point of contact between the low tundra, Walakpa Bay, and the Chukchi Sea. The ecology of the low tundra eco? system has been examined twice at Walakpa Bay (Wiggens and Thomas, 1962; Potter, 1972), but the ecological role of Walakpa Bay and the Chukchi Sea at Walakpa, has not to my knowledge been studied. At Walakpa, the low tundra ecosystem consists of six principal physiographic forms: (1) ponds and low-centered polygon pans; (2) sedge-willow mead? ows; (3) depressions between high-centered polygons; (4) high-centered frost polygons; (5) polygons with frost boils; and (6) coastal beaches and bluffs. The following descriptions of the physiographic forms are based on Potter's work in the Walakpa area (Potter, 1972). PONDS AND LOW-CENTERED POLYGON PANS.?The dominant relief of the tundra north and east of Walakpa is elongated, with oval-shaped basins, which have a longitudinal axis of approximately 9? west of north. These basins are in varying stages of being filled in and range from being well-drained to marshy with low-centered polygons. The vegetation of these ponds and low-centered polygon pans includes Arctophila fulva, which is found in the mud, wet gravel, and shallow water around the margins of the ponds. A second grass found around the wet margins is Dupontia fisheri. Grass-like dominants include Carex aquatilis and Eriophorum anqustijolium. T h e most common aquatic herbs are Hippuris vulgaris and Hippuris tetraphylla. Ranunculus pallasii and Ranunculus gmelini are found in shallow pools and lake margins. SEDGE-WILLOW MEADOWS.?Sedge-willow meadows are found on the gentle slopes of the principal drain? age channels. Because there is little relief, vegeta- tional variations are not pronounced. Over 90 per? cent of the vegetation consists of Carex spp. and Salix rotundifolia. Other vegetational types include Peta- sites frigidus, Dupontia fisheri, Poa spp., Luzula spp., Salix pulchra, and Saxifraga cernua. Several other species make up less than 1 percent of the rela? tive cover. DEPRESSIONS BETWEEN HIGH-CENTERED POLYGONS.? Depressions or troughs surround raised-centered poly? gons. These two types of topography are the most common around the Walakpa site. It is also in these depressions that the principal habitats of the lem? mings are found. The troughs are covered with a dense growth of Carex aquatilis, Dupontia fisheri, and Eriophorum spp. Petasites frigidus is found in clumps on the upper edges of these depressions. Extensive mats of sphagnum are sometimes found in the lower part. HIGH-CENTERED FROST POLYGONS.?High-centered frost polygons, when mature, can reach a height of 0.9 meter (3 ft). These polygons are well drained and have a dense mat of tufted vegetation. Phanero? gam cover is quite high. Salix pulchra makes up 31 percent of the foliage cover. Patches of willows cover the surface with a mat of woody stems. Carex spp. make up 30 percent of the relative cover of the poly? gons. Petasites frigidus, Luzula spp., and Salix arbutifolia are found near the centers of the poly? gons. The grasses, Poa spp., Arctagrostis latifolia, Festuca brachyphylla, and Dupontia fisheri are found on locally drier sites. The transitional area between the sloping sides of the polygons and the more moist depressions have tufts of Eriophorum spp., Carex spp., and Petasites frigidus. POLYGONS WITH FROST BOILS.?When high-cen? tered polygons become dissected by contraction cracks and frost boils, the plant types change. Minor changes result from lemmings, which use the cracks as runways and enrich the peat with nitrogenous wastes. Vegetation changes with improved drainage caused by the cracks. Major changes also result from mineral soil of silt and clay, which is thrust up from below by frost boils. Luzula spp., Eriophorum spp., and Carex spp. make up the dominant cover of these disturbed polygons. Other than crustose lichens, much of the disturbed surface has no vegetation for several years, except that found in the contraction cracks. Salix pulchra and S. arbutifolia grow over the surface from the edges, unless they are disturbed by additional frost activity. Rubus chamaeorus, cloudberry, Vaccinium vitis-idaea, and lingonberry are also found on disturbed polygons. INTRODUCTION COASTAL BEACHES AND BLUFFS.?The shorelines of both the Chukchi Sea and the sand splits, which separate the open sea from Walakpa Bay, are re? peatedly scoured by waves, ice, rolling sand, and gravel. For this reason they are nearly devoid of vegetation. In the protected bay areas the coastal spe? cies of Mertensia maritima, Arenaria peploides, Elymus mollis, and Puccinellia phryganodes are found. Patches of Cochlearia officinalis, Cerastium beeringianum, Stellaria humifusa, and Saxifraga caespitosa are also common. The exposed areas of the bluff face, brink, and gully sides are colonized by tap-rooted plant species such as Cochlearia officinalis, Taraxacum lyratum, Sagina intermedia, and Oxyria digyna. The dande? lion, Cochlearia officinalis, is found in these areas and could possibly be an important food plant. Other common species include Cerastium beeringianum and Saxifraga rivularis. Wiggins and Thomas (1962) have identified 59 species of plants in the Walakpa area. Of these, 17 percent are known to be used by the Alaskan Eskimo for possible food sources (Heller and Scott, 1967). These are Carex aquatilis, Eriophorum angustifolium, Salix pulchra, Oxyria digyna, Rumex arcticus, Honckenya peploides, Caltha palustris, Ranunculus pallasii, Cochlearia officinalis, Saxifraga punctata, Hippuris vulgaris, Vaccinium vitis-idaea, Pedicularis lanata, and Petasites frigidus. None of these plants was found in the archeo? logical investigations at Walakpa, probably because of the seasonal occupation of most of the cultural levels. It is a reasonable assumption, therefore, that plants were of minor importance as a food resource for the people who lived at Walakpa. Ethno- graphically, I observed only one plant, Petasites frigidus, being eaten. This plant has a broad, green leaf, which is eaten raw. Other plants are utilized for nonfood purposes, such as willows (Salix), which are used for floor matting in both houses and tents, packing muluks, diapers, and fire tinder (Potter, 1972:130). Young flexible stems of Salix pulchra were found in the archeological assemblage, used for bindings on scrapers and knives. The Caribou (Rangifer arcticus) is the principal tundra food resource for the Point Barrow Eskimo. Those caribou that are found in the Point Barrow area represent the northernmost occurrence of the Arctic caribou herd, which occupies a range of ap? proximately 362,600 square kilometers (140,000 sq mi), and numbered more than 242,000 animals in the 1970 census (Hemming, 1971:5). The primary winter range of this herd is south of the Brooks Range, where food resources can be found; however, some caribou stay on the Arctic slope until the snow gets too deep to find food, then they will move toward the coast, where food can be found in semi-open areas, which are relatively free of snow. The spring brings on the calving season and most animals move to the major calving areas lo? cated on the headwaters of the Colville, Ketik, Meade, and Utukok rivers. The important food during this season is the cottongrass (Eriophorum vaginatum) shoots. After calving, summer distributions are es? sentially random and caribou can be found anywhere west of the Savavanirktok River to the Arctic Ocean and the Chukchi Sea. Towards the fall the herd be? gins to drift slowly toward the tree line. The diet of the caribou, as described by Skoog (1968:81) and Hemming (1971:7?9), consists of willow shrubs (Salix spp.), sedges (Carex spp.), cottongrass (Eriophorum vaginatum) and lichens (Cladonia spp. and Certraria), which are all found at Walakpa (Potter, 1972:125). Caribou are often found in the summer and fall in the immediate area of Walakpa Bay; thus during historic times the area has been known as a good caribou hunting location. Other possible food animals found on the low tundra are musk ox (Ovibos moschatus) and moose (Alces americana). The bones of these animals are rare in most of the occupation levels at Walakpa and are not considered to have been important. Re? mains of the arctic fox (Alopex lagopus) and the wolf (Canis lupus) are found in the faunal collection. In historic times these animals were rarely eaten and were probably killed for their skins. It is possible that these animals were a prehistoric food resource, as in the Birnirk level there is a small species of dog that was apparently eaten. MARINE ECOSYSTEM For man, the main economic resources of the marine ecosystem consist of whales (Balaena mysti- cetus, Delphinapterus leucas), walrus (Odobenus rosmarus), bearded seals, (Erignathus barbatas), seals (Phoca vitulina, Phoca hispida, Phoca fasciata), and certain fishes. Ice covers the surface of the ocean offshore from the Walakpa site for most of the year and exercises an important influence over hunting activities. There are several types of ice and ice conditions based on the age of the ice and its relationship to the coast. These types of ice and their movements, due to cur? rents and winds, are of extreme importance in ex? ploitation of the sea. Sea-ice conditions and their effect on hunting has been summarized by Nelson SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 (1969). His observations are the basis of the follow? ing description. SUMMER ICE.?The ocean is usually ice-free for a very short period during the summer. If the summer is particularly cold, or if there are frequent, onshore winds, the ocean may be ice-bound all year around. Even during a relatively warm summer, ice floes and drifting ice are common along the shore. The ocean is usually ice-free, however, from late June through late September or early October. The open water of the ocean is not utilized to any great extent, and economic activities are gen? erally concentrated on land. Seals are hunted from boats, and walruses are found on the ice floes during late summer. Open-ocean whale hunting is not prac? ticed at Barrow, as the most effective ice-lead hunting is conducted in the spring and fall. The small Beluga whales are shot from the shore during this period but are not pursued. The hunter usually shoots them near shore and waits for them to be washed ashore. FALL ICE.?The new or young ice forms in the fall. Its occurrence and condition is erratic, as it is subject to movement by winds and currents (Nelson, 1969:11). New ice forms on the ocean and appears when onshore winds or currents wash it ashore. By the end of November the young ice is usually land- fast and will thicken quickly. Ice hunting is possible after it thickens, which is usually in December. An ice apron is the formation of new ice along the edges of open-ice leads in either young ice or winter ice. Ice aprons may develop rapidly or slowly, depending on the current direction. If the current parallels the lead edge, the apron will grow quickly and soon cover the entire lead. Most travel and hunting are done on the ice aprons. The aprons are thicker near the base and become thinner towards the edge. It is of extreme importance to be able to gauge the thickness and condition of apron ice and new ice. The color of the ice will indicate whether or not it is safe for travel. WINTER ICE.?There are two types of winter ice: (1) winter ice (ice of the year) and (2) polar ice (old ice) (Nelson, 1969:28). The winter ice is not common except during midwinter. It is thick enough that a hunter need not worry about it breaking. It can reach a thickness of 3.5 meters (10 ft), but when it becomes piled it reaches up to 18.3 meters (60 ft) in height. This type of ice is quite mobile and is subject to cracking and opening of ice leads. This causes the surface to become piled and jumbled with pressure ridges. Polar ice is not flat or piled as winter ice, rather it has developed a gently undulating surface. This is caused by weathering from the warmer summer temperatures. The Polar ice will form jagged, rough edges along its limits and cracks. Polar ice will de? velop large fissures up to 4.57 meters (15 ft) deep, which freeze over almost level with the surface (Nel? son, 1969:31). Polar ice, unlike winter ice, is usually not completely covered with snow. This is because the snow will not blow free over the moist surface of salty, winter ice, while the Polar ice is free of salt. Landfast ice is ice that has been driven up on shore or into the shallow water along the shore. It becomes solidly anchored to the ocean bottom or to the beach. When the offshore ice moves, due to currents or winds, the landfast ice will usually stay in place, and ice leads will open along its ocean edge. It may well be the most important type of ice for economic exploitation of the sea. Most of the hunting and traveling during the fall, winter, and spring are carried out on this type of ice. The effects of winds and currents on the ice pack and ice floes are of prime importance to the Eskimo hunter. During the fall, winter, and spring, the wind direction and velocity is more important than the currents. In fact, currents are sometimes deter? mined by the wind direction and velocity. The most common winter wind is from the north? east. This is an offshore wind, which causes the ice to drift seaward and opens up ice leads between the pack ice and the landfast ice. This is a dangerous time to be hunting out beyond the landfast ice. It is, however, an extremely favorable time to hunt seals along the leads from the landfast ice. It is also necessary for successful whale hunting. An east wind will have the same effect as a northeast wind on ice movement, but east winds are relatively rare at this time of year. When there is a southeast wind accompanied by a south or southwest current, the ice will remain fast against the coast. Onshore winds will also hold the ice landfast. During these periods, it is safe for a hunter to travel a good distance away from land in search of polar bears. A southeast gale is quite dangerous. It can open wide leads and remove land- fast ice. This is because a wind of gale velocity ac? companied by a high tide can lift the landfast ice from the ocean bottom and drift it out to sea. When traveling or hunting on ice during the winter, the hunter is obliged to test both winds and currents before leaving shore. Currents often shift before an approaching wind change and allow a cer? tain amount of predictability before ice travel. During the summer the currents are more im? portant than winds. Summer ice floes are carried northward along the coast by the strong and powerful south current. In the early summer a west wind will INTRODUCTION pack the ice up along the shore, but by late summer even a strong, onshore wind will not stop the ice from drifting along the coast or out to sea. These ice floes and their movements are extremely important for hunting seals, bearded seals, and walruses. It is important for the hunter to make the most of the summer landfast ice and the drifting floes. If the weather is favorable, a large amount of meat can be obtained and stored at this time. POND AND RIVER ECOSYSTEMS The pond and river ecosystems are not of extreme importance to this study. The lakes, ponds, and marshes have been analyzed above for vegetational composition as physiographic features of the low tundra ecosystem. As far as I could determine, none of the waters in the Walakpa area contain fish. Some lakes, such as Tusikvoak and others, are important for winter fishing, but they are long distances from Walakpa. During historic times the pond ecosystem has been secondarily utilized because waterfowl nest along the shores of these lakes. The only running stream in the Walakpa area is the Walakpa River. This river is not economically important, as it has no fish. Walakpa Bay does not contain fish, but occasionally a seal will go into it. Waterfowl and caribou are also taken in the bay. T o the east of Point Barrow, the Kulugrua, Ikpikpun, and Meade rivers are important in the summer. Many families will establish fishing camps there and net a great many fish. However, the only fish remains found at Walakpa were those of tomcod (Boreogadus saida), which are netted or jigged from the ocean. Prehistoric Climates and Cl imatic Change Little is known of the paleoclimatic conditions or changes that have occurred at Point Barrow. Only several minor indications are available. The first is that of the pollen analysis of radiocarbon-dated coastal peat by Colinvaux (1964:707). The results of Colinvaux' analysis indicate that the vegetation and climate of the Arctic coastal region have been rather stable over the last 4000 years. However, his studies were not aimed specifically at the later climatic conditions. Analysis of the formation of beach ridges by Hume (1965), Rex (1964), Brown (1965), Brown and Sell- man (1966), and Moore (1960) have indicated that there have been rises and falls in the level of the sea along the north coast of Alaska. These fluctuations of the sea level are summarized as (1) a period from approximately A.D. 500-1000, when the sea level dropped 2 meters below the present sea level; (2) a high period from A.D. 1100 to 1400/1500, when the sea level rose 3 meters; (3) a drop in the level of the sea between A.D. 1400 and 1500; and (4) high level between A.D. 1700 and 1850 (Figure 1, Column 1). These sea level fluctuations do not necessarily reflect climatic conditions or changes, but with additional data presented by Bryson and Wendland (1967), they tend to correlate with climatic fluctuations in other areas of the Arctic. For additional climatic data we must turn to other areas where work has been done in the North. Whether or not these data are applicable to the Bar? row area is problematical, but they are used as indica? tors of general climatic trends in the Arctic. Koch (1945) studied ice and ice conditions in eastern Greenland and Iceland. He consulted the written works on Norse observations of ice condi? tions and their effects on life in both Greenland and Iceland. Ice conditions and climatic changes were expressed by the necessity of the Norsemen to change their transportation routes between Greenland, Ice? land, and Northern Europe because of ice accumula? tions and ice floes. Koch also consulted accounts on grazing and agriculture in both Greenland and Ice? land. From these data he constructed graphs of ice conditions showing the number of weeks in which ice was around Iceland for 20-year periods from A.D. 860 to 1939 (Figure 1, Column 3). Koch (1945:354) concludes that during this span there were five major periods of climatic change: 1) In the period A.D. 800-1200 there was hardly any ice in summer near Iceland and the southern half of Greenland. 2) In the period A.D. 1200-1400 there was somewhat more ice in the summer near Iceland and the southern half of Greenland. 3) In the period A.D. 1400-1600 the ice decreased in quanti ty in the summer near Iceland and the southern half of Greenland. 4) In the period A.D. 1600-1900 there were exceptionally large quantities of ice in the summer near Iceland and the southern half of Greenland. 5) In the period A.D. 1920-1939 there was hardly any ice. Bryson and Wendland (1967) have proposed tenta? tive climatic patterns for some of the late glacial and postglacial episodes in North America. This work was done on the basis of correlations between recon? structions of past airmass regimes and biotic regions and, specifically, shifts of the northern limits of the boreal forests, glacial advances and retreats, and palynological studies. They (Bryson and Wendland, 1967:294-296) have constructed six climatic episodes (Figure 1, Column 2), which are of direct importance to this study: (1) The Scandic episode, A.D. 400-900, SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 in ? * -1 ? 1 "J t 0: X 1 - > I T < ?* a. BI RN I DATE 1900 - 1800 - 1700 _ 1600 . 1500 _ 1400 - 1300 _ 1200 _ 1100 _ 1000 . 900 . 800 _ 700 _ 600 _ AD 500 CHANGES IN SEA LEVELS ALONG THE NORTH ALASKAN COAST ( BASED ON HUME 1965 AND MOORE I960 ) A X w 1 f (1 ^ w 1 ^v. i^S^ ^ v 1 ^ 1 1 ) y S*. s x / I 1 V ^4-^ ^\ > | S 3 ' !3 < ' cr UJ | a. ? . + 3 FEET | - 3 FEET RELATIVE WARMING & COOLING TRENDS FOR NORTHERN NORTH AMERICA ( BASED ON BRYSON E. WENDLAND 1967) _i < UJ t r O a 0 UJ z 0 u. 0 < a. 0 1 -z < _ j H < O Z O z < 0 WAR MER COOLER AMOUNTS OF ICE AROUND ICELAND ( BASED ON KOCH 1945 ) ' t- z 0 2 < t r < _? UJ 0 u . 0 u l "I \ 1- 1 1- 1 *V I 51 UJ a _ i K K INCREASE IN AMOUNT 0 - ICE FICURE I.?Summary of climatic data for the Arctic. a cold period, much like the Atlantic. (2) Neo-At? lantic, A.D. 900-1200. During this period, open water appeared in the Canadian Arctic Archipelago, the tree-line readvanced into the tundra, summer rains extended farther into the southwest. Glaciers disap? peared from the U.S. Rockies. These evidences suggest weaker westerlies and more meridional circu? lation. Westerlies increased about A.D. 1200, terminat? ing the Neo-Atlantic. The northern edge of the boreal forest once again retreated southward. (3) Pacific episode A.D. 1200-1500 is a period of increased westerlies, followed by an expansion of the circum- polar vortex and a return to patterns more like the sub-Atlantic (cooler). A far south jet stream in winter associated with blocking highs over northwestern Europe and a reduced summertime penetration of tropical air northward across the United States and into Canada is indicated by historical data. (4) Neo- Boreal episode, A.D. 1500-1850, had cooler summers and cold autumns. Glaciers once again formed as far south as New Mexico. T h e Neo-Boreal ended in the mid-nineteenth century with the well-documented climatic change which was associated with a return to strong westerlies in mid-latitudes. Modern Cl imat ic Condit ions The Alaskan Arctic coastal plain is virtually a desert from the standpoint of rainfall. Precipitation is slight and ranges from 25 to 228 mm (1 to 9 in) per year, with a yearly average of 100 mm (4 in). Most of this precipitation comes during the summer in the form of sleet and rain. T h e average annual snowfall at Point Barrow is about 6.5 cm (26 in). Summer, the period of greatest economic potential, lasts from June through August, during which the sun shines 24 hours a day. The summer temperatures range from - 7 ? C (20?F) to as high as 18?C (64?F). After July the sun begins to set briefly in the early hours of the morning and by mid-November the sun will not be seen again for 72 days. Winter temperatures range from - 4 2 ? C to 12?C ( - 4 5 ? F to 18?F) (U.S. Department of Commerce, 1963:27.) Windstorms are frequent, and when the wind is blowing, the temperature must be calculated with a 40 30 20 - 10 - 1 1 1 II 1 \ \ C H I L L \ \ \ l N D E x \ \ \ 4 \ XCHILLV \ lN0ExV \ 3 \ \ CHILL \ " \ ^ CHILL INDEX 1 1 1 INDEX X ^ 2 ^ V 1 iN**,,*sfc. 1 1 CHILL INDEX 5 _ 1 1 1 T* ? 50 40 30 20 10 10 -20 -30 -40 TEMPERATURE IN DEGREES FARENHEIT FIGURE 2.?Temperature-wind-chill index. (Chill Index 1 = Com? fortable with normal precautions. Chill Index 2 = Very cold, travel becomes uncomfortable on overcast days. Chill Index 3 = Bitterly cold, travel becomes uncomfortable even on clear sunny days. Chill Index 4 = Freezing of human flesh begins, depending upon degree of activity, amount of solar radiation, and character of skin and circulation; travel and life in temporary shelter becomes dis? agreeable. Chill Index b ? Survival efforts are required; exposed flesh will freeze in less than one minute. Chart is based on data from the Naval Arctic Research Laboratory.) INTRODUCTION TABLE 1.?Climatic data from Point Barrow, Alaska, based on observations made between 1921-1953 and averaged per month over this period (based on Naval Arctic Research Laboratory, n.d.) MONTH Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Year ly Average E =3 E X ta E -9 -12 -8 7 24 39 46 44 34 22 7 -4 16 TEMPI- I'A : i !!??! ( E 3 E c s : -22 -25 -22 -8 13 29 33 33 27 12 -5 -18 4 - C en ?r - . E I E + J C OJ o CTI E ro S- i- OJ OJ > ? . c OJ o CT! E i - S- Oi a. < -40 -42 -39 -25 -5 20 28 27 18 -4 -20 -33 -45 3H E =3 E X (*3 s : OJ 3 O S- c OJ o > E ? t .2 .1 .1 .1 .1 .3 .9 .8 .5 .5 .3 .2 .1 WIND >, ? M i - O O 0 ? OJ i- > 01 Q . OJ c n i/i ra r- > E ?t 11.0 11.3 10.9 11.5 11.8 11.4 11.8 12.7 13.7 14.0 12.5 10.9 12.0 c o 0> T - c n + J ID u S- OJ OJ s - > - r - < Q ESE ENE NE E NE E SW E ENE NE NE ENE NE CHILL FACTOR E E 3 3 E E C X z: ?: 4 5 4 5 4 5 3 4 2 2 1 2 1 2 1 2 2 2 2 3 3 4 4 5 3 3 chill factor, which will lower the effective tempera? ture (Figure 2). The wind plays an extremely im? portant role in both the success and failure of eco? nomic activities; it affects the ability of Eskimos to hunt and travel and it also regulates the occurrence of game animals. Winds are usually from the north? east. The average wind velocity is about 19 km/hr (12 mi hr) (U.S. Department of Commerce, 1959:12). Table 1 is a composite chart showing temperatures, precipitation, wind velocity, wind direction, and the corresponding maximum and minimum chill factors for the period of 1921-1953. Site Description The Walakpa site is located on a 6.1-m (20 ft) high bluff on the Chukchi Sea, 19.3 km (12 mi) south of Barrow Village (Figure 3) . T o the south of the site is a sandpit which separates Walakpa Bay from the sea. North and east of the site is the low tundra, which is composed of the physiographic features previously described (pp. 4-5). The site contained 15 surface house structures of varying age, most of which were historic (Figure 4). These houses were oriented in two lines, one running north-south along the sea coast and the other east-west along the bay side of the bluff. The midden was situated at the point of the bluff between the two house rows. It is reported that there were once more houses between the midden and the ocean, but they have been washed away by the ocean eroding the bluff. There is no evidence in the area for prehistoric ice cellars. A single ice cellar, which was constructed in 1946, is located between the A area of the mound and the B area. This ice cellar cuts through both Thule and Birnirk occupation levels. Strat igraphy The Walakpa stratigraphy has been defined on the basis of the 1968 test trenches and the 1969 excava? tions. The development of the natural stratigraphy of the site has been worked out in detail by Duguid (1971). At least one occupation level occurs in the upper portion of the Gubik formation; however, the rest of the mound has a different origin. Several of the processes to be described below are probably the same as those that contributed to the formation of the Barrow unit of the Gubik. Excepting the cultural debris, I consider the mound a continuation of the formation of the Barrow unit. The first Walakpa stratigraphic process to be de? scribed is the eolian deposition. Much of the upper Barrow unit and the cultural stratigraphy are sepa? rated by windblown sands (Plate 2b). The origin of these sands is not completely understood; however, the Arctic coastal plain is particularly favorable for eolian erosion and deposition (Black, 1951:92). This is because of high winds, lack of topographic wind? breaks, and an abundance of silt and sand. If it were not for permafrost impeding surface runoff, and thereby making it possible for vegetation to grow in spite of small amounts of precipitation, this area would become a desert with shifting sands (Black, 1951:93). The stratigraphic units, which are of prime im? portance to this study (5, 6, and 7, pp. 13-14), can be differentiated by their color and sand composition. A R C T I O C E A N 0 0 5 5 10 10 IS 20 15 25 20 25 KILOMETERS MILES WAINWRIGHT \ NUWUK (BARROW VILLAGE) UTKIAVIK BIRNIRK ^ p Q | N T KUGOK \ BARROW PLOVER ISLANDS FIGURE 3.?Eskimo archeological sites in the vicinity of Point Barrow, north coast of Alaska. 10 SITE DESCRIPTION 11 ARCTIC OCEAN WILL ROGERS - WILEY POST MONUMENT FIGURE 4.?Topographic map of Walakpa site, showing mounds A and B. It was not possible to determine what caused these differences, but it is thought that they reflect sands of different origins, compositions or mixtures rather than soils as described by Britton (1967:35). The immediate sources for sand at Walakpa are (1) the ocean beach and large sandspit of Walakpa Bay, (2) the beach of Walakpa Bay, and (3) the tops of polyg? onal mounds stripped by winds. The parent forma? tion of most of these eolian sands is probably the Gubick formation consisting of fine-grained, white to yellow and tan quartz, and chert sands. The change in color of these deposits at Walakpa is perhaps indicative of climatic fluctuations and changes in the sea level (page 7), both perhaps reflected by changes in the mean annual wind direc? tion and velocity, and severe storm frequencies. A drop of 2 meters in the sea level would cause Walakpa Bay to become an inward draining basin and the alluvial sediments, which would normally be dis? bursed along the beaches to the northeast by ocean currents and ice floes, would then accumulate at the bottom of the bay. These sediments would be similar to the gray lake sediments as described by Brown (1965:44) . Furthermore, if there were a 2-meter drop in the sea level, areas now under water in Elson La? goon and Dease Inlet would be exposed to wind action and would contribute to eolian deposition throughout the entire Point Barrow area. When the bay was once again open, after a rise in the sea level, there would be an extremely high amount of sediment that would be cleaned out of the basin and be deposited in an alluvial fan at the mouth of the Walakpa River. This load of silt would sub? sequently be disbursed along the coast by ocean cur? rent and ice action. If this rising of the sea level were concurrent with a warming trend, there would also be an increase in the silt load of the Walakpa River, which would be highly variable in color and compo- 12 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 sition caused by the differential draining of inland lakes and bogs as the permafrost recedes. Additional sources of sand would be produced by solifluction causing numerous rivulets of sludge to descend from the face of the cliffs along the beach, which would be composed of variable types and amounts of sand (Rex, 1964:388). It should also be noted that waves reach the cliffs only during severe storms and erosion is especially pronounced during southwest storms (Rex, 1964:388). Also of importance for determining stratigraphic breaks is the occurrence of thin gravel layers. The cliffs in the Walakpa area are covered with a thin lens of gravel, which at first glance would be identi? fied as the result of solifluction. Gravels of this nature covered the Walakpa midden; however, no artifacts were found among the gravels. It was also observed that these lenses were interspersed through? out the stratigraphic column of the site. These gravel layers were apparently formed on the tundra surface after a period of Eskimo occupation and then covered by subsequent occupations. It is Duguid's (1971:4) contention that these gravel lenses were formed by a strong onshore wind pushing pack ice against a narrow margin of landfast ice, which is pushed up to the cliff forming a ramp. An offshore wind would subsequently push the pack ice back out to sea. During later storms, waves would push beach sand and gravel up over the ice ramp and onto the beach cliff. Over a period of years, erosion and precipitation would spread the gravel out in a thin layer across the surface of the cliff. T h e stratigraphy of the two major excavation units, A and B of the Walakpa site, is summarized below (Plate \a,b,c,d). Profiles were drawn of all trench walls. Changes in the geological deposition were identified and grouped into ten major strati? graphic units (Figure 5). The term "unit" designates discrete, natural accumulations of soil and does not relate to the cultural subunits which are described on pp. 15-16. The units are identified on the basis of their color, composition, and position. A AREA FIGURE 6 Unit 1: This uppermost layer consists of recent sod, gravel, and sand. It varies from 5 to 3 cm (0.5-0.1 ft) in thickness and is light gray in color. Artifacts are all datable to historic and modern times. Unit 2: A thin lens 6-3 cm (0.2-0.1 ft) thick of beach pebbles and sand. This unit is gray in color. It probably originated through the ice ramp process. There were no artifacts found in this unit. Unit 3: This unit is from 39 to 18 cm (1.3-0.6 ft) thick?a gravel lens containing artifacts of Nuwuk and Thule ages. It is composed of gravel and sod, along with sand and is gray-brown in color. It is probably the result of the construction of a house that was built during the Nuwuk period in Unit 4 (Plate 3a). Unit 4: A lens averaging about 60 cm (2 ft) thick of gravel, sand, and sod. It is coarse and gray-brown N955 AREA 5.0' 1.5 M FIGURE 5.?Schematic profile showing correlation of the stratigraphy of the A and B areas of the Walapka site. SITE DESCRIPTION 13 .UNIT I UNIT 5.0' ( 1.5 METERS) 1*^.^1 NUWUK ^Z^*3 EARLY THULE rzzzi LATE THULE EOLIAN OEPOSITS HISTORIC BIRNIRK GUBIK FORMATION ^q WALAKPA, DENBIGH FIGURE 6.?A area stratigraphy at the Walakpa site. in color. It is also made up of elements from Unit 5 that were disturbed by the construction of a Nuwuk age house. Unit 5: A 92-61 cm (3-2 ft) deposit of sand, clay, sod, and gravel. It is reddish to yellow-tan in color. This is the unit that contains the Late Thule occupa? tions. Unit 5 contains 20 independent levels of which 10 are occupation levels and 10 are sod and sand levels that separate each occupation. At the west end of the east-west profile test trench, these levels are all slumped together and have been completely soaked with oil from a sea mammal. This is probably the result of a historic Eskimo staking out a seal or walrus skin on the ground surface. This is the point where these levels emerge from the eroded midden. It ap? pears that this area was disturbed by both erosion and later Eskimo activity. Consequently, artifacts and cultural debris from this area have been excluded from the analysis. T h e cultural levels can be readily recognized. They are 6-3 cm (0.2-0.1 ft) thick. Bones, artifacts, sod, cultural debris, and gravel were found in these units. The occupation levels are dark brown to gray in color. Each is separated by eolian sands, which are yellow to tan in color. Unit 6: Unit 6 is black grading into yellow in color. It is 39-18 cm (1.3-0.6 ft) thick containing eolian sand interspersed with organic units. Artifacts, diagnostic of Early Thule, were found in one of the organic subunits; other subunits contained artifacts but none was diagnostic. Unit 7: Unit 7 contains logs and bones. It has at least one occupation level containing artifacts that are Birnirk in age. It is yellowish to brown and black in color, measuring an average of 92 cm (3 ft) thick. Unit 8: Thin lenses of fine gray and tan sands. These units are probably a continuation of the Gubik formation post-Walakpa Phase Denbigh occupation. No artifacts were found in this unit. Unit 9: A lens of organic matter and sand. It is black in color. This unit contains a distinct occupa? tion level with artifacts typical of the Arctic Small Tool Tradition. It has been identified as Walakpa Phase Denbigh. Unit 10: This consists of yellow-tan, fine-grained sands with interspersed gray, loamy streaks. It also contains orange and yellow clays, which are prob? ably limonite. Unit 10 represents the upper part of the Barrow unit of the Gubik formation. B AREA FIGURE 7 The stratigraphy of the B area has five major units, all of which contain one or more occupation levels. These units are not as well-defined as those in the A area, but it is possible to identify and correlate them on the basis of color and cultural compositions. Therefore, the unit numbers given below also cor? respond to the units of area A. Many of the units found in the A area did not occur in the B area. Unit 1: This is the current sod level, with the typical band of gray sand and clay. This unit contains artifacts of historic age. Unit 5: This unit is a sandy loam which is yellow- tan in color. Its upper portion, as well as some areas of the first unit, have been disturbed. This unit con? tains a single, black organic level which is 3 centi? meters (0.1 ft) thick in most areas. There is little in the way of bones and cultural objects in this level. Artifacts indicate that it is Late Thule in age (Plate 2c). Unit 6: This unit is quite thick, measuring up to about 1 meter (3 ft) thick. It consists of a yellow- 14 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 N950 5 . 0 ' (1.5 METERS) 1///I LATE THULE | C ^ y j EARLY THULE HISTORIC GUBIK FORMATION BIRNIRK FIGURE 7.?B area stratigraphy at the Walakpa site. gray sand, which is interspersed with six organic levels which are gray to black in color. These lenses are usually 6-3 cm (0.2-0.1 ft) thick and contain occupational debris of Early Thule age and cultural affiliation. Unit 7: This unit is quite complex. It is made up of nearly solid occupational debris of Birnirk affilia? tion. It consists of wood chips, baleen, bone, and logs, all permeated with oil. The unit has been di? vided into three subunits on the basis of the midden structure. It is felt that these three levels are quite distinct and they do separate when traced out from the midden. Units 8, 9, and 10: This is the upper part of the Gubik formation. It consists of yellow-tan, fine? grained sands with interspersed gray streaks. There are slight traces of an occupation level in the upper portion of the unit, but it was nearly removed when the Birnirk house was constructed. I believe that it was Denbigh in age. The correlation of the stratigraphic units of the A and B areas can be accomplished with the aid of the test trench segment stratigraphy (Figure 5). Unit 1, the recent sod, is, of course, quite readily recognized throughout the site. Units 2, 3, and 4 of the A area are apparently unique to that portion of the site. This is probably because of the manner in which they were deposited and their proximity to the beach. Unit 5 of the A area can be traced through the site and becomes the second layer (Unit 5) in the B area. The number of Late Thule levels in the A area is re? duced to only one in the B area. Unit 6 contains the Early Thule cultural levels. It is identified as the third layer (Unit 6) in the B area. Unit 7 (in the A area) contains Birnirk artifacts. Archeolog ica l Excavat ions of the W a l a k p a Site 1968 EXCAVATION The archeological excavations at Walakpa utilized a grid system of 1.5 meter (5 ft) squares, which was superimposed on the site area. A north-south base line was established, using true north as its orienta? tion. The base line was assigned the coordinate iden? tification of E 1000. The southeast corner of the base of the Will Rogers-Wiley Post monument is a reference point for the base line (Plate 2a). The corner of the monument was identified on the grid as N 1096.4, E 1000. Three brass caps embedded in concrete were placed along the base line. These brass caps were located at N 1000, E 1000; N 1050, E 1000; and N 950, E 1000. T h e brass cap at N 1000, E 1000 was used as the main datum point, with all elevations and grids referenced to this point. The grid point identification system was set up so that each 1.5 meter (5 ft) grid had four main refer? ence points, one at each corner. The corners were referenced to a point that was identified as N 1000, E 000, W 000, and S 000, a point southwest of the site in the Chukchi Sea. This point was used so that all of the site would be in the northeast quarter of a 2- axis grid. This was found to be quite useful, as all grids were located only on north and east lines. This system eliminates the confusion of four directional locators. Moreover, with this system, all artifacts can be referenced to each other without having to trans? pose directions. From this figure of a sample grid, one notes how to determine the exact location of the grids and how the positions of artifacts were plotted. T h e artifacts' location (N 943.9, E 804.2) were thus plotted hori? zontally on the site map without any other identifica? tion. A second system was used for rapid grid identifica? tion. This system was devised primarily for the ease of handling paper work. This system utilized both letters and numbers. The letters were assigned to grids located in the north-south plane, with the letter "A" being completely off the site area. All grids SITE DESCRIPTION 15 along the east-west axis of the site between N 895- 900 were assigned the letter "A"; all grids along the east-west axis N 900-905 were designated by the letter " B " and so on. The numbering system identified grids in their east-west position along the north-south axis. The numbers ran from 1 to 100, with number 1 being to the west of the site. For example, grid unit J-19 has the coordinate numbers N 940-945, E 935-940, while unit J-20 has the coordinate numbers N 940-945, E 940-945. Elevations during the 1968 field season were deter? mined by positions above or below the main datum point. The elevation of an artifact was calculated from its depth below the northeast corner of the ex? cavation square. The elevation of the northeast corner of the square was calculated according to its position in relation to the main datum point. The grids chosen for excavation in 1968 were those along a single east-west axis (N 940-945) and those along the two north-south axes (E 880-885 and E 920-925). These were chosen in order to provide a test trench that would transverse both axes of the major middens. Cross-sectional trenches were exca? vated into all other house structures to determine their age and relationship to the middens. Every other grid along these axes was excavated. The other grids were left intact in order to work out detailed stratigraphy of all four faces of the exca? vated grids. By excavating in this manner and inter? polating the stratigraphy between the excavated grids, a complete cross-section of the stratigraphy of the major middens was determined. The interpretation of the east-west test trench stratigraphy also provided evidence for the relationship of the stratigraphy of the two middens. The grids were excavated in arbitrary levels of 12 centimeters (1 ft). This depth was used to speed up excavation, but when occupational stratig? raphy was determined, natural living surfaces were excavated. All levels were troweled, and when pos? sible, the dirt was dried and screened. All structural features, artifacts, and other debris were mapped and recorded before they were removed. 1969 EXCAVATION In 1969, excavations were resumed a month earlier and even though our work was critically hampered by below seasonal temperatures and numerous snow storms, good results were forthcoming. Those areas labeled "A" and "B" (Figure 8) were chosen for excavation because the stratigraphic continuum de? fined by the 1968 test trenches indicated that they would be the most productive areas. Each area was excavated according to the grid system established in the 1968 excavations. Occupation floors, which were defined by the test trenches, were excavated as single units, exposing entire occupation surfaces. These sur? faces were left intact until all features were mapped and photographed. By using this method, specific activity areas were exposed; moreover, the relation? ships between these task specific-activity areas and other features such as houses, drying racks, and trash middens were observed. All cultural features including bone scraps and flaking debitage were collected, recorded, and ana? lyzed. Pollen and carbon samples were collected from each level. Several pollen profiles were made in each stratigraphic column, as well as a composite sample from each occupation surface. Surface pollen samples were collected from the site and the sur? rounding tundra. The major difference between the archeological techniques used in 1969 and those used in 1968 was the utilization of local H.I. (height above main datum) stations for elevation control. Two major H.I. stations were erected. The H.I. stations consist of three concrete reinforcement bars, which were placed in the ground, forming a small triangle with each reinforcement bar being 0.3 meter (1 ft) from the other two. The stations were placed on the highest ground nearest the area of excavation. A string was tied around the rebars at 61 centimeters (2 ft) above zero elevation. All artifacts, structural features, debris, and the upper surface of each occu? pation level at grid intersection points were recorded as to their positions below the H.I. datum point. This allowed fast, accurate recording of data without the use of a line and line level. Further, with one con? sistent point of elevation for each unit, all artifacts in that unit could be directly compared with the ele? vation of every other artifact from all levels in the unit. Excavations of area A, the largest excavation area, were terminated at a depth of 1.2 meters (4 ft) below the present ground surface. At this lowest depth, an Early Thule horizon was partially excavated. Over? lying the Early Thule level were eight distinct levels of Late Thule occupation. Above the Late Thule levels were two levels of historic Eskimo occupation, but these levels were, for the most part, disturbed by looting (Plate 3a). In addition, the 1968 test trenches indicated that the Thule occupation may be under? lain by Birnirk, Norton, Choris, and Arctic Small Tool levels. As yet, we have not systematically exca? vated these occupations. In area B, an historic level, one level of Late Thule, 16 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 :?1 \ N 25 50 FEET 10 METERS FIGURE 8.?Contour map of the A and B areas of the Walakpa site. Contour interval equals one foot (0.3 meter). six levels of Early Thule, and three Birnirk levels were excavated. Below the Birnirk occupation, there is evidence that an Arctic Small Tool level may have once existed, but it has been nearly destroyed by the construction of the Birnirk house. ADDITIONAL EXCAVATIONS IN THE WALAKPA BAY AREA Test excavations were made in two new sites found in the Walakpa Bay area during the 1969 season. The Coffin site, located on a high bluff about a mile east of Walakpa on the north beach of the bay, yielded pottery and ground slate in unquestionable associa? tion with typical flint artifacts of the Arctic Small Tool Tradition. Since a level consisting of this same assemblage was found in the main Walakpa excava? tion, we now have considerable control over this hori? zon. This level has been dated by C-14 at 1450 B.C ?520 (GAK-2299), and is designated as the Walakpa phase of the Arctic Small Tool Tradition. Appar? ently this assemblage represents a late Denbigh hori? zon, transitional into the Choris horizon. Facing the Coffin site from across a large creek basin is another archeological locality known as the Kahraok site. From this small surface site were re? covered projectile points similar to the lanceolate points of the Old Cordilleran tradition. Other artifacts found were burins, burin spalls, and large blade tools, all distributed around a cache of 32 cores. These artifacts are similar to both the Akmak assemblage and to the Driftwood Creek com? plex (Anderson, 1970; Humphrey, 1970). The surface reconnaissance of the Point Barrow area produced what appeared to be six additional Arctic Small Tool sites, several Norton sites and sev? eral uncharted Eskimo sites. Based on a composite interpretation of these several sites and both sum? mers' excavations, a tentative scheme for cultural suc? cession of the Walakpa site areas includes the follow? ing horizons, from latest to earliest: Modern Eskimo, SITE DESCRIPTION 17 Historic Eskimo, Late Thule, Early Thule, Birnirk, more extensive excavations are required for the Ipiutak, Norton, Choris, Walakpa Phase Denbigh, earlier material. Only the Birnirk, Early Thule, and Denbigh, and Kahraok. Late Thule occupations that have direct bearing on Although many of the above early levels are impor- the problem of the origins of Thule culture and tant to an understanding of American Arctic prehis- Thule culture development in the Point Barrow area tory, they will not be considered in this monograph; will be analyzed and interpreted here. Analysis and Classification of Artifacts H U N T I N G A N D F I S H I N G E Q U I P M E N T Harpoon Gear HARPOON HEADS Except where noted, the classification used in the following treatment of harpoon heads is that em? ployed by Ford (1959) and Collins (1937c). Several new types of harpoon heads have been identified and given names that are compatible with the already existing classifications. BIRNIRK OPEN SOCKET.?Fourteen specimens of the Birnirk open socket variety were found in the Walakpa excavation (Plates la, \bd-h, 37a, 49a-d). Unlike the Birnirk open socket harpoons found by Ford at the Birnirk site, which were all of antler, four of the Walakpa specimens are made of ivory and ten are made of antler (Ford, 1959:75). All specimens have barbs placed on a lateral edge with an inset side blade facing the barb. On nine specimens, where the location of the socket could be determined, it was found that the spur occurred on the right edge of four and on the left edge of five. Three have the open socket flanked by two lashing slots, while six have only one lashing slot. The rest of the base is grooved for hafting. All line holes are round. On four complete examples, "the range of the lengths of these harpoon heads is from 14.4 to 8.8 cm, with an average of 11.9 cm. This compares to a range length of 13.5-7 cm and an average of 10.5 cm on Ford's Birnirk open socket harpoon heads (Ford, 1959:75). Measurements from the upper edge of the line hole to the tip of the harpoon point range from 8.75 to 6.8 cm, with an average of 7.7 cm. The ivory specimens are all shorter than the antler specimens, which have a range of 7.1-6.9 cm from line hole to tip of harpoon point. Decoration is simply straight incised lines that di? verge from the median ridge to approximately half? way between the line hole and the point of the harpoon. One of the antler specimens has two short converging lines placed over the line hole and on the socket surface; the area between these lines is also in? cised. A unique specimen has two scallops that are similar to nonfunctional side blade slots. These occur on the edge of the outside spur. On either side of the spur are incised crescent-shaped lines. None of the ivory specimens contain any decoration above the line hole. It is also interesting to note that the location of the sockets on the antler specimens are not determined by the cancellous structure of the antler, for on five specimens the socket is carved into the hard surface. On four others it is carved into the cancellous structure. The inset side blades are made on small, thin semi- rectangular flakes. Three of these flakes are made from a gray chert that comes from the Siksikpuk for? mation in the Brooks Range. One is made from a clear chalcedony. NAULOCK OPEN SOCKET.?Six specimens of the Naulock open socket variety were found, all made from antler (Plates 6b, lb, c, 15a). T h e range of lengths of these harpoon heads is 9.2-7.6 cm, with an average of 8.4 cm. On three specimens, three gracefully carved points which make up the spurs are carved (Plates 7c, 15a). These are placed directly below the sockets on the left edge of the harpoon. One specimen has only two points, one much larger than the other (Plate lb). The larger point has four lateral points. Unlike the specimens reported by Ford (1959:79), one has the open socket placed on the hard surface rather than on the cancellous side of the antler. Three of these harpoons have the open sockets flanked by two lash? ing slots. One has only one lashing slot. They have round line holes, which are beveled toward the base of the harpoon to accommodate the harpoon line. The blade slit is located in the same plane as the line hole. The decoration on these specimens consists of two converging lines located on both sides of the harpoon point. T h e lines begin at the lashing slot and con? tinue beyond the blade slit. The incised lines do not reach the edge of the harpoon and do not serve as ornamental barbs. Ornamental barbs appear on the li ANALYSIS AND CLASSIFICATION OF ARTIFACTS 19 edges, but are simply the result of the depression carved into both lateral edges to form nonfunctional side blade slots. TUQUOK OPEN SOCKET.?Two antler specimens were found which resemble Ford's (1959:79) Tuquok type (Plates 6a, 15b). Although in certain aspects they are quite different, I will not attempt to describe a new type with only two specimens, and consider these examples as a variety of the Tuquok type. These specimens are small compared to Ford's and measure 9.4-8.7 cm in length, while Ford's measure on the average of 10-16 cm in length (Ford, 1959:79). The spur of this type is not bifurcated and rather than a vestigial third prong, one has four distinct prongs symmetrically placed below the open socket (Plate \5b). The other has only two prongs (a third prong is broken), which are also placed directly be? low the socket (Plate 6a). The open sockets are flanked by one lashing slot. One has a deeply cut groove running from the slot across the harpoon to the socket. The other has the groove cut completely around the harpoon. Both examples have side blades located on both edges of the blade. The side blades are made from chert. The line holes are round and beveled towards the base of the harpoon. Other than the four basal spurs, the decoration is simply two converging lines that start between the line hole and the edge and nearly three-fourths the length of the harpoon blade. TASIK OPEN SOCKET.?Only one example of the Tasik type was recovered in the Walakpa excavations (Plate Id). This specimen is 8.6 cm in length, which compares to the range of 13-7.5 cm reported by Ford (1959:82). This specimen, like Ford's, is made of antler. The harpoon head has a single asymmetrically placed lateral spur. The inside of the spur is scalloped in the shape of a nonfunctional point placed on the upper margin of the spur. It is oval in cross-section. Both sides are slightly constricted by the presence of nonfunctional side blade slots. Ford does not men? tion the presence of these side blade slots on any of his specimens. Two lashing slots are present on either side of the socket. The line hole is rather small and placed in the same plane as the blade slot. It is slightly beveled towards the socket to accommodate the harpoon line. Ornamentation consists of incised lines that run from the line hole to the blade slot. Two incised lines are located along both edges below the nonfunctional side blade slots and extend almost to the base of the harpoon point. It also appears that the small, nonfunctional point on the spur is for ornamentation. Sicco O P E N SOCKET.?Three examples of the Sicco Open Socket type were recovered at Walakpa (Plates 58b, 63c). Only one specimen was complete and its length is 8.9 cm; this is slightly smaller than those described by Ford (1959:83). Its spur has one point with five facets placed lateral to the socket. All of the examples have two lashing slots, which flank the open socket. Blade slits occur in the same plane as the line hole and socket. In cross-section these harpoon heads have six facets. They are slightly constricted at the line hole and have faceted triangular-shaped prongs on either side of the blade slit. Decoration, which occurs on only one specimen, consists of elongated triangles placed between the line hole and blade slot. The center areas of the triangles have been filled with incised lines. Neither specimen has vestigial side blade slots, such as those reported by Stefansson from Point Barrow (Ford, 1959:85). NATCHUK OPEN SOCKET.?Ford believes that the Natchuk open socket evolved from the Birnirk open socket by the elimination of stone side blades and the reduction of the multipronged spur to one point (Ford, 1959:83). He further believes that it is a type that represents a step in the development of the Thule ii harpoon head (Ford, 1959:83). Evidence from the Walakpa site would suggest that this type occurs very late in the Birnirk sequence and is con? temporary with the Birnirk type harpoon head in the uppermost Birnirk level at Walakpa. The differences between the two types of harpoon heads are that the Natchuk type is usually more delicate, less elaborate and does not have inset side blades. Three ivory specimens of the Natchuk open socket were recovered at the Walakpa site, two from levels that represent middle period Birnirk occupation (Plate 49e, /), and one from an Early Thule level (Plate 58a.) Of those from the Birnirk level, one is a finished specimen while the other is a nearly finished preform. The two Birnirk specimens may demon? strate an early period of Natchuk development, as they are more closely related to the Birnirk open socket variety in that they have multipointed spurs and the barb is placed at the center of the blade. As Ford (1959:83) noted, Natchuk specimens are smaller than the Birnirk type. These two measured 8.9 and 8.8 cm in length, as compared to the average of 10.5 cm for the Birnirk type. The spurs have two points and are placed on the opposite side of the lateral barb. The line hole in the finished specimen is round while in the unfinished specimen it appears to have been intended to be triangular. The unfinished specimen has three nonfunctional slots to replace the side blades, while the other con? tains two scallops, unlike the nonfunctional slots. One specimen has two lashing slots, while the other has 20 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 only one slot with a cut groove extending three- fourths of the way around the diameter of the har? poon point. The unfinished preform provides details on the manufacture of the open socket. Three cuts are made: two long cuts run parallel to the long axis of the harpoon and one short cut placed at a right angle across the top of the two parallel cuts. This leaves an uncut section, which is chipped out; then the rough surfaces are ground off. Decoration consists of converging incised lines starting at the lowest por? tion of the nonfunctional slot and/or scallops, which end and do not meet above the lateral barb. On the complete specimen an intricate scallop has been carved under the bark. The Early Thule Natchuk specimen from the A area is extremely interesting, as it more closely re? sembles the Thule 11 type than the Birnirk open socket variety (Plate 58a). It has only a one-pronged spur, which is laterally placed. It retains the non? functional side blade slot. Unlike Ford's specimen, however, and more nearly like Thule 11, the lateral barb has been placed near the tip of the harpoon rather than at the center of the blade. It has a small round line hole with beveling between it and the socket for the line. It has two lashing slots. The specimen is 9.15 cm in length. Its decoration is con? fined primarily to two converging lines above the line hole with two matched sets of inward pointing lines which do not meet. T H U L E II OPEN SOCKET.?The Eastern Thule 11 is the most distinctive of the Thule types. Eleven specimens were recovered at Walakpa (Plates 49g, 58c, 59a, b, 63e, 15b, 118a). Three of these were complete and the rest were damaged. Seven of the harpoon heads were made of ivory, while the rest were of antler. The range of lengths is 8.9-8.5 cm. This compares to a range of 15-8 cm from Birnirk, with an average of 9.5 cm (Ford, 1959:86). These harpoon heads can be characterized by hav? ing two symmetrically placed lateral barbs, open sockets and single-pointed spurs. Three slight varia? tions appear in the lashing techniques. Six specimens have two lashing slots cut into either side of the socket and converge toward each other. A groove is cut between them on the opposite side of the open socket. One specimen has two drilled holes on either side of the socket. Two specimens have only a groove cut around the diameter of the harpoon head from one side of the socket to the other. Four specimens have triangular line holes, while the others are round. One specimen has two round line holes. This appears to be accidental, as the socket cuts into the lower line hole and the second hole was subsequently drilled. Occasionally on the outside edges of the body of the harpoon head there appears a pronounced ridge. The ridge runs from slightly below the line hole to the tip of the spur. It appears that these were carved to give greater support for the lashing slot. Decoration ranges from no modification to a straight line carved from the line hole extending up the center of the harpoon to the tip. An incised, up? side down Y occurs over the line hole on one exam? ple, while another has a very simple triangle carved above the line hole. T h e final example has a triangu? lar-line hole, the top of which terminates with a short incised line. Both the line hole and the incised lines are inside two converging lines that run from the lashing slots to about two-thirds the length of the harpoon head. One, possibly two, specimens of Thule 11 harpoon heads occurred in the Walakpa-Birnirk occupation. The first is a broken ivory specimen (Plate 49g). In most respects it is similar to those described by Ford (1959) and Mathiassen (1927a), but it seems to have many decorational features that are more similar to Birnirk than to Eastern Thule and is identical to a specimen from Kurigitavik (Collins, 1940:562). The length of this specimen is 7.25 cm. On the specimens from Birnirk, Ford (1959:86) reports that they have single pointed spurs. On the Walakpa-Birnirk speci? men there is a single pointed spur, which has been faceted into a pentagonal cross-section, with one small but very pronounced spur placed slightly above the main spur and to the opposite side of the socket. The open socket is flanked by two lashing slots with con? necting lashing grooves. T h e line hole is triangular. Two barbs are placed symmetrically at the tip of the harpoon head but are broken. The decoration is quite elaborate. A set of two incised lines are placed above the line hole, with one starting below the hole and meeting above the line hole where they form a single line and continue up the blade of the harpoon. A triangular-shaped groove has been cut between the line hole and the incised lines, which act as a decorative extension of the tri? angular line hole. Incised lines are cut into the blade from the ends of the barb cuts. Below the barbs, and symmetrically placed to either side of the line hole, are two gracefully carved vestigial barbs. These barbs have incised lines cut into the blade from the termination of the barb cuts. T h e faceted spur is the final form of decoration. This particular harpoon head, even though it is considered here as a Thule n, has features that are characteristic of both Thule and Birnirk cultures. It is a good example of a harpoon head belonging to the time when Thule types were beginning to emerge ANALYSIS AND CLASSIFICATION O F ARTIFACTS 21 from Birnirk types. Therefore, it is considered ?n early form of Thule n. The second example, which may be a Thule n harpoon head found in the Birnirk occupation at Walakpa, is a broken tip (Plate 376). Although there is no way of making a positive identification, what? ever type it may have been would have had two symmetrically placed lateral barbs. The only type of harpoon head that occurs with this type of barb arrangement is the Thule n. T H U L E HI OPEN SOCKET.?Eastern Thule m is represented by only two antler specimens (Plates 63 b, 80a). In regard to general characteristics, they have no side barbs but are slotted for end blades. The slots are cut so the blade is set parallel with the line hole. They have round, medium-sized line holes and both sides have lashing slots. They have open sockets with lateral spurs. Mathiassen's (1927a: 25) specimens range in length from 10.4 to 7.4 cm, while those re? covered at Walakpa are 11 and 6.7 cm in length. The smaller specimen is decorated by two incised lines, which run from the lashing slot along both edges for a distance of approximately two-thirds the length of the point. In the manufacture of the socket on the larger specimen, the line hole was accidentally damaged. A small ivory wedge was placed in a slot behind the line hole. This wedge kept the open socket separated from the line hole. T H U L E IV CLOSED SOCKET.?Thule iv specimens, as described by Mathiassen (1927a: 25), have closed sockets, with the blade slit cut at right angles to the direction of the line hole. Eastern Thule iv is not reported by Ford from the Point Barrow area. At the Walakpa site, four, or possibly five, specimens of this type were recovered. Three of these were in unques? tionable association with the Birnirk occupation. There is no doubt about their occurrence in situ in the Birnirk occupation. The implication is that, like Eastern Thule n, Eastern Thule iv also has its imme? diate origin during the Birnirk period and is indeed fully developed before the end of the Birnirk period. One of these closed socket specimens (Plate 15c) is made of ivory. Its total length is 8 cm. The prongs have a thick diamond-shaped cross-section, while the base has been faceted into ten sides. The line hole is round and in comparison with other specimens from this level is rather large. The hole is placed in a shallow concave surface on each side of the harpoon, with a shallow bevel extending towards its base. The end blade slot is placed at a right angle to the line hole. It has a single-pointed spur occurring below the socket. Other than the faceting at the base of the harpoon point, there is no decoration appearing on this specimen. The Second specimen of Thule iv (Plate le) comes from the test trench in a Birnirk level. This speci? men, measuring 6.8 cm, is smaller than the example described above. The prongs have a diamond- shaped cross-section and the base has eight facets. The line hole is round and is proportionally smaller than the first specimen. It is placed in a shallow concavity with a bevel leading back to the base of the harpoon. Unlike other specimens from Walakpa, this one has a slight ridge at the front of the line hole. The end blade slit is placed at a right angle to the line hole. The faceted spur is placed below the socket. A por? tion of an ivory foreshaft still remains wedged into the closed socket. A third specimen from the A area was also found in association with Birnirk type artifacts (Plate 6c). This specimen is not complete and no overall length could be determined. The blade slit is situated at a right angle to the line hole, with the prongs being diamond- shaped in cross-section. The line hole is relatively small, and it is placed in a concave Surface, which occurs on each side of the harpoon. The socket is also relatively small. The spur is placed at a right angle to the socket. The spur is not faceted as in the first two specimens and there is no decoration. The fourth specimen of Thule iv (Plate 63d) made of antler was found in the Early Thule unit of the B area. The prongs have been broken, but it is possible to determine that the blade slot was at a right angle to the line hole. The specimen is broken and no measurements could be made. The line hole is round but it is not placed in the concave portion of the harpoon. The cross-section of the harpoon is almost rectangular, which is quite different from the Barrow or Nuwuk types and fits the faceted cross-section of the other Eastern Thule types from Walakpa. THULE na.?I have here designated four harpoon types as Thule na, b, c, d. These new designations are used to remain parallel with the Thule-n type harpoon head, of which these are simply variations. They do not imply any temporal ordering. Only one broken specimen (Plate 97a) of this variety was recovered at Walakpa. Larsen and Rainey (1948) recovered one of this type in their Western Thule house at Point Hope. In their general characteristics, they are open socket harpoons, which are very closely related to both Thule ii and the Natchuk types. The singular dif? ference is that they are similar to Thule n, but have only one lateral barb rather than two. The specimen from Walakpa is broken; however, it has a triangular line hole, no decoration, and is made of ivory. T H U L E nb.?This type is also a variety of Eastern Thule II (Plate 936). The only difference is that these 22 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 specimens are cut for an end blade slot. In all other features they are identical to Thule 11. Only one specimen was recovered at Walakpa. It measures 7 cm in length and is made of ivory. It has an open socket, two lashing slots, and a single-pointed spur placed at a right angle to the socket. Two lateral barbs were present but one is broken. The specimen has a round line hole. Decoration is a simple incised line running from the line hole to the end blade slot. T H U L E IIC.?This variety of harpoon head was found in the St. Lawrence Island Punuk assemblage and is called type iv(a) by Collins, (1937c:211) (Plates 816, c, 87a, e). It is closely related to Thule 11, but differs in that it has multiple sets of barbs. Most of these specimens come from the Late Thule levels at Walakpa. In general description, they may be characterized as having open sockets and two or three sets of lateral barbs. End blades are not present. The line holes on four specimens are round. The spur is placed at right angles to the socket and has only one point. On three specimens there are lashing slots on both sides of the socket, while a fourth has two round holes drilled on both sides of the socket. The lash? ing holes are connected by cut grooves. Only one specimen is complete (Plate 87a) and its length is 11.8 cm. Decoration consists of incised lines extending from the line hole past, or at least to, the end set of barbs. On one specimen the decoration is much more elaborate. A mark, which may indicate ownership, is placed above the line hole. On both edges of this specimen there are two incised lines, one on either side of the edge facet. These lines all have short incised lines, which turn away at approximately 90? and nearly meet the ends of the main incised lines. The back facets are bifurcated. T H U L E nd.?This variety of Thule has not been described in the literature. Two ivory specimens were found at Walakpa (Plate 81a, d). These har? poon heads have, as their most characteristic fea? ture, blades that have neither barbs nor end blades. They have a long slender pointed blade. In its essen? tial structural features, this type is similar to Ford's (1959:82) Tipiruk-type. Ford sees this as a more slender and ancestral version of Mathiassen's (1927a: 24) Thule 1. Collins (1937c:211, pi. 71:12, 13) also de? scribes a variety similar to these from St. Lawrence Island. The complete Walakpa example, however, is simply a Thule 11 type, which has a long barbless blade. It may well indeed be the end point of the change from a Tipiruk to Thule 1 type. The smaller specimen is incomplete (Plate 8Id). We can determine, however, that it has an open socket and a round line hole. T h e most interesting feature of this harpoon head is the decoration that is incised on the blade. The larger specimen measures 16.9 cm in length (Plate 81a). It has an open socket with the spur placed at a right angle. T h e spur has a single point with gracefully scalloped edges. Lashing slots flank the socket. The line hole is triangular in shape. On the blade a fluting effect has been achieved by the removal of the core pulp from the walrus tusk. Deco? ration consists of an ownership mark placed above the line hole. The mark is an inverted Y, which has two slashes across the base. This same ownership mark occurs on several antler arrowheads in the Late Thule levels. The edge facets have been incised with a single line located on all four facets of the spur. These lines are bifurcated near the base and have two pairs of falling lines, which meet at the base. UTKIAVIK CLOSED SOCKET.?This variety is closely related to Thule nd (Plate 101a). It has multiple barbs, a round line hole and a closed socket. The spur has a single point and is placed below the socket. Three of these were recovered at Walakpa. Two antler specimens were incomplete. One made of ivory was complete; its length is 11.6 cm. Decoration consists of a simple incised line beginning at the line hole and running approximately two-thirds of the length of the blade. One specimen has an inverted Y design over the line hole. NUWUK CLOSED SOCKET.?Nuwuk Closed Socket harpoon head, in addition to having a closed socket, has spurs containing single points, which are placed below the socket. Line holes are round and the end blade slot is placed in the same plane as the line hole. The prongs are triangular to elliptical in cross- section. The main body and the spur are faceted to round in cross-section. Line holes are beveled towards the socket. Three examples of Nuwuk closed socket harpoons were found in the Late Thule levels (Plates 846, 93a, 108a) and one from an Early Thule level at Walakpa (Plate 75a). One (Plate 108a) is made of ivory and three are made of antler. They range in length from 8.4 to 7.5 cm. These specimens are decorated with incised lines. The major design motifs are converging lines forming series of triangles. T h e upper ends of the triangles have been etched solid. T h e inverted Y motif occurs on two harpoon heads over the line hole (Plates 75a, 108a). PUNUK II (f)(x).?This harpoon head type is repre? sented by one specimen in the Early Thu le level (Plate 63a). It has an open socket flanked by two lashing slots and a bifurcated symmetrically placed spur. The line hole is round and is slightly beveled ANALYSIS AND CLASSIFICATION OF ARTIFACTS 23 towards the spur. Two nonfunctional side blade slots occur at right angles to the line hole. The blade is diamond-shaped in cross-section and the blade slot is located on the same plane as the line hole. In cross- section the main body of the harpoon is almost rec? tangular in shape. The lateral edge expands directly below the side blade slots, then falls at a slight angle, where two additional triangular slots have been carved. This specimen measures 9.6 cm in length. Decoration is quite elaborate and consists of two incised lines beginning at the bifurcation of the spur and extending up the blade on both sides. At a point directly across from the upper edge of the lashing slot, the outside line turns back to the upper edge of the spur. At the exact place where the outside line turns, a hole inset with fossil ivory, has been drilled. From the edge of the spur the outside line turns back up the blade and rejoins the inside line near the line hole. One line extends up both sides to a point just below the end blade slot. On both lateral surfaces the single inside line has been carved. This line starts slightly above the line hole and proceeds to a point opposite the end blade slot. The line then turns back on itself; after a short distance, it terminates with a set of drilled holes, which have been inset with fossil ivory. WHALING HARPOON HEAD.?The only specimen of strictly whaling equipment is a whaling harpoon head from the Late Thule occupation (Plate 84a). This harpoon measures 25.6 cm in length. It is made of bone. It has a closed socket and the end blade slot is located at a right angle to the line hole. The line hole is located near the ventral edge. It is extremely large, measuring 2.7 by 1.7 cm. The line hole is not particularly beveled, but it is placed in deep con? cave surfaces carved on both sides of the harpoon point. An ownership mark occurs on the dorsal sur? face near the socket. It consists of a single line slant? ing slightly to the right. Another short line extends at a 90? angle to the right of the main line. The spur is faceted into five sides. HARPOON HEAD PREFORMS.?Eleven examples of harpoon-head preforms were recovered. As these were in various stages of manufacture, they indicated the steps in the manufacture of harpoon heads. The first step in the manufacture of open socket harpoons is the cutting of a rectangular antler strip. The second step is that of cutting one end at an angle (Plate 6h), which will eventually become the spur of the harpoon. The edges are then faceted and the point is roughly cut. Next, the blade is ground to a sharp edge, as is the point. At the upper edge of the angular-end cut there is a deep notch (Plate 37c, d). This gives additional definition to the spur and creates a flat surface from which the socket can be carved. Next, the line hole is drilled (Plate 6i). (On one example the hole is drilled completely through the preform; in another case, holes are drilled from both sides and meet near the center of the preform.) The spur is then shaped and ground into its final shape. The next step includes the cutting of the barbs, side blade slots, and lashing slots, and the final step is the manufacture of the socket (Plate 6;). The socket is made by cutting two parallel grooves up to a point near the line hole. Another groove is cut that transects the ends of the parallel lines. This leaves a central portion, which is chiseled out, and then the rough surface is ground off. Very few specimens of the closed socket preforms were recovered. The step after quartering the antler blank is that of cutting an angular surface across one end (Plate 85a). The other end is bluntly pointed by a series of facets and the diameter of the harpoon head body is also faceted. These are then ground smooth (Plate 85a). The end blade slot and the sockets are then manufactured. As no specimens were found representing these final steps, I cannot say which is manufactured first. HARPOON END BLADES Several types of harpoon end blades were found in the Walakpa middens. In general they conform to the type described by Ford (1959:96-97). T Y P E ONE.?The first type has fluted or hollow ground faces (Plate 856). This type occurs only in the Late Thule levels. Six specimens, all of gray slate, represent this type. They range in length from 5.3 to 4.2 cm. Their maximum widths range from 2.2 to 1.4 cm. The maximum widths of the flutes range from 1.4 to 1.2 cm. The style of the base ranges from concave to straight. The edges are ground all the way to the margin of the flute. Below the flute the edges meet to form a diamond-shaped cross-section. The flutes are horizontally scarred; probably the results of manu? facturing or from the wedging of these end blades into the prongs of the harpoon. These end blades appear to be very strong, as the only breakage, which occurred on five of the speci? mens, was that on the upper quarter of the blades. This appears to have resulted from stress in harpoon removal. T Y P E TWO.?Type two is closely related to the first type, but rather than being hollow ground, they are ground flat (Plate 87/). The cutting edges are not as steep as those of type one. At the termination of the hafting surface there is a steep rise to meet the edges. The edges are ground into four facets. 24 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 The type-two examples were recovered in the Late Thule occupation. The complete specimen is made of red slate, and the other of gray slate. On the only complete specimen, the base is ground flat; the length is 5.1 cm, the maximum width is 2.3 cm, and the maximum width of the hafting surface is 1.8 cm. T Y P E THREE.?Type three is the most numerous type of end blade found at Walakpa. These are both Early and Late Thule in age (Plates 59d, 73a, 776, 87d, 93d, e, 1016). There are fourteen examples: eleven of gray slate and three of red slate. These harpoon end blades are extremely thin and well made. The major portion of the end blade consists of the hafting surface. The edges are steep, narrow, and come together at the tip. The bases are ground flat. This type of harpoon end blade appears to be the least structurally sound of the types. All but one of these specimens is highly fragmentary. They ap? pear to break near the tip, at the center, and at the base of the blade. The measurements range from 3.7 to 2.3 cm in length. The maximum width is 2 cm, and the maximum width of the hafting surface is 1.6 cm. The final class of harpoon end blades includes a large variety, presumably for whale or walrus har? poons. Three tip fragments of gray slate make up the first type. These correspond to type three of the smaller variety (Plate 59c). No measurements are meaningful, but they are extremely wide. The haft? ing plane extends nearer to the tip than do those of type three of the small variety. All of these specimens are Early Thule in age. Specimens which do not fit into Ford's classes are lenticular in cross-section, with no distinct bevel on the edge (Plates 81 ft, 84c, d). The only near-complete specimen is greater than 7.3 cm in length, and it is approximately 4.6 cm in width. The base is ground flat. It is Late Thule in age. Nine harpoon end blade preforms were found (Plate 59e). These are roughly triangular-shaped pieces of chipped and ground slate. They range in length from 5.8 cm to slightly greater than 2.3 cm. The method of manufacture appears to be flaking the preform to the approximate shape with final shaping by grinding. The grinding of the edge is the last step. HARPOON FORESHAFTS On the upper end of the harpoon shaft is affixed the harpoon foreshaft. The foreshaft is used to connect the main shaft with the harpoon head. These artifacts occur in two varieties, large and small, but structur? ally and morphologically they are the same. The size difference is probably related to function, with the larger type being used for walrus, bearded seal, and possibly whale, while the smaller are used for small seals. Eight ivory specimens were recovered in the Bir? nirk levels, but only one is complete. Seven larger specimens make up the first class (Plates 15/-ra, Vie, 49h, i). One (Plate 15m), nearly complete, can be used for descriptive purposes. The foreshafts are slender, oval to rectangular in cross-section, and made of ivory. Both ends are tapered and round. On those specimens, which are complete enough for analysis, all have one small line hole near the butt end for attaching a line to the harpoon shaft or socket piece. The upper end is carved to fit the socket of the harpoon head. No total lengths are available, but for two specimens the length from the line hole to the haft end is 5.1 and 6.7 cm, respectively. On one, the length from the line hole to the tip is 16.4 cm. These foreshafts tend to break in two places?at the tip near the harpoon head and at the line hole or im? mediately behind it. These would be the two places of maximum stress during impact. The smaller variety is represented by one complete specimen and a broken end from the Birnirk levels (Plate \5i). They are structurally identical to the larger type except that one end is scarred for hafting purposes. This indicates that the larger type is made to separate from the harpoon shaft, while the smaller remains attached. The length of the complete speci? men is 10.5 cm, with the distance from the line hole to the shorter end being 2.1 cm, and 8.1 cm to the longer end. Only three foreshafts were found in the Early Thule occupation. Two are broken fragments of the larger variety (Plates 63g, 68a). They are oval in cross-section, measuring 1.5 by 1.4 cm at their widest point and tapering to a narrow blunt point. T h e complete example (Plate 806) is of the smaller variety. It is pointed at both ends. One end is closer to the widest portion of the shaft than is the other. The rectangular slot is in the widest portion. Lash? ing grooves are cut on one side of the slot. T h e speci? men has four sides, three of which are slightly fluted. The fourth side is convex. The foreshaft measures 9.1 cm in length and 1 cm at its maximum width. The line hole is 1.2 cm long and is situated 6.6 cm from one end and 1.4 cm from the other. One complete foreshaft and three fragments were found in the Late Thule levels (Plate 119a). T h e three broken speci? mens appear to have been from broken, larger fore- shafts. An extremely large, bone harpoon socket piece was found in the lowest Birnirk level (Plate 16e). This tool is 32 cm long and 2.4 cm wide. T h e distal end ANALYSIS AND CLASSIFICATION OF ARTIFACTS 25 contains a socket hole for engaging the butt of the foreshaft. The hole is 1.6 cm deep and 1.25 cm wide. At the other end it is flat on two surfaces with a hole drilled completely through the shaft. The hole is located 6.4 cm from the end. This end of the artifact is beveled to a round point. HARPOON SHAFTS Harpoon shafts are not common because wood was extremely valuable and a broken shaft could be re? used for another artifact. Five small fragments and one large fragment were recovered (Plates 68/, 101ft). The measurements of the diameters of these frag? ments range from 3 to 2.5 cm. The longest section (not illustrated) is 89.5 cm in length. It is from a Birnirk level. Fortunately, one end of this example was that which was attached to the harpoon socket. This harpoon shaft is beveled on one edge to fit the socket. The hafting surface was manufactured by cutting away one-half of the shaft at a slight angle. The length of this surface is 6.6 cm. From the lower end of the outside surface of the haft surface it is slightly beveled toward the end, where it culminates with a raised lip. There is wood discolora? tion, which outlines the binding of the socket to the shaft. The hafting surface has not been drilled. Another Birnirk specimen is a fragment of the butt end of the shaft. The end of this shaft has a drilled socket to receive the tang end of the ice pick. This hole is 2 cm in diameter and 5.4 cm in depth. This correlates with tang measurements on certain Bir? nirk ice picks. As in the hafting technique of the socket, this end is also slightly tapered toward the butt with a lip at the end. The wood is discolored from the binding of the ice pick. The last artifact was probably never utilized for a harpoon shaft, as it bends markedly to one side. It is probably a discard from the manufacture of a shaft. This section of shaft measures 40.5 cm in length. Its largest diameter is 3 cm and tapers to 2.3 cm. Al? though the surface has undergone the final smooth? ing stage of manufacture, it has not been polished. It had been cut from a larger shaft by incising the circumference to a depth of 0.3 cm; then it was snapped off. This specimen is from the lowest Birnirk level. HARPOON ICE PICKS The ice pick is attached to the butt end of the harpoon and is used to break ice from seal breathing holes. According to Ford (1959:102), they appear to change through time and are possibly diagnostic of cultural change. The ivory ice picks recovered by Ford from the Birnirk period fall within one of his five major classes of harpoon ice picks, that is, class A. He found that this was the only type that occurred in the Birnirk level (Ford, 1959:102): Type A. Only one type occurs in the Birnirk finds. It is made from the distal end of the slender tusks of young female walrus. Towards the points, flat facets have been ground to sharpen the picks. . . . At the proximal end there are a slop? ing shoulder and tapering tang made to fit into a socket in the wooden harpoon shaft. Two complete examples of type A were found in the Birnirk levels at Walakpa (Plate 49/, k). Both of these are ivory. The measurements are 17.7 and 18.8 cm in length and 2.3 and 2.1 cm in maximum diam? eter, respectively. The length of the hafting tangs are both 6.5 cm with their maximum diameter being 1.8 and 17 cm. Fourteen broken examples were recov? ered (Plates 686, 81i, 91 e, 1196, c). Their distribu? tion seems to be fairly general throughout the various occupation levels of the site. This does not conform to the pattern discovered by Ford, where this type was strictly a Birnirk manifestation. Another complete specimen was recovered from the lowest Birnirk level at Walakpa (Plate 16d). This example may make up a subtype of Ford's A group, but in many respects it is different. Although it is made from the distal end of a walrus tusk, it does not have the shoulders nor is it cut into a cylindrical shaft. Rather, it is simply a quartered section of tusk with three pronounced sides. It is pointed at both ends, with adz cuts on one end for hafting. Its total length is 28.5 cm, which is 4.5 cm longer than any reported by Ford. Two other examples of possible ice picks, which have not been described in the literature, were found in the Birnirk levels (Plates 166, 49/). Only one of these antler examples is complete. It is a long, 17.5 cm, slender, knife-shaped piece of carved antler. The blade is diamond in cross-section, having four beveled edges. At 10.1 cm from the tip of the blade there are slightly rounded shoulders, with the rest of the arti? fact being oval in cross-section. At the hafted end, adz cuts occur. On the face of the complete specimen is a small hole cut into the blade. The hole does not go completely through the blade. A single example of Ford's type B was found in a Late Thule level (Plate 115a). It is made from antler with a round shaft and the tip is semicircular in cross- section. The distal end is beveled to a point, while the proximal end is socketed to fit the harpoon shaft. There is a shoulder 5 cm above the proximal end. The area between the socket and the shoulder is roughened for lashing. This ice pick measures 19.5 cm in length and 2.4 cm in diameter. 26 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 The second variety of Late Thule ice pick, recov? ered at Walakpa, is represented by four broken examples (Plate 10lg). These specimens are made of antler. In cross-section the blade is triangular. T h e longest specimen measures in excess of 29.5 cm in length. The other two examples are small blade sections. The tang on the longest specimen is wedge- shaped, with a single bevel along one edge. The beveled surface has been roughened for hafting by adz cuts. The final examples of ice picks from Walakpa are probably closely related to Ford's type E (Ford, 1959: 102-104). The major difference is that this variety does not have the cut scarf of the type E, and the blade tends to be oval in cross-section. The blade is carved round in a very crude fashion. The tang is wedge-shaped, having been beveled on one edge. All surfaces of the tang are scarred for hafting. There is no shoulder to offset the tang from the blade. This specimen dates Late Thule in age. It measures 19.4 cm in length. The tang is 8.6 cm in length, with a width of 2.2 cm. Another extremely interesting ice pick was found in the upper Late Thule level (Plate 93i). It is made from a split caribou antler. The blade is curved gracefully to a sharp point. The tang is wedge-shaped, the wedge being produced by cutting the split antler at a slight angle until it intersects the os-compacta of the antler. The inside cancellous structure has been scarred for hafting. At the upper end of the tang a hole is drilled. Slightly above the hole two notches are carved in both lateral edges. The interior of these notches is roughened. Near the tip of the tang, on the outside surface, is a shallow notch. These are prob? ably hafting devices. The total length of this ice pick is 33.8 cm. The tang measures 12.1 cm in length, with its maximum width being 3.3 cm. Seal Poke Gear FLOAT BARS Ford describes three major varieties of float bars: (1) small bars with raised ends late in time, (2) large round bars with rounded ends, Birnirk in age, and (3) curved bars, Birnirk in age (Ford, 1959:104). T h e float bars from Walakpa generally follow Ford's model. Ford's Type 1: No examples of type 1 were found in the Late Thule levels, but they did occur in the historic houses. Ford's Type 2: Five examples of the second type were found in the Birnirk and Early Thule levels (Plate 70a, 6). It is hard to determine whether these are float bars or baleen basket handles. It is most probable that they are float bars, as the ends have not been perforated for basket suspension. They are round in cross-section with the widest portion in the center and slightly tapered to either end. This type does not have the raised lip on the ends as does the smaller type 1 mentioned above. T h e range of lengths for these bars is 22.9-15.1 cm. T h e average length is 18.6 cm. The range of maximum measure? ments of the center diameter is 2.7-1.7 cm, with an average of 2.1 cm. Ford's Type 3: One example of the curved type was found in the Birnirk level. This specimen is crescent- shaped (Plate 17a). Its length is 18.3 cm. T h e center is concave. The length of the concavity is 4 cm. Below the concavity the ends taper to a point and turn downward. SEAL FLOAT MOUTHPIECES As the name implies, the seal float mouthpiece was used for the inflation of seal floats. Two examples of these artifacts were found at Walakpa. One is of Birnirk age and one is Early Thule age. The Birnirk specimen, like those of Ford's (1959: 105) was crudely manufactured in comparison to the later forms (Plate 47c). It is a short, cylindrically shaped tool with a wide groove encircling the entire shaft. The groove is used in the attachment of the mouthpiece to the float. It has an extremely small hole cut through the length of the cylinder. It is 3.6 cm in diameter and its length is 4 cm. T h e width of the attachment groove is 0.8 cm. A single example of an Early Thule seal float mouthpiece was found (Plate 69c). This is an ivory specimen that measures 3.4 cm in diameter and 2 cm thick. It is concave on the inside and convex on the outside. A small hole was drilled through the center. It probably has an attachment groove; however, the mouthpiece is still attached to a portion of the seal skin and the groove is not visible. SEAL FLOAT MOUTHPIECE STOPPERS Four examples of mouthpiece stoppers were found. Three are of Birnirk age and one is Late Thule in age. The ivory specimen measures 4.8 cm in length (Plate 51c). It has at the outside end a short, tabular handle, which would fit the fingers. T h e remaining portion of the tool is slender and tapers to a blunt point. Around the shaft, spiraling scars are visible, which resulted from the stopper being forced into the mouthpiece. A wooden specimen of Birnirk age is identical in form to the one described above (Plate 5Id). It has a ANALYSIS AND CLASSIFICATION OF ARTIFACTS 27 short, tabular handle placed above a long, tapering plug. It measures 5.8 cm in length. Another Birnirk specimen was also made of wood. It consists of a round piece of wood, larger in diameter at one end and tapering to the small end. At the small end the wood is peeled back a distance of 1 cm. This has resulted from its insertion into the mouthpiece. It measures 3.3 cm in length, 1.1 cm in maximum diameter, and 0.4 cm in minimum diameter. The last specimen is Late Thule in age (Plate 81*). It, too, is a cylindrical-tapering piece of wood. It measures 3.7 cm in length, with the diameter tapering from 1.2 cm to 0.6 cm. SEAL FLOAT PLUGS These plugs are used to close holes in the floats. A wooden Birnirk specimen is quite well made and is possibly a float mouthpiece without the hole having been drilled (Plate li). It is polished and appears to be a finished specimen. It is 2 cm in length and 3.4 cm in diameter. It has an attachment groove carved around the outside surface, which measures 0.8 cm in width. The second specimen, also Birnirk in age, is crudely made, possibly an unfinished tool. It measures 4.6 cm in length and its diameter is 3.4 cm. The attach? ment groove is slightly offset towards the inside face. The groove is 0.8 cm in width. DETACHABLE LANCE POINTS Ford describes an artifact known as the detachable lance point (Ford, 1959:105). This is a wooden shaft, measuring 37 to 41 cm in length. The tip has a crude stone blade. The proximal end has a conical point, which fits into the harpoon socket piece. A shallow notch near the butt serves to retain a line that holds the lance point in place. A possible example of this artifact was found in the Late Thule occupation. The blade end is broken, but it appears to have once had a blade slot. The length is greater than 36.5 cm and its diameter is 2 cm. The butt end has a conical point. The major difference, which may mean that it is not a lance point, is that it does not have the retaining notch. Ice Hunting Gear WOUND PLUGS The wound plug is used to plug the wounds of animals to conserve the blood in the carcass. These have also been described as plugs to hold air blown under the skin of a seal to make it float (E. Nelson, 1899:147, Ford, 1959:106). Ford (1959:106) feels that there has been little change in the morphology of this artifact through time. I would have to agree, as in our finds this artifact does not change through time. Four wound plugs are from the Birnirk occupation (Plates 11, m, lib, c). They can be described as blunt, pointed, tapering pieces of wood. They range from 9.5 to 8.5 cm in length. Ford (1959:106) indicates that his specimens have grooves encircling the larger end of the artifact. These specimens from the Walakpa site were attached at the opposite end. One has an encircling groove at the narrow end, while two others have perforated holes at the same end. One of these holes is square while the other is round. A preform was also recovered from a Birnirk level. It is nearly complete except that there has been no arrangement made for attachment to the bundle. Another specimen (Plate 51ft), which I will describe separately because it is unique, was found in an upper Birnirk level. The most distinguishing feature of this example is that at its wide end it has two flanges on both edges. It has a round hole drilled at the pointed end and measures 9.2 cm in length. Two additional specimens come from the Late Thule occupations. Here, again, the holes have been drilled in the pointed end. WOUND PINS Wound pins from Walakpa can be separated into two varieties. One is exclusively Birnirk, while the other is Thule. Four ivory wound pins were found in the Birnirk levels (Plates Ij, k, lid, 51g). They tend to be round in cross-section, but near the tip, four facets make a diamond-shaped cross-section. The upper end is flared and gives the artifact a nail-shape in appear? ance. These do not have the sharp edges reported by Ford (1959:106). The range of lengths is 11-7.8 cm. The diameters range from 0.7 to 0.5 cm. A bone specimen from a Late Thule level (Plate 87;) more closely resembles Ford's specimens (1959, fig. 41). In cross-section it is flat to lenticular. The base is expanded and has a short handle-like central section, which results in the pin looking like a minia? ture dagger. Directly below this handle there are shoulders, which expand to the blade. The edges are ground sharp. The length measures 9 cm. Its width is 0.8 cm and its thickness is 0.4 cm. ICE SCOOPS The ice scoop, as described by Murdoch (1892:308- 309), was a baleen netting lashed inside an oval-shaped 28 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 antler strip. It has a long handle attached at one end. These were used for removing ice chips when cutting holes in the ice or gathering chips of fresh water to melt. Two small sections and one large section of an ice scoop rim were found in the Birnirk levels (Plate lle-g). They measured 2-2.7 cm in width. The dia? meter of the scoop on a nearly complete specimen is roughly 18 cm. This is identical to the diameter of those from the Birnirk site (Ford, 1959:107). Along the bottom edges of these rims are lashing slots through which the baleen netting was attached. These slots, on the most complete specimen, average 3 cm apart, while on the broken rims they average 1.9 cm. These slots are connected on all but two specimens by grooves used for additional lashing support. One specimen had two holes drilled at both ends to lash the ends together. On another specimen, an additional hole was drilled on the upper edge. Ford (1959:107) recovered at least one rim with this type of attachment mechanism. The last Birnirk specimen is made from antler. Rather than binding holes, this example has a narrow tang with lateral spurs. The tang is either inserted into a slot or it overlaps the other end. A single broken Late Thule specimen was also recovered (Plate 114g). It is made of antler and is too fragmentary to yield measurements. TWO-PIECE ICE SCOOPS The two-piece ice scoop was used to skim ice from fishing holes and to remove snow and hoar frost from clothing (Murdoch, 1892:309). It is made from two semicircular pieces of antler, baleen, or horn lashed together to form a scoop. The two halves are attached to a handle. Ford (1959:108) speculates that this was not a popular tool in early periods but came into more general use later. Our meager evidence tends to support this thesis. Only four halves were found, all of Late Thule age (Plates 82ft, i, 111ft). None of these were found in the Early Thule or Birnirk levels. Two halves (Plate 82ft, i) were found that were undoubtedly part of the same artifacts. These are made from antler measuring 12.8 by 9.1 cm. The edges are smooth and blunt. On the inside edges there are four matching holes drilled in each half, which are for lashing the halves together. Two lower lashing holes are square. The upper two holes are round and were probably used for the handle attachment. A third antler specimen (Plate 111ft) is different and may not be part of a two-piece ice scOop. It has the general shape and outline, but the holes for the attachment of the second half are peculiar in place? ment and there is no provision for handle attachment. It measures 13.7 cm in length and its maximum width is 5.2 cm. It has two holes placed close together and near the juncture of the inside edge and the distal end. The distal end and the outside edges are ground sharp. The inside edge is ground flat to fit to the other half. The fourth example of a two-piece ice scoop is made of baleen. The handle end is incomplete, but the overall shape must have been semicircular. The outside edges are rounded while the inside edge is flat. Along the inside edge four holes were drilled for the attachment of the other half and the handle. SEAL DRAG HANDLES As the name implies, a seal drag is used to drag home seals. It consists of a thong about 46 cm (18 in) long attached to a handle. T h e thong is attached to the seal's lower jaw. The handle can be used for attachment to a longer line or to a dog harness (Murdoch, 1892:256). One specimen of a seal drag handle was found at Walakpa. It is Late Thule in age (Plate 10\e). The handle is carved from dense bone, such as a walrus mandible, and resembles a seal's head. A hole is drilled from a tang at the base of the head to a point just below the mouth. A second hole goes through the tang at a 90? angle to the original hole. These holes serve to attach the thong to the handle. The eyes and nostrils are drilled holes. A mouth is repre? sented by an incised line. T h e handle measures 4.5 cm long and 1.5 cm wide. SEAL SCRATCHERS The seal scratcher is used to make a scratching noise on the ice, which was thought to attract seals to the hunter (Murdoch, 1892:253). Two very fine complete specimens were found in the Birnirk occu? pation at Walakpa. Ford (1959:102) did not recover any from the Birnirk site. The first specimen fits Murdoch's (1892, fig. 253) drawings of those being used in historic times at Point Barrow (Plate 50d). It is 21 cm long and 5 cm at its widest point, which is just below the claws. It is manufactured from a single piece of wood, which resembles a seal's flipper. T h e wood is carved to allow seal claws to slip over five spurs at the distal end. The claws are attached by baleen cords. Two holes, cut side by side, are placed 8 cm back from the distal end. The holes are cut at an angle so that they intersect ANALYSIS AND CLASSIFICATION OF ARTIFACTS 29 each other and allow the baleen to pass through. The holes do not penetrate the ventral side of the tool. There are four baleen cords that are tied around the claws, passing through the holes and then back up to be attached to another claw on the other side. No provision has been made for attaching a thong to the proximal end of the scratcher. The second specimen is made from a carved piece of wood (Plate 506). When the handle portion is held in the hand, the claws bend downward. This specimen has only three claws. The claws are ap? parently made to be covered by seal claws, as in the specimen reported above. There are two holes drilled at the ventral side of the scratcher, which must have been used in claw attachment. The tool is carved to leave a thicker cross-section immediately behind the claws. The holes are drilled in this thicker section. The proximal end of the handle has a hole drilled through the side. In cross-section the handle is rectangular and flattens to a wide triangular cross-section near the claws. The length of this specimen is 18.1 cm. SEAL RATTLES The cottonwood block and ivory bangle seal rattle reported by Ford (1959:109) was not found at Walakpa; however, a possible rattle was found in the lowest Birnirk level (Plate 18). It is made from two split caribou scapuli attached to a short wooden handle. The handle is 20.5 cm in length. A notch is cut at one end and a hole drilled 7.2 cm back from the end. A baleen loop passes through the hole and around the notch. The distal end of each scapula is drilled and another loop of baleen attaches both pieces. The two loops are joined together, allowing the scapuli fragments to hang loose. When the handle is jerked, they bang together producing a clicking sound. NETS AND N E T WEIGHTS Nets were used for capturing seals and fish. One fragmentary example of a net was recovered at Walakpa. It is made of thin baleen strings tied together producing diamond-shaped openings. It is Late Thule in age. An example of a net weight made of dense bone was found in a Birnirk level (Plate 47d). It is circular in shape and on one end it has two sets of holes, which are used to attach the weight to the net. It is 5.2 cm in diameter. Two stone weights of Late Thule age were found (Plate 116a, 6), which are heavy quartz beach pebbles. They are not modified but are bound with baleen lashing. The lashing has a major knot on the upper surface with four bands encompassing the stones. Rawhide lashing is attached around the central knot. This lashing is used to attach the stone weights to the net. These weights are 10.3 and 9.2 cm in diameter. At la t l Equipment ATLATL DART BOARDS The atlatl dart boards were used for casting both seal darts and bird darts (Murdoch, 1892:218). A possible dart board handle was recovered in an Early Thule level (Plate 63/). It was purposefully cut in half just above the forefinger platform. The groove for the atlatl shaft runs the full length of the handle. Three finger holes are located on the left margin, with a notch to rest the middle finger below the forefinger platform. The opposite side is carved round to fit into the palm of the hand. An example of a distal end of a throwing board was found in the same Early Thule level (Plate 63n). The end is carved blunt. The dart shaft groove terminates at 4.1 cm from the end. Above the end of the groove a hole is drilled to receive the hook that engages the dart butt piece. On the reverse side there is a shallow hole of undetermined function. ATLATL HOOKS A possible atlatl hook made of ivory was found in a Late Thule level (Plate 85n). Its total length is 2 cm. It has a relatively long hafting tang and a short, pointed barb for engaging the dart butt piece. DART SHAFTS AND BUTT PIECES Dart shafts are hard to identify unless a diagnostic feature can be found. There is only one candidate for a dart shaft in the Walakpa collections and its identification is highly tenuous. It is of Early Thule age. If it is a dart shaft it is the butt end. The end of the shaft has been carved to receive another artifact such as a butt piece. The shaft is narrow at the butt, 1.2 cm in width, and expands to a maximum diameter of 1.9 cm. The shaft is diamond-shaped in cross- section. The total length of this fragment is 24.7 cm. The shaft is highly polished. A dart butt piece was recovered from the Birnirk occupation. It is 2.7 cm in length. The socket is circular with a concave base. Four tiny holes of unknown function are drilled into the rim of the base. 30 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 The diameter of the socket is 0.1 cm. The hafting tang is 2.2 cm in length. It is wedge-shaped and none of the surfaces have been scarred for hafting (Plate 37f). HARPOON SOCKET PIECES The type of harpoon socket piece found at Walakpa is similar to those found in Eastern Thule collec? tions (Mathiassen, 1927a: 28). Three specimens were recovered in the Early Thule levels and one from a Late Thule level. In general form they are very similar. The first two specimens are of the same style. These measure 6.1 and 4.6 cm in length, respectively. They are cylindrical in form with the inside portion of the bone carved out, leaving a thin cylinder. The socket opening on the only complete specimen (Early Thule) is 1.8 cm in diameter. Around both the upper and lower lips there is a groove. The second variety is Early Thule in age. It is essentially the same as those described above, with the exception that it does not have the grooved rings around the lips. The surface is highly polished and the socket end is slightly flared. Its length is 5.4 cm and the diameter of the socket is 1.7 cm. These sockets function by sliding down over the distal end of the dart shaft. Since the holes are large and round rather than rectangular, they must have been used for larger lance points or bladder darts. One possible preform for a socket piece is from a Late Thule level. This is a cylindrical-shaped piece of bone, which measures 5.4 cm in length. At both ends of the cylinder, holes have been started but were not completed. INFLATION NOZZLES One example of an inflation nozzle was found in an Early Thule level (Plate 59/). It is made of ivory and measures 2.7 cm in length. It is cylindrical with a hole drilled through the center. A raised portion at one end forms a thin lip. The center section is indented with attachment scars cut three-fourths of the way around the diameter. The lip at the distal end is larger and has a transverse tapering base. In diameter this artifact measures 1.3 cm. BIRD DART POINTS AND SIDE PRONGS Murdoch (1892:210-214) reports that atlatl cast bird darts were used at Point Barrow. These darts consisted of a single end point or occasionally two points with additional prongs attached near the midcenter point of the shaft. At Walakpa four specimens of bird dart side prongs were recovered. No end points were found. All of these specimens fit into either Ford's (1959:117) class A or B. Three of these specimens from the Walakpa site were found in the Birnirk levels while the last was from an Early Thule level. Ford's (1959:117) class A side prongs are flattened ovals or lens-shaped in cross- section and have barbs on both edges. Two of this type were found in the same Birnirk occupation level. One is complete while the other is a small fragment. The complete specimen (Plate 51a) was manufactured from antler. It measures 14 cm in length. There are four side barbs on the inside edge and three barbs on the outside edge. The barbs are made by cutting into the blade at an angle greater than 90?. T h e point of the lower barb tapers back to the cut made for the barb above. A lashing slot is cut near the base of the prong. At the edge directly above the lashing slot there is a small notch. This is undoubt? edly used for lashing. Both faces of the base of the prong are also scarred for lashing. The fragment of a side prong is made of ivory (Plate 516). It has a lashing slot and edge notch. T h e edge notch is placed above the lower end of the slot. The faces of the base have been slightly roughened for hafting. Specimens of Ford's (1959:117) class B are similar to those of his class A except that they are triangular in cross-section with barbs placed on all three edges. A Birnirk specimen is made of ivory. Its total length is 10.9 cm. There is only one side barb on two edges and none on the third. T h e edges with the side barbs are gracefully carved with an incised line running the full length of the prong on each side. There are three slightly raised borders and at the center of each there is a lashing notch. A lashing slot is carved slightly off center and closer to the barbed edges. On the barbless edge and slightly above the upper end of the lashing slot there is a relatively large notch. It has a tapering tang, which has not been scarred for lashing. An Early Thule specimen is made of ivory (Plate lie). Its length is 10.9 cm. All three edges have small side barbs. One edge has at least two while the other edges have at least one. An incised line is carved on both sides of each edge. T h e lashing slot is placed near the base. It is impossible to determine if there were any lashing notches. Another type of side prong for a bird spear oc? curred in the Late Thule occupation. It was not described for Point Barrow by Ford but is similar to a specimen described by Giddings (1964:31, 34) for Nukleet. This ivory artifact measures 11.5 cm in ANALYSIS AND CLASSIFICATION OF ARTIFACTS 31 length. It is flat to oval in cross-section with side barbs on both edges. The outside edge has only one barb and there are two areas for lashing, one carved at the base and the other a short distance above the first. The inside edge has two barbs. The base of the inside edge is beveled and the surface is scarred. HARPOON DART HEADS Giddings (1952:54-55) found that on the Kobuk River there were two sizes of harpoon dart heads. The smaller type has been termed "harpoon arrow point." Two examples of the smaller variety were found at Walakpa. Both were made of antler. A broken specimen of this type was found in the Birnirk level. The blade is diamond in cross-section and has only one side barb on each edge. A line hole is drilled into the shank, but does not appear to be beveled. The length from the line hole to the tip of the blade is 3.6 cm. A complete example, measuring 7 cm, was recovered in the Late Thule level. Both edges have one side barb. Toward the tip on one edge it looks as though there may have been an additional barb. The edges curve outward and resemble a Christmas tree. A line hole is drilled in the shank of the dart head. On the os-compacta surface of the antler the line hole is beveled for the line. The opposite side has not been beveled. The length from the line hole to the tip of the blade is 5.2 cm. The base is carved to a wedge-shape. The butt is flat with an edge that measures 0.7 cm along its long axis. Another extremely interesting example came from a Late Thule level. This long antler harpoon dart head is 30.7 cm in length (Plate 89/). Three deeply cut side barbs occur on one edge. At the termina? tion of each cut for the barbs, there is a long incised line, which runs almost to the beginning of the next incised line. The incised line of the last barb runs completely to the end of the dart head. The tip is socketed to receive a stone end blade. Two line holes are drilled at the base, one directly below the other. On one side of the blade the line holes are connected with a short groove. The tang is beveled on all sides and terminates with a sharp point, which is scarred for hafting. B o w and A r r o w Equipment Bows Two specimens of bows were recovered at Walakpa, one complete and the other a fragment. Both are of the "Arctic Type" (Murdoch, 1884:316) or a single- piece self bow. These are the type n bow in Hamil? ton's (1970:46) classification. The complete bow is Birnirk in age (Plate 13). Although small, 76 cm long, it is probably a func? tional hunting bow rather than a toy. It is a short spruce bow, nearly elliptical in cross-section, being flatter on the back than on the ventral surface. The handle is narrow, but thick in cross-section. It has recurved wings. The nock ends were carved by splitting the wood on both sides and breaking the edges off. Although no traces of the sinew backing are present, marks on the ventral surface indicate that it was once present; there are six bleach marks on one end and four on the other. The broken specimen comes from the Early Thule occupation (Plate 68&). It consists of a recurved end. It would have been a larger specimen than the one described above. At its maximum width it measures 4 cm and it is 1.3 cm thick. The wide wing tapers to a carefully manufactured nock. Two wedge-shaped notches were cut on both sides of the end leaving a small wooden central peg for the attachment of the bowstring. At 25.7 cm from the nock the wing is curved. It is at this point that the bow was broken. Just above the break the edges have been roughened for the attachment of the sinew backing. There were at least five lashes, each having six strands for holding the sinew backing in place. The bow is elliptical to rectangular in cross-section. MARLIN SPIKES Marl in spikes were used to raise parts of the sinew cord when putting the backing on a bow (Murdoch, 1892:292). Although no tools such as Murdoch illus? trates were found at Walakpa, several were found which are similar to those illustrated by Giddings (1952:72) from the Kobuk River. The three examples from Walakpa were found in the Late and Early Thule levels and fit the same basic type (Plates 77ft, 83g, 89;). They are long and slender, slightly curved, and with pointed rib sections. All have holes drilled near a blunt end. The most nearly complete specimen has an ownership mark below the hole (Plate 83g). The ownership mark consists of a bifurcated line placed over the suspen? sion hole. An incised line extends 1.8 cm down the spike from the point of bifurcation. This marlin spike measures in excess of 18 cm. In maximum width they are 1.5 cm, 1.3 cm, and 1.1 cm, respec? tively. SINEW TWISTERS A single bone sinew twister was recovered from a Birnirk level (Plate 8e). It is rectangular in cross- section, measuring 11.5 cm long and 1.1 cm wide. At 32 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 either end and on opposite sides flanges are carved producing notches to keep the sinew in place. ARROWSHAFTS In order for a wooden shaft to be classified as an arrowshaft, it must be either complete or have one or the other distinguishing end described below. Two types of arrowshafts were found in the Walakpa levels. The first type was for hafting deer arrow? heads (Murdoch, 1892:201). The haft end of this type of arrowshaft does not change in appearance from Birnirk to modern times. Eighty-one examples were found in the Walakpa mound. Two are Birnirk in age, six are Early Thule (Plate 73ft), and 73 are Late Thule (Plates 85g, 896, 94ft-;', 102*', 108/). The ends of these shafts are cut at right angles to the shaft. A conical hole is drilled into the flat end of the shaft. This conical hole received the tang of the antler arrowhead. The deer arrowhead is wedged into the conical holes up to and beyond the shoulder of the arrowhead tang. The stele and pile are wrapped with baleen lashing. The pile is wrapped to a point above the shoulder of the tang and the lashing extends about 3.8 cm up the stele. The stele occasionally split when used, which allowed the pile to slip in its haft. Consequently, we see examples of these haft ends being purposefully cut above the splits and being discarded. Four specimens were recovered in which the broken arrowhead tangs are still in place. The diameter of the stele at this end ranges from 1.1 to 0.6 cm. The second type of haft end was used on Bear arrows (Murdoch, 1892:202; Plates 102g, 115i). There are six examples of this type of stele in the Walakpa collection, all of which are of Late Thule age. These artifacts do not have the concave ends, but are expanded and flattened out with the surfaces taper? ing to a blunt end. The ends have been split laterally, and a stone point is inserted into the split. The haft? ing is reinforced by sinew bindings, which extend up the stele for at least 2.5 cm. One of these specimens has an ownership mark carved above the lashing scar. The ends of these hafts range from 1.81 to 1.1 cm in width and average 1.3 cm. An Early Thule arrowshaft is unique (Plate 89a). It has four projections carved in the haft end of the shaft. Directly below the projections, a notch has been carved around the diameter of the shaft. Below the notch the shaft is square in cross-section with four hollowed-out sections on each side. This gives the impression of a worn-out drill bit. T h e shaft is broken but would have measured in excess of 30 cm. It measures 1.1 cm in diameter with the diameter of the point being 1.1 cm. The nock end of arrowshafts from the Walakpa site sort into three distinct types. These types seem to be unique to certain culture types. T h e earliest variety found in the Birnirk levels can be charac? terized as having a triangular-cut nock (Plate 48c). The stele is slightly widened just below the nock and is round in cross-section. Of five specimens of this type only one had any direct evidence of fletching. This consists of two shallow cuts or scratches across the stele just below the nock. T h e range of diameters on this variety is 0.8-0.6 cm, with an average of 0.7 cm. The second variety appears to be characteristic of the Early Thule levels (Plates 68g, ft, Ilk). It is similar to the one described above, the major differ? ence being that these tend to be flat in cross-section rather than round. The nock end is more expanded and the upper part of the stele tends to be flat and expanded. The nock has a large V-shaped cut on both outside edges. Thirteen examples of this variety of arrowshaft were recovered. Seven of these have evidence for fletching. Two types of fletching were utilized. In the first, the feather is bound to the stele without any modification. Binding marks occur directly below the nock end and at an average of 0.5 cm below the nock. The widths of the upper binding scars range from 1.4 to 2.2 cm. T h e width of the lower binding ranges from 2.2 to 1.2 cm. T h e outside length of the fletching ranges from 14.1 to 11.3 cm. One of these examples has a small notch placed on one edge near the center of the fletching, which may be for stabilizing the fletching. T h e second technique for fletching is cutting notches below the nocks, around which a binding holds the feathers to the stele. Two specimens have one notch, while another has a series of four notches. The diameters of the stele on this type ranges from 1.1 to 0.7 cm. The third type of arrowshaft is found only in the Late Thule levels (Pates 85/, 102ft, \08i-k, 118e). This type was utilized up to historic times. T h e majority of this type do not have an expanded notch. T h e general form for these arrowshafts is a round stele, which gradually flattens out to an oval cross-section near the nock. T h e butt of the nock is flat with a U-shaped notch in the center. T h e diameters of these arrowshafts range from 1 to 0.8 cm. T h e fletching is attached by cutting a slight groove directly below the nock. No lower grooves have been cut. Binding marks occur on the upper end of one specimen, which measures 2.5 cm in length. One shaft has traces of red paint on the stele but no design was discernible. ANALYSIS AND CLASSIFICATION OF ARTIFACTS 33 FEATHER-CUTTING BOARDS Three wooden feather-cutting boards were found at Walakpa. All three are Late Thule in age. These boards are thought to be used for either cutting feathers for fletching or for cutting sinew thread (Ford, 1959:138, Giddings, 1952:81). The first specimen is a long, thin, bipointed piece of wood (Plate 89&). It is rectangular in cross-section. There are 34, 30, 15, and 11 irregularly placed notches on the four respective edges. The surfaces are slightly concave with both vertical and horizontal cutting scars. The board is highly polished on all edges and surfaces. It measures 17.8 cm in length, 3.3 cm in width, and 1.2 cm in thickness. It is possible that because of the notches, the artifact may be a game counter; but due to the cutting scars on the surfaces, I feel that it is probably a cutting board and the notches may serve as gauges. The second specimen is an elongated oval that is rounded at both ends (Plate 85;). It has a semi- lenticular cross-section. The flat surface has numer? ous vertical cutting scars with fewer horizontal scars. It is 9.7 cm long, 3 cm wide, and 0.8 cm thick. The last specimen is also an elongated oval in out? line. Its cross-section is semilenticular. The flat surface is also scarred with vertical and horizontal scars. One end is thinned to a tang. It measures 10.4 cm long and 3.3 cm wide. In thickness it measures from 0.7 to 1.6 cm. This specimen has a dubious identification, and the surface scars may be the result of manufacturing rather than cutting. ANTLER ARROWHEADS Antler arrowheads have caused an extremely in? teresting and equally confusing problem in Eskimo prehistory. It is thought that the arrowheads are almost as useful as harpoon heads for establishing cultural chronologies for the north coast of Alaska (Ford, 1959:124). The most distinguishing feature is apparently the tang. Collins (1937c:324) first described the important types of tangs, which he termed the "tapering tang," the "knobbed tang" and the "spurred tang." Ford (1959:124) used this same general taxonomic system for his Point Barrow specimens, although he included placement of barbs as another important type determiner. It was observed in the Walakpa specimens that several different types occurred within the same level and that in levels where ownership marks occurred the same ownership marks were found on specimens of purportedly different types. This raises the ques? tion as to whether the variability of types is temporal, functional, or fortuitous. As the antler arrowpoint is reported to be used only for caribou hunting (Murdoch, 1892:201-207), it would be extremely difficult to determine functional differences. The possible difference that could occur would be the manner in which the caribou were hunted, e.g., single kills, herd challenges, drives, surrounds. The principle for killing the animal with bow and arrow, however, would be essentially the same. The varia? bility in antler arrowpoints is probably not func? tional, but probably has more to do with the individ? ual manufacturer's whim at the time of manufacture. However, these variable were all compared and include the following categories: (1) shoulder type: curved, square; (2) knobs: curved knobs (strong, weak, strong upper portion), circular, lateral (offset, symmetrical), no knobs; (3) barbs: single barb, uni? lateral multibarb, multiple-bilateral barb (symmet? rical, asymmetrical), no barb; (4) spur: short, long; (5) cross-section: round, lenticular, triangular, dia? mond-shape, combinations; (6) ownership mark: present, absent. Using these twenty-three variables on a sample of over 400 antler arrowpoints from Walakpa, Birnirk, Utkiavik, and Cape Prince of Wales, it was found that only four specimens conformed to a single type, while each of the others then represented a unique type of its own. A comparison of the arrowheads was made using only the first three categories: (1) shoul? ders, (2) knobs, and (3) barbs. This comparison indicated that the form of the barbs did not constitute type indicators. The shoulder and knob form ap? peared to be the only traits that seemed to change through time, or to indicate types. All five barb forms were found on the various shoulder and knob classes. It therefore appears that Collins (1937:324) was correct when he discussed types based on only shoulder and knob forms. On the basis of these comparisons we found that there were eight types of antler arrowheads, which indicated general trends through time. These eight types and their varieties, based on barb placement (used only in a descriptive sense), are discussed below with the earliest forms first and younger forms last. T Y P E I, TAPERING TANG.?This type is thought to be one of the earliest, and at Walakpa it appeared only in the Birnirk levels. This type is called the "tapering tang" by Ford (1959:128). In his classifica? tion, however, he includes types with both a strong central knob and those without knobs, which may be temporarily significant. For these reasons I have split these types into two distinct groups. Type i has a rounded shoulder, long, tapering tangs and no knobs. Five of these were found, all of Birnirk age. There 34 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 are three varieties of side barbs: (1) no side barb, (2) multiple, bilateral barbs, and (3) multiple-bilateral barbs with end blade slots. Three barbless examples were found. Two larger examples (Plates 8a, 51 q) measure 16.1 and 15.1 cm in length. The tangs are 3.3 and 2.8 cm in length, respectively. The blades are slightly expanded near the tip. No ownership marks are present. The third barbless example (Plate 86) is extremely short (9 cm), but otherwise it is identical to those discussed above. The tang measures 3.3 cm long. There is only one example of the multiple-bilateral barb variety (Plate 5lp), which is 16.3 cm long, with the tang 2.8 cm long. It has a side barb placed on both sides of the blade. These barbs are part of the blade structure. One barb is slightly higher on the blade than the other. These barbs are both long, narrow, and extremely sharp. The last specimen of this type is unique for this collection (Plate 19a). It has a bifurcated tip for the insertion of an end blade. The barbs are short and are part of the blade structure. One barb is placed above the other and on the opposite side of the blade. The length of the arrowhead is 16.4 cm; the tang is 2.6 cm long. There is no ownership mark on the arrowhead. This arrowhead was found in sterile clay a few centimeters below the lowest Birnirk level. It is unknown if it is Birnirk in age or older. It is the author's opinion that it probably belongs to the Birnirk level and was forced into the wet clay from pressure by the occupants of the site. T Y P E II, KNOBBED TAPERING TANG.?This type is similar to type 1 and is the type that Ford (1959:128- 132) included in the tapering tang category. Twenty- eight examples were recovered at Walakpa. These were found most commonly in Birnirk and Early Thule levels, but four* were found in Late Thule levels. Most of these were originally long arrowheads but appear to have been broken and resharpened. The range of lengths is 20.7-6.8 cm, with an average length of 11.3 cm. The majority of these arrowheads have a single lateral barb. These barbs group into three types: (1) those with the main blade terminating in a barb at a distance of over one-half the length of the point (one of these is Birnirk in age, Plate 5In, and three are Early Thule, Plate 63ft, ;); (2) those with the barb being placed almost directly at the end of the blade (one example of this type is Birnirk in age, Plate 8c); (3) those with a single barb carved in the side and are not a part of the main structure of the blade (two are Early Thule, Plate 68e, two are Late Thule). T h e varieties of barb manufacture occur together in sev? eral levels and are not to be taken as functional variables. Four small specimens have no lateral barbs of any kind and appear to be resharpened examples of the larger varieties. Three are Early Thule in age (Plate 59ft-;) and one is Late Thule (Plate 91k). Two examples have multiple-unilateral barbs. A Birnirk specimen is rather unique (Plate 51o). The end of the blade has been widened before it tapers to the point. T h e first barb is part of the blade struc? ture and, like those described above, it is placed about halfway down the length of the blade. T h e second barb is quite small and is placed directly below the first barb. In the case of a Late Thule example, the barbs are evenly spaced on the blade but do not make up part of the blade structure (Plate 118c). The tang on this specimen has been roughened by lateral cuts for hafting purposes. No ownership marks are found on any of these specimens. T Y P E HI, CONICAL KNOB.?The conical tang variety was first described by Larsen and Rainey (1948:169). This type was associated with the Western Thule levels at the Ipiutak site. Ford (1959:129-136) found this type at both Birnirk and Nunagiak. Ten specimens of these arrowpoints were found at Wal? akpa: two from the uppermost Birnirk levels, five from the Early Thule levels, and three from the Late Thule levels. I would therefore conclude that these are probably Late Birnirk types and are transitional forms between Birnirk and Early Thule , and persist into Late Thule times. The most characteristic feature of these antler arrowheads is the conical tang. The upper edges of shoulders of the cones are rather sharp and intersect the taper of the rounded shoulder. T h e cone is short in comparison to the tapering tang varieties. At least two types of barbs are present, single barbs and multiple-unilateral barbs. These measure from 16.3 to 13.1 cm in length. The length of the tangs from the shoulder to the conical tip range from 3.5 to 2.7 cm. The average of the tang lengths is 3 cm. T h e length of the conical portion of the tang ranges from 2.1 to 1.1 cm, with an average of 1.6 cm. T h e longer tangs are from the Birnirk levels. The most common form of barb is that of the single barb. All of the single barb types have the barbs carved independently of the blade structure. One, of Early Thule age, is a short barb that extends out from the blade at a steep angle (Plate 68d). Two other Early Thule examples have relatively long barbs that are slightly parallel to the blade (Plate 63*). T h e cuts that separate the barb from the blades on these specimens are long and cut quite deeply into the blade. ANALYSIS AND CLASSIFICATION OF ARTIFACTS 35 The only example of multiple-unilateral barbs is Late Thule in age. It has five small barbs along the central portion of one side of the blade. These barbs are manufactured like the barbs of a leister prong. Six broken examples, four Early Thule (Plates 58d, 59fc, 73/, 80c), and two Late Thule cannot be analyzed for blade types. The Birnirk specimens are both small tang sections that were broken just above the shoulder of the cone. One of the Birnirk tangs was found still hafted to a broken arrowshaft. T Y P E IV, OFF-SET KNOBS.?A single level from the Walakpa site had this interesting type of knob (Plate 75c, d). There are two knobs on either side of the tang, one placed slightly above the other. The shoulders are not square but do not slope as much as those in the previous classes. This type of tang has not as yet been reported for North Alaska, but it is found in the Eastern Thule collections from Canada (Ma? thiassen, 1927a: 35-36, pi. 9:4). Only two specimens were recovered and they are incomplete. Therefore, the only measurements to be discussed are the tangs. The total lengths are 2.4 and 3 cm. The lengths of the tangs below the upper knob are 1.7 and 1.6 cm, respectively. No ownership marks occur on these two arrowheads. T Y P E V, TAPERING TANG, WITH A SLIGHT KNOB.? This unique type came from the uppermost levels of the Early Thule period. These can be characterized by a slight shoulder and a very slight knob. Two types of barbs occur. The Early Thule example is a multiple-unilateral barb (Plate 80d). It has one barb present, carved away from the blade structure. Another barb (Plate 84 rH U ffl W ir k c ffl rH % r l M o u n d O c c u p a t io n A-1 A-2 A-3 A-4 A-5 A-6 A-7 B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 B-9 B-10 Late Thule Test Trench Early Thule Test Trench Birni rk Test Trench Totals Perforated St Notched Teeth 1 1 2 1 1 1 1 5 13 Beads 1 1 2 Labrets 1 1 2 Toys 0) id rl ?P Ga me s e a l V e c e s 01 01 01 C M-rl ?H 73 id Q< 01 p & p s e r t e li s y K a y m in g y B o w 3 c o o ffl O CJ H Q trl IS E-l 1 - 1 2 1 - 1 - - 1 - 1 1 - - - _ _ 1 - - - 1 4 - 2 1 3 1 44 2harms & Amulets 5 1 1 1 8 Bear Jaw 1 1 rH > J C 0 -P ft 3 u 0 0 a rH id ?p fH 1 1 1 5 2 2 0 0 1 1 2 2 1 1 3 3 2 3 10 41 1 Two effigies, one from the Early and one from Late Thule occupation, are carved whales (Plates 69g, 82g). The carvings are simply outlines and no fea? tures are apparent other than the flukes. T h e whales measure 21.2 and 4.5 cm in length, and 2.3 and 1.7 cm in maximum width, respectively. The last effigy found at Walakpa is a grass caribou from a Late Thule level (Plate 107a). T h e body is made of grass wrapped by baleen strands. T h e legs are small twigs inserted into the body. T h e caribou does not have antlers. T h e length of the effigy is 6.3 cm and stands 4.4 cm high. T h e body is 1.1 cm in maximum width. Found with the caribou effigy were three small broken arrows (Plate 101 b-d). Al? though comparatively larger, it is believed that they are associated with the caribou and possibly repre? sent some form of sympathetic magic. T h e arrow shafts are all broken but measure in excess of 8.5 cm in length. In diameter they measure 0.4 cm. Fletch? ing still remains on one specimen. T h e fletching con? sists of a small feather wrapped at both ends by a baleen string. The feather, which is split at the center but not at the end, is still attached to the arrowshaft. Bear J a w s An interesting small polar bear jaw is pictured in Plate 56. It is a lower mandible found in a Birnirk level. The jaw is unmodified. A baleen cord is tied around both of the ascending rami. T h e cord was probably used to hang the jaw on a wall or a post. Modern Eskimo use parts of skulls of various animals for charms and amulets (Murdoch, 1892:434, Spencer, 1959:282). This jaw may, in fact, be a Birnirk ver? sion of one of these charms and reflect a variation of the bear cult. % Table 5 is a summary of all of the community ac? tivity artifacts that were recovered in each level at Walakpa. MISCELLANEOUS AND UNIDENTIFIED ARTIFACTS Belt Toggles At Walakpa three examples of belt toggles were found, all of Birnirk age. These are square, flat to pyramidal-shaped artifacts. The flat specimen is made of bone (Plate llh). It measures 3.8 by 3.9 cm in width and 1.2 cm in thickness. One side is" polished with the center being slightly elevated. Near the center on the opposite side, two rectangular holes have been cut at an oblique angle so that they intersect. T h e other specimen, which is made from antler, measures 3.8 by 3.7 cm (Plate l lg) . Its thickness is 1.7 cm. T h e front face has a high center with a ANALYSIS AND CLASSIFICATION OF ARTIFACTS 65 drilled hole, which is inset with an ivory plug. On the reverse side two oblique holes are drilled near the center. The third Birnirk belt toggle looks like a quarter of a sphere with two flat sides. The rounded outer surface has a ridge at the center. One side is polished while the other side has the two obliquely cut, rec? tangular holes. This ivory specimen measures 3.6 cm in length, with a width of 1.1 cm. Only one example of an Early Thule toggle was recovered. It is an incomplete bone specimen. In overall shape it is rectangular, with one end curved, on which there is a small slot, while at the other end a large hole. The toggle measures 2.6 cm in length, 1.2 cm in width, and 1.2 cm thick. Ivory Chains An extremely well-carved chain with three oblong interlocking links was found in the lowest Birnirk level (Plate 36a). Its maximum length is 8.7 cm, with the links being 4, 3.6, and 3.2 cm, respectively. It appears to be identical with those reported by Larsen and Rainey (1948, pi. 70:1-3) from the Ipiutak site. A single ivory link or circle was found. It is also Birnirk in age. This ivory artifact is nearly circular and measures 2.1 cm on its outside diameter (Plate 51m). Unidenti f ied Ivory Art i fac t s Many ivory objects, which are probably parts of other artifacts or ornaments, were recovered at Walakpa. One of the most interesting is a thin piece of ivory with a small hole bored into its center (Plate 36b). At either end, and pointing in opposite direc? tions, are two small projections, which are reminiscent of half of a whale's fluke. On both edges two sets of bifurcated appendages are carved. Its length is 4.15 cm and its maximum width is 0.98 cm, with the thickness being 0.52 cm. It appears to be a zoo- morphic figure, such as two seals or whales, facing opposite directions. Another interesting artifact is a thin strip of ivory which has been curved like a bracelet (Plate 36c). It has a series of wrapped baleen strands tied to either end. Near the apex of the curve, a slight nipple has been left on one edge with a shallow cut in its center. It is impossible to specify the function of this tool, but I feel that it is probably a saw similar to a hack saw. The baleen strings, acting along with a fine grit sand, would constitute the blade, and tension could be regulated by pressure applied to the ivory strip. T o support this identification, a similar baleen string saw was found at the Utkiavik site. On the Utkiavik specimen the blade was broken in the process of cutting ivory, and it was left in its original position when it was discarded. An interesting Late Thule artifact is a thin ivory cylinder measuring 4.4 cm in length (Plate 112d). Its maximum diameter is 0.8 cm. One end is ground to a point with two flat sides and a hole has been drilled, which goes through both sides. The opposite end is truncated with a slightly raised lip going completely around the circumference. The lip is slightly larger and pointed at one end. A socket has been drilled in the truncated surface. This could possibly be a toy harpoon socket piece. An unidentified Late Thule artifact may be a thimble holder. It is a highly decorated piece of ivory. The edges have three scallops which terminate in a blunt point. A square hole is cut near the point. At the other end, a slot is cut up the center of the tool shank. Prongs to either side of the slot are broken. Incised lines run along all of the edges. Two short lines run from the tips of the scallops toward the center of the artifact. Five short lines are found around one side of the rectangular cut. Two designs, which suggest ownership, are located above the end slot. The tool is oval in outline and weald have measured in excess of 7.8 cm. It is 1.4 cm at its widest point and 0.7 cm thick. Another Late Thule ivory object could be a cup- shaped, fat scraper or a chin rest for a bow drill (Plate 103a). It is a thin, 0.3 cm, concavely carved artifact. It is 7.1 cm long and 4.7 cm wide. All edges are rounded and polished. A cut, which is keyhole in shape, is placed near the center of the artifact. Miscellaneous Coal A single Late Thule disk is made of fine-grained bituminous coal (Plate 85o). It may be a labret. It is quite thin, measuring 0.6 cm. The diameter is 3.2 cm. It is ground flat with holes drilled into the center from both sides, but not meeting. Unidentif ied Bone Art i fac t s Two unidentified Birnirk bone tools were found. One is a long, thin, highly polished tool (Plate 51r). It is 13.4 cm long, 1.3 cm wide, and 0.7 cm thick. The edges of the proximal end have two scallops. The distal end comes to a sharp, flat end. Another bone artifact is a long, 18.3 cm, pointed peg (Plate 8d). It is 1.3 cm wide, 1.8 cm thick, and is 66 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 oval in cross-section. The proximal end is crushed and appears to have been beaten with a hammer. T h e distal end is pointed. A baleen string is tied around the center of the peg. Unident i f ied W o o d e n Art i fac t s Seventeen unidentified artifacts were found in the Birnirk levels. A short, wooden handle-like object was recovered (Plate 51 d). It has a square end with a hole drilled through the center. Behind the square end the handle is round and comes to a blunt end. The tool measures 18.5 cm in length. An object shaped like a propeller measures 11 cm in length and 1.1 cm at its maximum width (Plate 11/). It is lenticular in cross-section. It is carved from a piece of wood that contained a knot hole. The knot is at the center of the artifact and has been removed. T o either side of the hole there are short, tapering, flat wings, which terminate in rounded ends. A long, thin stick measuring more than 20.5 cm in length and 1.7 cm in diameter was recovered (Plate 51 e). Two notches are carved on one edge, and one on each side of the pointed end. Several unmodified wooden sticks and pieces of wood were found in the Birnirk levels (Plates \4a,e, 36/). These pieces of wood were tied together and undoubtedly made up part of a more complex tool. Four unidentified Early Thule artifacts were made out of wood. The first is a split tool, which measures in excess of 12.7 cm in length, 4.5 cm in width, and 3.3 cm in thickness (Plate 19b). On one surface a large hole measuring 2.8 cm in diameter was drilled near the center. Two small holes are drilled to either side of the large hole. On one edge, two holes are drilled even with, and at a 90? angle to, the small holes on the artifact. Another broken piece of wood is rectangular in cross-section (Plate 80/). It measures 2.1 cm in width and 0.8 cm thick. The end that is not broken is cut to a tip with a slight shoulder and has a hole drilled in it. A large piece of wood has a truncated, oval shape. One end is beveled with a notch for adjoining to another artifact. Discoloration at this end of the artifact indicates that it has been lashed around the beveled portion. It measures 2.2 cm in width and 2 cm in thickness. An interesting type of artifact of unknown use was found in both the Early and Late Thule levels (Plates 66/, 14n, 16h, 80g, 92h, 112ft). These are small shafts that resemble arrowshafts. They have a nock end with V-shaped cuts. The opposite ends are expanded at the tip. It may be that they are small toy arrows. However, one has a cord tied around the center and the tip has been burned (Plate 66/). They may be small, bow-drill spindles, but if this is the case I can? not account for the nock ends. Also the string around the specimen depicted in Plate 66/ is not set up for a bow-drill string. It is looped on one side and the unattached ends are both on the same side. These range in length from 18 to II cm and 0.6 to 0.3 cm at their maximum diameters. Two wooden artifacts of Late Thule age may be small cases for holding needles, end blades or other objects. The larger specimen is broken (Plate 83e). This case looks as though it was made from a broken wound plug. It is 9.8 cm long. Its maximum width is unknown. The center part of the case is carved hollow. There is a notch around the wide end, which possibly held a lid over the top. T h e other end is cut out and would not serve as a containing end. The second case is extremely small (Plate 85p). It is 2.8 cm long, 1 cm wide, and 0.5 cm thick. The ends are rounded with lashing grooves cut around the outside surface. T h e inside is hollowed out. A chain of five wooden balls are of Late Thule age (Plate 84i). The chain is 6.2 cm long and 1.7 cm in diameter. It is highly reminiscent of the chain pend? ant pictured by Mathiassen (1927a, pi. 30:18). How? ever, it does not have any suspension mechanism. Two short sections of wood that are the same kind of artifact was found in a Late Thule level (Plates 84&, 105n). The complete specimen is 16.2 cm in length. They are both 2.2 cm wide, and 2 cm and 1.7 cm thick, respectively. They are ovoid in cross- section. At opposite sides on both ends the wood is cut in a curved fashion to attach an elliptical object. Behind the cuts on both edges a hole is drilled. At the center, another hole is partially drilled for a brace. A thin stick is cut on both ends and resembles a crude toy bow (Plate 118ft). It is 12.5 cm long and 0.8 cm wide. A long, thin spatulate-shaped artifact measures 44.5 cm in length. At its maximum width it measures 2.8 cm; its thickness is 0.8 cm. T h e distal end of the artifact is blade-like, with its widest portion near the blunt tip. At the proximal end there is a shoulder with a tang made from a cut on both edges. The width of the shoulder is 2 cm, the width of the tang is 1.4 cm, and its length is 1.8 cm. A hole has been drilled through the tang. A broken artifact with an elliptical cross-section measures 2.5 cm wide and 1.5 cm thick (Plate 105m). The only unbroken end has a rounded, blunt end. Four notches are carved on both faces of the artifact below the end. ANALYSIS AND CLASSIFICATION OF ARTIFACTS 67 A broken piece of wood that measures in excess of 38 cm has four large square holes cut into one face (Plate 83i). These squares are 1.8 cm by 2.6 cm. A portion of what appears to be a round wooden ball was also recovered. In diameter it would have measured 9.1 cm. The rounded surface is roughly cut and not finished. A split piece of wood gives the appearance of a spoon. It measures 9.3 cm in length. At its maximum width it is 1.7 cm, and its minimum width is 0.6 cm. In thickness it is 0.3 cm. I do not believe it is a spoon as the surface is broken and it appears that it may have been originally round in cross-section. An example of what the above-described artifact may have been was also recovered. It is a round shaft but at one end it widens out considerably then tapers back to a blunt point. The end is broken so its length is unknown, but it is more than 11 cm. The diameter at the greatest point of expansion is 1.8 cm, while the diameter of the narrow end of the shaft is 1.2 cm. Another paddle or spatulate-shaped object was found. It is carved flat on one face, while the rest of the artifact is rounded. A long, round handle is broken. The other end is oval in outline. It would have measured in excess of 14.3 cm in length. The widest point of the blade measures 1.5 cm. The handle is 1.1 cm in width and 0.7 cm in thickness. Unident i f ied Baleen Art i fac t s One of the most unique unidentified baleen speci? mens comes from the Birnirk occupation (Plate 14c). It is a circle made of a baleen strip with baleen wrapped around it. In diameter it measures 10.2 cm. Another piece of baleen is a long parallel-sided strip of baleen, which is bent into a square crook-shape. The bottom of the crook is drilled with a baleen swivel passing through the hole and attached to the circle. The end of the crook is notched and tied back to the main strip with a baleen cord. The end of the strip is also notched. In tlength the strip measures 28.9 cm. It is 2.4 cm wide. Another piece of baleen measuring 24.7 cm long by 5.1 cm wide has both ends rounded (Plate 36ft). It appears similar to a bull roarer such as Ford (1959: 226, fig. 112b) describes, but it has not been drilled for suspension. In the Early Thule levels three silhouettes of un? known objects have been cut out of baleen. One, measuring 22.8 cm long and 1.4 cm at its widest point, looks like a long knife blade and handle (Plate 69ft). The blade is lunate in shape and terminates with a point. The handle is rectangular and is set off from the blade by notches cut on both edges. The second specimen is 17.3 cm long and 2.3 cm at its widest portion (Plate 70c?). The blade is also lunate in shape with a sharp point. The handle is narrower than the blade, but expands at its base. The third specimen also has a lunate blade and a rectangular handle (Plate 66e). However, the handle is wider than the blade. This artifact measures 13 cm in length. Its maximum width is 1.2 cm. Another silhouette is probably part of a more com? plex tool (Plate 69/). It has an elongated diamond shape measuring 16.5 cm long. At its widest portion it measures 2.5 cm. Six holes have been drilled in a straight line, evenly spaced from one point to the other. It is bent at the center, causing the points to turn upward. Another baleen strip has been tied in a small circle. It measures 6.2 by 4.1 cm in diameter (Plate 69b). Inside the circle is a webbing of baleen string with a small sliver of wood woven into the web. Very little baleen of any kind was found in the Late Thule levels. Among those pieces that were found were two elongated, diamond-shaped silhouettes like the one described from the Early Thule period (Plates 92z, 112a). The major difference being that these two do not have the drilled holes. They measure 12.2 and 10.2 cm in length. Their maximum widths are 1.7 and 1.6 cm, respectively. A silhouette shaped like a knife also occurred in the Late Thule period (Plate 100c). It has a lunate blade and a short, narrow handle. It measures 18.7 cm in length. The widest portion of the blade measures 3.6 cm. Coils of baleen lashing were found from each oc? cupational period (Plates 14d, 66ft, lll-n, 85q,r, 1126). Several skin lashing cords were also found. The cord depicted in Plate 57a has a small loop on one end. The one in Plate 14b has a loop on one end. The one in Plate 85q, has a loop on one end and a small twig at the other end. Miscellaneous Stone Art i fac t s Thin, bifacially flaked, flint preforms were found in both Birnirk and Early Thule levels. Fifteen are Birnirk (Plates 43b,c,i, 52i), two are Early Thule (Plate 65a), and one is Late Thule (Plate 90;) in age. These are relatively thin pieces of stone, which appear to be exhausted cores. Many show use on the edges. They have not been assigned to any formal artifact class as the use patterns may have resulted from being used as scrapers, knives, or adzes. They range in size from 14.1 by 13.7 cm to 3.8 by 3.7 cm. All are made from fine-grained quartzite, which can be found along the beach at Walakpa. 68 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 Two exhausted conical cores were found in Late Thule levels. Both are made from Sik-sik-puk chert and do not show any indication of secondary use. The complete specimen measures 6 cm long, 5.5 cm wide, and 2 cm thick. Two large stone artifacts were found in the Birnirk levels (Plates 23d, 44a). These are both made of slate. They have been flaked into the desired shape and then the flat surfaces were partially ground. Neither has sharp edges although one has a slightly beveled edge. They resemble gigantic ulu blades, but do not seem to have functioned in that capacity. Among the stone artifacts which have not been identified are a series of slate nodules (Plates 29/, 45d, 53/, 16a,b, 114a). These nodules occurred in all three occupation units. They are all of the same general shape. They are round to oval in cross- section, with a central portion having a larger diam? eter than the ends. The ends are flat to rounded. Some ends were naturally flat but others were ground to this shape. Any irregularity on the stone is ground away to leave the circular cross-section. They range in size from 20 to 4 cm in length, with an average of 7.6 cm. In maximum diameter they range in size from 6.2 to 1.2 cm, with an average of 2.7 cm. Several small hematite nodules were found. Three were Birnirk (Plates 29e, 53c) and three were Late Thule in age. These have been ground, producing faceted surfaces. T h e grinding was presumably done to get the fine, red powder for making red paint. A notched quartzite pebble was found in a Birnirk level (Plate 29b). It measures 3.7 by 2.8 cm. The notch runs around the entire center of the long axis. It is flat on one surface. T h e pebble may have been used for a sinker, a bola weight, or possibly an atlatl weight. Two small, flat, sandstone pebbles, one Birnirk (Plate 45c), and one Early Thule (Plate 61;) in age, have a hole drilled through one edge. They may have been used for sinkers. T h e larger specimen measures 5.7 cm and the smaller specimen is 2.8 cm in diameter. The debitage produced in flaking chert or flint artifacts from Walakpa was very sparse. A total of 816 flakes were found in all the levels combined. Faunal Analysis Introduct ion The excavations at Walakpa produced over 30,000 animal bones. In addition, a few fish scales, gastro? pods, egg shells, and quantities of oil-soaked fur and feathers were recovered from the prehistoric levels, all of which have not been identified as to species. Only 18,523 bones were used for the faunal sample, as the others were from disturbed areas of the site and historic levels that are not a part of this study. Ninety-two percent of the 18,523 bones under study, or 17,066 bones of the faunal sample, were identified as to species and bone elements. Over one-third of the 1457 unidentified specimens were bird bones. Procedure Identification of faunal elements was established by comparing each bone or bone fragment to specimens in the Department of Biology reference collection of the University of New Mexico. The comparative osteological collections available included complete skeletons of common seal, fur seal, bearded seal, arctic fox, and musk ox. A coyote skeleton was used to determine the dog bones; black and grizzly bear skeletons were used for the polar bear; and a mule deer skeleton substituted for caribou. These were supplemented by miscellaneous caribou, walrus, and polar bear elements collected at Point Barrow. All cervid remains of deer size, and too small to be moose, were assumed to be caribou Rangifer arcticus. Similarly, all small seal bones were considered to be Phoca spp., while those of medium large size were considered to be bearded seal, and those too large for bearded seal were referred to as walrus. There was no reference skeletal material available for whales at the time of the study. This material was dealt with by measuring the forelimb bones and extrapolating size by comparisons with illustrations of whale skeletons. In the case of small whales, the ratio of the length of the humerus to the total length of the animal is about 1 to 25 (Romer, 1966). All humeri material from Walakpa seems to belong to this small whale. No evidence was found for larger whales. The available bird reference collection included: (1) loon, (2) black brant, (3) Canadian goose, (4) snow goose, (5) pin-tailed duck, (6) old squaw, (7) buffle- head, (8) merganser, (9) grouse, and (10) gull. Marine material was absent and references made to such species is by inference only. Identifications in the case of birds was attempted only for selected elements that are the most distinctive in the different species. The elements used were the skull, corocoid, furculum, humerus, tibia, tarsal-metatarsal, synsacrum, and sternum. Bones that had been modified into tools or that showed use-marks were considered as artifacts and were not included in the faunal identifications or count. Both butchering marks and dog-tooth marks were observed on many of the bones but no attempt was made to record them formally. The faunal remains were recorded by the prov? enance units of level and square. Where the bone concentrations were dense, each provenance unit was recorded on a separate data form. In many cases so few bones and fragments were present that they were merely listed. Bones were sorted into species, then into specific bone elements, after which the number of each ele? ment for each species was recorded together with other information outlined below. Eleven species of mammals were found at Walakpa: (1) ground squirrel, Citellus spp.; (2) polar bear, Thalarctos maritimus; (3) fox, Alopex spp., (4) wolf, Canis lupis; (5) dog, Canis familarus; (6) common seal, Phoca spp.; (7) bearded seal, Erignathus bar- batus; (8) walrus, Odobenus rosmarus; (9) whale; (10) caribou, Rangifer arcticus; (11) musk ox, Ovibos moschatus (Table 6). Twelve species and four families of birds found were classified and include: (1) loon, Gavia spp.; (2) brant, Branta spp.;(3) Canada goose, Branta cana? densis; (4) snow goose, Chen hyperborea; (5) pintail duck, Ana acuta; (6) shoveler, Spatula clypeata; (1) scaup, Aythya marila; (8) old squaw, Clangula hye- malis; (9) bufflehead, Bucephala albeola; (10) king eider, Somateria spectabilis; (11) common eider, Somateria mollissima; (12) ducks, Anseridae; (13) ptarmigan, Lagopus spp.; (14) shore birds, Charad- riidae; (15) gulls, Laridae; (16) thick biller murre, Uria lomvia; (17) owls, Strigidae (Table 7). 69 70 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY N U M B E R 2 0 TABLE 6.?Distribution of bones of various species of animals by number and percentage from the various occupation levels of each culture phase, Walakpa site Culture Phase, Mound, and Occupation Level u le fl 0) -P J Th ul e E ar ly ? ?H C ?rH CQ A-1 A-2 A-3 A-4 A-5 A-6 A-7 B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 B-9 B-10 Totals % of Totals Ground Squirrel No. % 6 .6 - - - - _ - - - - - 1 .1 - - - 7 .3 NO 3 - 1 5 3 1 - 1 2 - 1 2 4 9 2 34 - Bear % .1 .2 .3 .1 .2 - .4 .7 .2 .3 .6 .2 .1 .2 No. 2 1 4 4 8 4 3 - 1 - 1 2 4 26 112 172 - Fox % .2 - .2 .7 .2 .7 1.1 .6 .3 - .2 .3 .6 .7 2.2 .9 No. 13 4 4 3 8 12 14 3 1 4 3 2 6 13 63 143 296 Doq % 1.4 .2 1.1 .5 .5 1.1 3.9 .6 .4 1.4 1.1 .7 - .9 1.9 1.7 2.8 - 1.6 2 3 9 Seal No. 385 508 255 242 424 446 106 230 121 159 172 198 274 340 404 ,020 ,193 ,477 - % 41 29.2 68.8 43.8 26.8 46.5 29.8 47.6 54.5 56.5 60.5 70.5 71 55.1 59.3 55 62.6 - 51.2 Bearded Seal No. 13 10 6 3 6 4 3 4 1 1 2 2 6 13 70 57 201 % 1.4 .6 1.6 .5 .3 .3 .8 .8 .4 .3 .7 - .5 .9 1.9 1.9 1.1 1.1 Walrus No 5 2 1 - 3 5 1 2 1 2 2 1 8 10 52 4 99 - ? 1 1 1 % .5 .1 .2 - .2 .4 .2 .4 .3 .7 .7 .2 .3 .4 .4 .1 5 CRANIAL ELEMENTS.?Maxillae and mandibles for each species were recorded as right or left. Fragments were also sorted into right and left when possible. When the tooth row was at least partially present, jaws were recorded as to their probable age. T h e clas? sification used (juvenile, mature, or old) was based on a subjective examination of the degree of tooth wear. Initially, an unsuccessful attempt was made to determine ages of caribou from the faunal remains based on tooth eruption, wear, and shedding. Our tentative classification was based on those of Bochoud (1966), Skoog (1968), and Lucifer (n.d.). During the summer of 1971, Dr. Lewis R. Binford of the Uni? versity of New Mexico made a collection of caribou mandibles at Anaktuvuk Pass, Alaska. These were animals killed during a known year and season and were of known ages. This collection was provided for our use and proved invaluable for age determina? tion of the caribou of comparable time periods. How? ever, caribou killed during the Birnirk time period did not have the same type of eruption and wear as did the latter examples. This difference in wear may be attributed to a different climatic episode, a shift in the caribou migration schedule or changes in territory, which allowed the animals to graze on softer feeds in early spring and late fall. These pos? sible changes may have caused the teeth to be less highly worn than on those animals that did not graze on softer vegetation so long into the year. LONG BONES.?Long bones were recorded as right F A U N A L ANALYSIS 71 41 Whale No. % 3 .3 5 .2 - - 3 .2 7 .6 3 .8 1 .2 2 .9 1 .2 2 .3 4 .6 7 .2 3 .1 Caribou No. % 473 955 97 244 974 380 212 151 65 74 67 50 44 194 127 960 1,009 50.5 55 26.2 44.2 61.5 38.7 59.3 31.3 29.2 26.3 23.5 17.8 11.4 31.4 18.6 26.2 19.1 Musk Ox No. % - - 1 .2 - - - - - - - - 1 .2 - - - 1 .1 Birds No. 8 23 6 29 22 17 5 2 14 5 7 38 32 44 120 288 % .8 1.3 1.6 5.2 1.4 1.6 1 .9 4.9 1.7 2.1 9.9 5.2 6.5 3.3 5.6 Unidentified No. % 29 230 27 135 101 13 84 29 27 33 22 23 24 58 335 287 3.1 13.2 4.9 8.5 10.2 3.6 17.4 13.1 9.6 11.6 7.9 5.9 3.9 8.5 9.2 5.6 Total Per Level 1 1 3 5 937 .740 371 553 ,584 983 357 483 222 282 285 281 386 617 681 662 099 % of Total Count 5.1 9.4 2 3 8.5 5.3 1.9 2.6 1.2 1.5 1.5 1.5 2.1 3.3 3.7 9.7 7.7 6,076 660 1457 32.8 . 01 3.5 7 . 9 18,523 100 and left. On the humerus and femur the degree of epiphyseal fusion was recorded. The fusion was recorded in three categories: unfused, partially, and completely fused. When possible, proximal and distal fragments were treated as if they were complete bones, but if they were badly broken they were recorded as long-bone fragments. Shank fragments were identified as to ele? ment whenever possible. When this was not possible, they were recorded in the general category of long- bone fragments. Due to the lack of size overlap in the populations present, fragments could usually be assigned to a species if not to the particular bone element. Caribou metapodials were recorded with the long bones as right or left metatarsals or metacarpals. Fragments were sorted into front and rear legs when possible; otherwise they were recorded as metapodial fragments. Seal metapodials were considered as miscellaneous flipper elements instead of long bones, because they are not long bones in a morphological sense. BODY ELEMENTS.?Pelvic elements were recorded as complete pelvis, ilium, ischium-pubis, acetabulum, or pelvic fragments. These were also identified as to right or left side when possible. Vertebrae were recorded in the categories of cervi? cal, thoracic, lumbar, sacral, and caudal. Fragments were also recorded in these categories whenever pos- 72 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 TABLE 7.?Distribution of bones of the various species of birds from the various occupation levels of each culture phase, Walakpa site Culture Phase, Mound, and Occupation Level m le L at e TT r T hu le rly rd H tf a u m A-1 A-2 A-3 A-4 A-5 A-6 A-7 B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 B-9 B-10 Totals % of Totals Loon No. % - 2 6.9 1 4.5 - - - 1 7.1 1 14.3 - - - 5 1.7 10 - 1.5 Brant No. % - - - - - - - - - - - - 5 1.7 5 .7 No 1 - - - 1 1 4 2 2 5 3 20 - Goose % 3.3 5.9 - - - 7.1 14.3 10.6 6.2 4.5 4.2 1. 3.1 Snow Goose No. % - 1 4.3 _ - - - - - - 2 1.6 3 .4 Pintail Duck No. % 1 12.5 - 2 9.1 - 1 20. _ _ - - - - - 4 .6 Shoveler No. % - _ 2 6.9 - - - _ - - - - - - 2 .3 Scaup No. % 1 12.5 - 3 10.4 2 9.1 - 1 20. . ; - - - 6 2.1 13 1.9 Old Squaw No. % - - _ - 3 13.6 - - 1 7.1 2 40. 1 14.3 - - 5 11.3 3 2.5 3 1. 18 2.7 Bufflehead No. % - - _ _ - - _ _ - 1 .3 1 .1 sible; otherwise they were called miscellaneous verte? brae fragments. Ribs "were recorded in three categories: complete, proximal, and fragmented. FOOT BONES.?Tarsal and carpals were not sepa? rated but were recorded as foot bones. T h e cubo- navicular was recorded separately. T h e metapodials were not subdivided except in the case of the caribou as noted above. First and second phalanges, both front and rear, were recorded as one category. The third phalange, or ungle, was recorded separately. Proximal and distal fragments of the metapodials and phalanges were also recorded. Measurements In addition to classification, a set of measurements was taken on selected elements and recorded on the same data forms. Dog skulls were measured to see if there were any differences in sizes. Seal femurs, humeri, and tibias were measured for maximum length and maximum width at the distal ends. SEAL LONG-BONE MEASUREMENTS.?It has been pointed out by McLaren (1962:177) that the size of seals is directly proportional to the length of the seal pup suckling period. T h e suckling period is termi? nated when the ice breaks up in the spring. There- FAUNAL ANALYSIS 73 King Eider No. % - - - - _ - 23 8. 23 3.5 Common No. 5 21 6 21 13 13 - 2 9 3 4 34 15 20 103 227 496 Eider % 62.5 91.5 100. 72.4 59.3 76.5 - 40. 64. 60. 57.1 89.4 46.9 45.5 86. 78.9 75.5 Ducks No. % 1 12.5 - _ - 1 4.5 3 17.6 _ _ 1 20. - - - - - _ - 2 .7 8 1.2 Ptarmiqan No. % - _ - - - - - _ _ - 1 3.1 - 1 .1 Shore Birds No. % - - - - 2 100. 2 14.2 - - - _ - 1 .3 5 .7 Gulls No. % - - - - - - - - - _ 5 4.2 12 4.2 17 2.6 Thick Billed Murre No. % - - - - - - - 14 43.8 17 38.6 2 1-6 - 33 5. Owls No. % - 1 4.3 - - _ _ - - - - - 1 .1 Total Per Level 8 23 6 29 22 17 - 5 2 14 5 7 38 32 44 120 288 660 % of Total Count 1.2 3.5 .9 4.4 3.3 2.6 - .7 .3 2.1 .7 1. 5.8 4.9 6.7 18.3 43.6 100. fore, if the ice breaks up earlier in the year the pups will be smaller, and if the ice breaks up later in the year the pups will be larger (McLaren, 1962:177). If the rise and fall of the sea level at Point Barrow indicates climatic changes, these changes would affect the seal suckling period, which would be reflected by variations in the sizes of seal bones found in the dif? ferent occupation levels at Walakpa. Measurements were then taken on seal long bones in order to see if in fact there was a correlation between seal size and the proposed sea levels or climatic episodes. As it has been demonstrated by Usher and Church (1969:126) that there is a relationship between seal length and girth that can be used to predict seal weight, a biometric analysis of seal long bones was made on seal specimens coming from various archeo? logical levels. If there were any significant changes in seal sizes, it was hoped that they would correlate with proposed climatic shifts. Biometric measurements were taken on seal humeri, femurs, and tibias. Measurements taken on the humerus were articu? lar length and distal condylar width (Figure 9), on the femur were the notch length and the distal condylar width (Figure 10), and on the tibia, for diaphyseal length (Figure 11). The major problem of this experiment was that the sample size was too small. 74 Articular s Length Distal Condylar Axis SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY I NUMBER 2 0 Distal Condylar Width FIGURE 9.?Seal humerus showing the distal condylar width and articular length. As juvenile variability would skew the result, only the bones of adult animals were used. Also the bones had to be complete. Both of these features reduced the sample size to the following: humerus?Birnirk (40), Early Thule (14), Late Thule (12); femur?Birnirk (66), Early Thule (22), Late Thule (0); tibia?Birnirk (32), Early Thule (5), and Late Thule (14). Notch Length Distal Condyle Axis Distal Condylar Width FIGURE 10.?Seal femur showing distal condylar width and notch length. )? Diaphyseal Length FIGURE 11.?Seal tibia showing diaphyseal length. The means of the measurements of the humeri widths are 2.39 cm (Birnirk), 2.39 cm (Early Thule), and 2.38 cm (Late Thule) . T h e lengths are 9.42 cm (Birnirk), 9.26 cm (Early Thule) , and 9.47 cm (Late Thule). This indicates that the humerus width did not vary in the three phases examined. However, there is a high correlation between the lengths of the Birnirk and Late Thule lengths, while the Early Thule lengths are significantly different from the other two. This may indicate that suckling periods for seals during the Birnirk and Late Thule ages may have produced seals that were longer than those of the Early Thule period. The mean femur width for Birnirk is 3.6 cm, while for Early Thule is 3.7 cm. T h e length mean was 7.63 cm for Birnirk and 7.7 cm for Early Thule, respectively. The tibia diaphyseal lengths for Birnirk had a mean of 15.4 cm, while for Early Thule the mean was 14.44 cm. Late Thule was again longer with a mean of 15.7 cm. Here, as in the measurements for the FAUNAL ANALYSIS 75 humerus, we see that the Birnirk and Late Thule seals appear to be longer than those from the Early Thule levels. It should be pointed out that the slight tendency of correlation of longer seal humeri in the Birnirk and Late Thule levels does correlate with cooler climatic periods and the lowering of the sea level. However, this may be insignificant, as the sample is small. The other problem that, must be considered is that no work has been done on the osteological remains of modern seals from various localities to see if in fact the bones reflect the body size as reported by McLaren (1962), and Usher and Church (1969). DOG SKULL MEASUREMENTS.?The measurements of dog skulls from the Birnirk levels at Walakpa indi? cate that there is a possible bimodality in the dog population. The most common size of dog is quite small as compared to modern Greenland sledge dogs. On ten selected adult animals the condylo-basal length ranges from 15.8 to 17.7 cm. The zygomatic breadth ranges from 10.7 to 9 cm. The palatal lengths range from 9.1 to 7.9 cm, and the upper carnassials range from 1.9 to 1.6 cm. They are probably the same type of dog as those at Ipiutak, but are slightly smaller. They fall into the range of Siberian dogs and Samoyeds. However, a single Birnirk-age specimen, which may be a wolf, has a condylo-basal length of 23 cm. This dog compares more favorably to the size of Greenland sledge dogs and those from Labrador as measured by Olaus Murie (in Larsen and Rainey, 1948:255-259). NUMBER OF ANIMALS PRESENT.?In order to deter? mine the minimum number of individual animals of each species from each occupation level, the maxi? mum number of animals represented by the absolute number of each skeletal element was calculated. Special attention was paid to the number of right and left sides and distal and proximal ends of bones with maximum possible pairs being calculated. The ele? ment with the largest number per level was taken to represent the minimum number of accountable indi? vidual animals present in each occupation level. I believe this estimate of minimum numbers of individuals per species is fairly accurate; however, it may be that the number of individuals was higher. Whole bodies or body parts may have been removed to other campsites, or there may have been errors in recovery. The important point, however, is that the basic calculation is on the low conservative side. Analys is b y Archeologica l Level and C u l t u r a l U n i t s The analysis of the horizontal distribution of the bones in the various levels was not particularly useful. Most bones were found in association with trash middens, which were irregularly structured with ele? ments introduced from discarded portions of meat when the animal was butchered and from choice meat packets when they were either consumed or processed for the drying racks. There was little evidence that the middens were disturbed by dogs or wolves. Few bones showed extensive scars from being chewed by these animals. The evidence suggests that dogs were present at the site and were staked out in an area that was not excavated. As seal and caribou make up the primary animals utilized by the Walakpa inhabitants, the elements of these animals have been grouped into "meat packets." The meat packets are differentiated by the amount of available meat occurring on unit cuts. The unit cuts are large meat packets (upper fore and hind limbs), small meat packets (lower fore and hind limbs, trunk and certain head portions), and waste packets (hooves, feet and certain head portions). The presence or absence of these parts in each occupation level was tabulated as a basis for establishing the minimum number of individual animals represented. When the minimum number of animals is known, the maximum number of possible elements can be deter? mined. The percentage of the number of bones re? covered, calculated with respect to the expected num? ber of bones found in a single animal, and multiplied by the estimated number of animals, should indicate what portions of the animals are present or absent. It is important to consider several variables when discussing the meaning of the occurrence or non? occurrence of bones representing meat packets, be? cause they reflect to a great extent the time of year during which the animal was killed; therefore they are potentially significant. The variables are (1) the type of camp, i.e., a base camp or hunting camp; (2) the distance from the camp to the kill or kills; (3) the number of animals in the kill or kills; (4) the num? ber of people available to process the meat; (5) the mode of transportation available to carry the meat back to camp, i.e., sled, boat, pack dog or by hand/ foot; (6) weather conditions, e.g., windy, icy, snowy; (7) ice and snow conditions; (8) whether or not the meat was frozen when butchered; (9) what the ulti? mate destination might be, i.e., immediate or near- future consumption by man and/or dogs, storage for later consumption, drying or ice-cellar freezing; (10) season of animal year (if antler or hide is not suitable for artifact manufacture due to seasonal changes, it will be discarded). Several assumptions based on these variables and on observations of Eskimo butchering practices have been made about the significance of meat packets. 76 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 One assumption is that if the animals were killed at or near the site, most, if not all of the elements, will be represented at some time at the site. Whereas, if the animals were killed far from the site, many waste portions such as the feet and certain head elements will not be taken back to the camp. If antlers, tongues, brains, and noses were desired, they would be extracted in the field. This would eliminate the necessity of carrying many extra pounds back to the camp. This is particularly likely when there is no transportation available, or when the number of animals killed exceeds the capacity of a sled or boat. In this case, even large meat packets will be dis? carded. In some cases, meat will be cut from the bones in the field leaving heavy elements such as humeri or femurs at the kill site. When a level represents a base camp, most of the elements from the large meat packets should be present. Irregularities in the numbers of elements found in a base camp probably represent meat con? sumed, dried, or fleshed out at the hunting camps, or stored in ice cellars to be consumed later. Waste cuts found in the base camp probably represent an oc? casional animal killed near the base camp and/or portions introduced for dog food. The evidence also indicates that when dealing with a hunting camp, particularly a seasonal task-specific hunting camp, the bones found in extraordinary numbers represent that specific activity, while bones from other species found in small quantities were introduced into the hunting camp from the base camp. That is, provisions were brought from the base camps to sustain the hunter for the duration of the encampment or until he obtained fresh meat. Certain elements may also have been introduced to provide raw materials for the manufacture of tools. Base camps should be central collection points with meat being imported from various task-specific hunt? ing camps. Although these camps are considered as primarily winter sites, bones should represent ani? mals that are killed during all seasons of the year. SEALS From examining the occurrence of seal elements from Walakpa, we find two general trends (Figures 12-16). The first is found in levels A-1 through B-8. Here we see a low percentage of skulls, relatively low percentage of the trunk bones and high percentages of leg and flipper bones. In levels B-9 and B-10 we see nearly the reverse of this trend, with skulls, lumbar vertebrae, and ribs being high. The larger leg meat packets are also high in relationship to -the smaller meat packets, and the flippers are extremely low. It is also observed that the number of individual seals in levels A-1 through B-8 is very low, while in B-9 and B-10 it is extremely high. This has been inter? preted to mean that the upper levels are hunting camp sites of brief duration, while the lower levels represent base camps. The scarcity of skulls from the levels that represent hunting camps could be explained by the fact that the skull is prized for both the brain and eyes (Spencer, 1959:373). These are apparently taken back to the base camp. Thirty-three to 58 percent of the skulls from the minimum number of seals killed were found in the base camp levels. This is extremely high and supports the hypothesis expressed above. This is also affirmed by Murdoch (1892:434), who indicates that skulls are usually brought home and regarded as sacred. The trunk, as represented by the lumbar vertebrae and ribs, does not figure significantly in the hunting camp levels, but a relatively high percentage of these were found in the base camp levels. Spencer (1959: 373) indicates that the choicest portions of seals are the heart, liver, and shortribs. This would account for absence of lumbar vertebrae and ribs from the hunting camps and their presence in the base camp levels. No mention is made by either Spencer or Murdoch of the leg and flipper cuts. It appears from our data that the back leg bones are introduced into the base camp more frequently than the front leg bones. Ap? parently the back leg and flippers are either con? sumed or left at the hunting camp. In former days dogs were fed blubber and seal entrails (Spencer, 1959:374). Rarely was a dog given seal meat; but the blubber was an important source of dog food, as it took little of it as compared to either seal meat or caribou meat to sustain a dog. The archeological significance of the change in size of seal elements between the lower B levels and the upper B levels can only be assessed by examina? tion of the population dynamics of modern seals. McLaren's (1962) work on seals in the eastern Canadian Arctic has profound implications for the seal analysis at Walakpa. Even though there is a pos? sible bimodality of seal sizes at Walakpa, I believe that these animals are all the Phoca hispida or ringed seal. The ringed seal is the most common seal found in these northern latitudes, and is the only species of small seal that currently resides the full year in these waters (Burns, 1970:447). T h e size variation is not only influenced by sex and age, but also by ice conditions. The size of adult seals from high arctic localities averages 15 percent longer than that of adults from southeast Baffin Island, while those from FAUNAL ANALYSIS 77 a LEVEL A-1 MINIMUM NUMBER OF ANIMALS 13 100- 99 - 9 0 - 8 3 - 8 0 . 7 5 - 7 0 . 6 5 " 6 0 - 5 5 - 5 0 - 4 3 - 4 0 - 3 5 - 3 0 - 25 - 20 - 15 - 1 0 - 3 - PERCENTASE RECOVERED BONES EXPECTED NUMBER HEAD M AN D IB LE A TL A S 0 0 26 23. 3 13 TRUNK A XI S 23. 3 13 CE RV IC AL 12. 8 65 TH OR AC IC LU M BA R 8. 15 195 22. 14 65 SA CR AL 3. 1 39 R IB S UPPER FRONT LEG SC AP UL A PR O XI M AL HU M ER US D IS TA L H UM ER US LOWER FRONT LEG PR O XI M AL RA OI US D IS TA L RA OI US UPPER REAR LEG P E LV IS PR O XI M AL FE M UR D IS TA L FE M UR LOWER REAR LEG < -< D IS TA L TI B IA FO O T OR FL IP PE R 8. 30 364 50. 13 26 27. 7 26 31. 8 26 58 15 26 1 3B. 10 26 15 4 26 23 6 26 19 5 26 W* 22 26 50 13 26 0 0 988 - - - - - - - - - - - - - - - - - - b LEVEL A-2 MINIMUM AN1"MBALS?FH 100- 95 - 9 0 - 65 - 8 0 - 7 3 - 7 0 - 6 5 - 6 0 - 3 5 - 3 0 - 4 5 . 4 0 . 35 - 3 0 - 25 - 20 - 10 - 5 - PERCENTAGE RECOVERED ./SONES EXPECTED NUMBER HEAD M AN D IB LE A TL A S 14. 3 22 27. 3 11 TRUNK 9. 1 11 < 9. 5 55 TH OR AC IC LU M BA R 12. 19 165 15. 8 55 SA CR AL 2 UPPER FRONT LEG SC AP UL A PR O XI M AL HU M ER US D IS TA L H UM ER US LOWER FRONT LEG S 3 j = UPPER REAR LEG P E LV IS PR O XI M AL D IS TA L FE M UR LOWER REAR LEG ?< D IS TA L TI B IA gi 1. 2 33 10. 32 308 36. 8 22 18. 4 22 36. 8 22 86. 19 22 6R 15 11 23. 5 22 45. 10 22 50. 11 22 73. 16 22 68. 15 49. 4 1 0 22 836 - - - - - c LEVEL A-3 MINIMUM NUMBER OF ANIMALS 6 100 - 9 5 - 9 0 - B S - 8 0 - 7 5 - 7 0 - 6 5 - 6 0 - 5 5 - 3 0 - 4 5 - 4 0 . 3 3 - 3 0 - 25 - 2 0 - 13 - 1 0 - 3 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAD I A TL A S 2 5 . 3 12 n n 6 TRUNK AX IS 33 ? 6 CE RV IC AL 3 9 30 TH OR AC IC LU M BA R 12. 11 LiS. 20 ft 30 SA CR AL R IB S 6. 1 M. 11, 18 168 UPPER FRONT LEG SC AP UL A PR O XI M AL HU M ER US D IS TA L HU M ER US LOWER FRONT LEG PR O XI M AL RA O IU S D IS TA L RA DI US UPPER REAR LEG PE LV IS PR O XI M AL FE M UR i\ LOWER REAR LEG PR O XI M AL TI B IA D IS TA L TI B IA FO OT OR FL IP PE R 25. 3 12 58, 7 12 67, 9 12 67, 8 12 J i 4 12 17, 2, 12 17, 2 12 0 0 12 v7, a 12 59, 7 12 28. 127 4 5 6 " - - " d LEVEL A-4 MINIMUM NUMBER OF ANIMALS 10 95 - 9 0 - 85 - 6 0 - 7 5 - 7 0 - 6 5 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 - 3 3 - 30 - 25 - 20 - 15 - 10 - ?- PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAD 2 A TL A S 0 0 20 10. 1 10 TRUNK s 1 THORA CI C LU M BA R 10. 1 10 8 4 50 11 16 150 18. 9 50 SA CR AL 1 UPPER FRONT LEG l O S DISTA L H UM ER US LOWER FRONT LEG RA OI US DI ST AL RA DI US UPPER REAR LEG UJ PR O XI M AL FE M UR j ? LOWER REAR LEG PR O XI M AL TI B IA < - i- ? FO OT OR FL IP PE R 7. 2 30 1. 3 280 25. 5 20 5 5 . 11 20 60. 12 20 70. 14 20 60. 12 20 15. 3 20 1 0 . 6 20 25. 5 20 4 5 . 9 20 40. 8 20 7. 5 4 /60 - - - - - - - - - " FIGURE 12.?The attained percentage of the expected number of bones for the minimal number of recovered seal bones from the Walakpa site: a, level A-1; b, level A-2; c, level A-3; d, level A-4. 78 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBKK 2 0 a LEVEL A-5 MINIMUM NUMBER OF ANIMALS 12 100- 9 5 - 9 0 - 8 5 - 8 0 - 7 5 - 7 0 - 6 5 - 6 0 - 55 - 5 0 - 4 5 - 4 0 - 3 5 - 3 0 - 23 - 2 0 - 15 - 10 - 5 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAO M AN OI BL E I \ \ 13. 3 24 0 0 12 TRUNK X CE RV IC AL 0 0 12 10. 6 60 TH OR AC IC LU M BA R 10. 18 180 20. 12 60 SA CR AL R IB S 0 0 36 16. 54 336 UPPER FRONT LEG < PR O XI M AL HU M ER US D IS TA L HU M ER US LOWER FRONT LEG PR O XI M AL RA OI US O IS TA L RA DI US UPPER REAR LEG UJ PR O XI M AL FE M U R DI ST AL FE M U R LOWER REAR LEG PR O XI M AL TI B IA D IS TA L TI B IA [ FO OT OR FL IP PE R 46. 11 24 83. 20 24 75. 18 24 58. 14 24 50. 12 24 38. 9 24 21. 5 24 25. 6 24 100. 24 24 88. 21 24 8. 77 912 - " - - - - - b LEVEL A-6 MINIMUM NUMBER OF ANIMALS 13 100- 95 - 9 0 - 85 - 80 - 7 5 - 7 0 - 6 5 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 . 3 5 - 3 0 - 25 - 20 - 15 - 10 - ?" PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAO M AN DI BL E A TL A S 2X. 6 26 15. 2 13 TRUNK < 0 0 13 CE RV IC AL < LU M BA R 6. 4 8. 16 65 195 8. 5 65 SA CR AL 1 0 0 39 2. 9 364 UPPER FRONT LEG SC AP UL A PR O XI M AL HU M ER US D IS TA L HU M ER US LOWER FRONT LEG PR OX IM AL RA DI US OI ST AL RA DI US UPPER REAR LEG P E LV IS PR O XI M AL FE M U R <1 LOWER REAR LEG PR O XI M AL TI B IA O IS TA L TI B IA FO O T OR FL IP PE R 35. 9 26 46. 12 26 42. 11 26 1 42. 11 26 42. |1 26 23. fi 26 19. 5 26 31 R 26 81, ?1 26 42. 12 ?11 122 26 988 - " - _ - c LEVEL A-7 MINIMUM NUMBER OF ANIMALS 7 100- 9 5 . 9 0 - 85 - 8 0 . 7 5 - 7 0 - 6 3 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 - 3 5 - 3 0 - 25 - 20 - 13 - 10 - 5 - PERCENTAGE RECOVERED 80NES EXPECTEO NUMBER HEAD M AN DI BL E A TL A S A / 0 0 14 ' \ 14. 1 / TRUNK *< tr x 5 i LOWER REAR LEG < D IS TA L TI B IA FO OT OR FL IP PE R 0 .2 1 504 36. 13 36 22. 8 36 0 0 36 47. 17 36 58. 21 36 25. 9 36 5b. 20 36 6 1 . 22 36 69 . 25 36 67 . 24 36 3. 42 1368 - - - - - - b LfVEL B - 7 MINIMUM NUMBER OF ANIMALS 13 100 ? 95 - 9 0 - 85 - 8 0 - 75 - 7 0 - 6 5 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 - 3 5 - 30 - 25 - 20 - 15 - 10 - 5 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAO < 2 A TL A S \ \ 1 9 . 5 2 6 8 . 1 13 TRUNK A XI S 8 . 1 13 < TH OR AS IC LU M BA R 8 . 5 65 1 6 . 3 1 195 4 9 . 32 65 SA CR AL RI BS UPPER FRONT LEG SC AP UL A PR O XI M AL HU M ER US DI ST AL HU M ER US LOWER FRONT LEG PR OX IM AL RA DI US OI ST AL RA OI US UPPER REAR LEG PE LV IS PR O XI M AL FE M UR D IS TA L FE M U R LOWER REAR LEG PR OX IM AL TI B IA O IS TA L TI B IA FO OT OR FL IP PE R 4 . 1 26 1 7 . 6 1 3 6 4 2 7 . 7 26 6 2 . 1 6 2 6 2 3 . 6 26 5 4 . 1 4 26 4 6 . 12 26 2 3 . 6 26 7 7 . 2 0 2 6 6 2 . 16 2 6 1 5 . 4 26 1 9 . 5 26 4 . 39 9 8 8 - - d LEVEL B - 9 MINIMUM NUMBER OF ANIMALS 6 8 100- 95 - 9 0 - 85 - 8 0 - 75 - 7 0 - 6 5 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 - 35 - 3 0 - 25 - 20 - 15 - 10 - '- 0 - PERCENTAGE RECOVEREO SONES EXPECTED NUMBER HEAO Z < 2 3 . 32 L36 2 5 . 17 6 8 TRUNK A XI S CE RV IC AL TH OR AS IC LU M BA R SA CR AL 1 UPPER FRONT LEG SC AP UL A PR O XI M AL HU M ER US DI ST AL HU M ER US LOWER FRONT LEG PR OX IM AL RA OI US DI ST AL RA OI US UPPER REAR LEG PE LV IS PR O XI M AL FE M U R DI ST AL FE M UR LOWER REAR LEG PR O XI M AL TI B IA O IS TA L TI B IA OR FL IP PE R 3 2 . 22 6 8 1 6 . 53 3 4 0 1 5 . 1 5 4 1020 6 2 . 2 1 1 3 4 0 8 . 1 0 2 0 4 1 6 . 3 1 3 1904 3 4 . 4 6 L36 9 0 . L22 1 3 6 7 9 . 1 0 8 1 3 6 3 7 . 5 0 1 3 6 3 7 . 5 0 1 3 6 2 9 . 4 0 1 3 6 8 5 . 1 1 6 1 3 6 7 2 . 9 8 1 3 6 7 2 . 9 9 1 3 6 6 5 . 8B 1 3 6 4 . 0 2 1 4 516( - . - - - - - - - - - - FIGURE 15.?The attained percentage of the expected number of bones for the minimal number of recovered seal bones from the Walakpa site: a, level B-6; b, level B-7; c, level B-8; d, level B-9. FAUNAL ANALYSIS 81 LEVEL B-10 MINIMUM NUMBER OF ANIMALS 162 9 5 - 9 0 - 8 5 - 8 0 - 7 5 - 7 0 . 6 3 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 - 3 5 - 3 0 - 25 - 2 0 - 15 - 1 0 - 3 . PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAD M AN D IB LE 59. 191 324 A TL A S 24. 39 162 TRUNK A XI S 15. 24 162 CE RV IC AL 12. 94 BIO TH OR AS IC 10. 241 >430 LU M BA R 35. 280 810 SA CR AL 8. 38 186 R IB S 18. 7 94 1536 UPPER FRONT LEG SC AP UL A 57. 185 324 PR OX IM AL HU M ER US 73. >36 324 D IS TA L HU M ER US 67. 216 324 LOWER FRONT LEG PR OX IM AL RA DI US 62. 201 324 DI ST AL RA DI US 61. 197 324 UPPER REAR LEG % 26. 84 324 PR O XI M AL FE M U R 93. 300 324 OI ST AL FE M UR 85. 75 324 LOWER REAR LEG |< 41. 133 324 O IS T A L TI B IA 44. 142 324 FO OT OR FL IP PE R 3. 384 1 2 . m - FIGURE 16.?The attained percentage of the expected number of bones for the minimal number of recovered seal bones from the Walakpa site, level B-10. southern Hudson Bay may be 8 to 10 percent shorter (McLaren, 1962:171). The reason for this size dif? ference is that a longer suckling period is permitted in the more northerly localities, which results in larger, more vigorous pups. If the climatic condi? tions during the Early Thule occupations were sig? nificantly warmer with longer summer weather con? ditions, the ice would tend to break up sooner in the spring and seriously shorten the suckling period of the ringed-seal pups. This would in turn cause the seals to be smaller and would account for the size distribution noted above (p. 74). The maintenance of a balanced population size of the ringed seals is of great importance for the suc? cessful exploitation of these animals. Land-fast ice is the most important variable limiting the number of reproducing females. Within the area of land-fast ice, there are limited numbers of sites which are suitable for birth lairs and reproduction is unsuccessful else? where (McLaren, 1962:172). Each seal requires a territory of fixed size so that there is an upper limit to the number of animals inhabiting a given area of ice. McLaren has calculated population densities of seals based on ice types: (1) Within one mile of shore and surrounded by further ice, there will be approxi? mately 35 seals per square mile of ice. (2) Within one mile of land but exposed to open water on its seaward edge, there will be about 10 seals per square mile. (3) More than one mile from land, there will be about 5 seals per square mile of ice. If the Eskimos of a region are successfully to ex? ploit an area for a long period of time, then the mortality rate imposed by hunting and natural causes must not exceed the recruitment into the seal popula? tion. In order to achieve this balance of population, the annual kill should not exceed 10 percent of the total population at the beginning of the seal year (McLaren, 1962:175). When calculating the minimum number of seals represented by Birnirk level B-10, I find that at least 162 seals were killed. By using optimum modern hunting conditions, with a seal density of 35 indi? viduals per square mile, it is found that 162 seals totally depletes four square miles of ice area. In order to maintain a balance in the seal population, the Eskimos would have had to exploit a minimum of 3.5 seals per mile. If the 162 seals from level B-10 represent this type of minimal exploitation, the Es? kimos would have had to hunt 46 miles of coastline. Based on element distribution, if our remarks about proximity of kills to the camp site are valid, then a great percentage of the seals were killed locally. We can then assume that the local seal population would have been seriously depleted for several generations and would not support a continued Eskimo habi? tation. The above remarks are based on modern climatic conditions and with the proposed warming trend in the climatic situation correlated with a decrease of suitable pupping areas, the ringed-seal population may actually have been significantly smaller. Given this situation and the fact that I did not calculate any loss of seals due to ice conditions, I feel that a mini? mum exploitative area of 46 coastal miles is not out of line and should possibly be increased several fold. This is possibly the reason that Birnirk camp sites were of brief duration and moved often. It is also because of this type of maximizing exploitative strategy that we see a rapid spread of Birnirk camp sites. CARIBOU The same pattern was found in the distribution of caribou elements (Figures 17-21) as occurred with the seals. The elements again point to the same type of campsites as indicated by the seal bones. In levels A-1 through B-8 there is a higher incidence of skulls. 82 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 a LEVEL A-1 NUMBER OF ANIMALS 4 100- 95 - 9 0 - 6 5 - 8 0 - 7 3 - 7 0 - 6 5 - 6 0 - 5 5 - 5 0 - 4 5 - 4 0 - 35 - 3 0 - 2 5 - 2 0 - .!- , 0 - s - PERCENTAGE RECOVERED BONES EXPECTEO NUMBER HEAD s 13. 1 e i 3 100 8 8 0 0 4 TRUNK s 0 0 4 1 0 0 16 5 0 0 56 2 0 0 24 1 ; / / / / / 0 0 4 IB. 17 56 UPPER Is o 2 w 2 LOWER S 3 i! ii i\ UPPER I 2 ? / ? 1 A r\ A / A / / ' / / / / / \ / 1 / 1 38. 3 8 0 0 8 13. 1 8 5 0 . 4 8 50 4 8 0 0 8 2 5 . 2 8 50. 4 8 0 0 8 0 0 8 LOWER REAR LEG l o i - E? 2 tt D IS T A L M E rA TA R SA L FOOT I 0 0 8 38. 3 8 50. 4 8 50. 4 8 ?5. 4 16 2 3 i 25. 4 16 16, 15 96 - - - - - - - - - - - - - - - - - - " C LEVEL A-3 MINIMUM NUMBER OF ANIMALS 4 100- 95 - 9 0 - 85 - eo_ 75 - r o ? e s - 6 0 - 5 3 - 3 0 - 4 3 - 4 0 - 35 - 3 0 - 2 5 - 2 0 - , 5 - 1 0 - 5 - PERCENTAGE RECOVERED BONES EXPECTEO NUMBER HEAD 2 36. 3 8 3 3 13. 1 8 0 0 4 TRUNK 0 0 4 13. 2 16 D 18. 10 56 2 17. 4 24 i 1 25. 1 4 7. 7 96 UPPER FRONT EG | II o 2 ? 2 LOWER FRONT LEG || ii ii 2 g ? 2 UPPER REAR LEG J 2 ^ x 2 ^ 2 25. 2 8 38. 3 8 0 0 8 38. 3 8 88. 7 8 13. 1 8 0 0 8 0 0 8 0 0 8 13. 1 8 LOWER REAR LEG 2 ? h aj - s L 2 O IS T A L M E TA TA R SA L FOOT 13. 1 8 50. 4 8 25. 2 8 38. 3 8 13. 2 16 2 3 1 6. 1 16 1 . 1 96 - " b LEVEL A-2 MINIMUM ANIMALS 6 100- 9 5 - 9 0 - 8 5 - 8 0 - 7 5 - 7 0 - 6 5 - 6 0 - 55 - 5 0 - 4 5 - 4 0 - 35 - 3 0 - 2 5 - 2 0 - . 5 - . 0 - 5 " PERCENTAGE RECOVERED BONES EXPECTEO NUMBER HEAO 5 75. 9 12 d LEVEL A _ 4 NUMBER OF 9 5 - 9 0 - 6 5 - 0 0 - 75 - 70 - 65 - 6 0 - 55 - 5 0 - 4 5 - 4 0 - 35 - 3 0 - 2 5 - 2 0 - 15 - 1 0 - 5 - PERCENTAGE RECOVERED BONES EXPECTEO NUMBER s 3 42. 5 12 0 0 6 HEAD 5 0 0 2 i 3 | 0 0 2 mo I i TRUNK a E J 2 50. 3 b 25. 6 24 18. 15 84 22. 8 36 | 1 UPPER FRONT LEG 1 100. 6 6 60. 86 144 58. 7 12 TRUNK = 1 I 2 A A 0 0 1 0 0 4 21 3 14 n n 6 1 ? O IS TA L H U M E R U S LOWER FRONT LEG 1 = - I D H 2 a ^ 2 s. ? 2 REAR LEG ; 2 ? x 2 H 1 0 0 12 12, 5 12 inn, 12 12 100 12 12 4? 5 12 42 5 12 16: 2 12 tfl 6 12 0 0 12 LOWER REAR LEG ? ? O IS T A L T IB IA P R O X IM A L M ET A TA R SA L D IS T A L FOOT I 3 2 3 P H A LA N G E S 6 0 . 6 12 V> 11 1? 7 12 6 4 17. n 4 31 12 24 24 144 - - - - - - UPPER FRONT LEG i 0 0 1 n n 24 n 2 l i O 2 in 3 LOWER FRONT LEG 2=> 11 i i 2 g ? s ? 2 UPPER REAR LEG UJ x 2 K 3 1 1 1 I n n i J l u l l 0 2 2 2 2 2 2 2 2 2 LOWER REAR LEG P R O X IM A L T IB IA S- - i ?- 2 ? 2 FOOT 1 3 ! 2 2 2 2 4 4 24 FIGURE 17.?The attained percentage of the expected number of bones for the minimal number of recovered caribou bones from the Walakpa site: a, level A-1; b, level A-2; c, level A-3; d, level A^t. FAUNAL ANALYSIS 83 a LEVEL A - S NUMBER OF ANIMALS 6 100- 9 5 - 9 0 - 8 5 - 6 0 - 7 3 - 7 0 - 6 5 - 6 0 - 5 5 - so- 4 3 - 4 0 - 3 5 - 3 0 - 2 3 - Z 0 - 15 - 1 0 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAD I 32 . 11 12 C LEVEL A-7 NUMBER OF ANIMALS 3 LOO. 9 5 - 9 0 - 85 - 8 0 . 7 5 - 7 0 ^ 6 3 - 6 0 - 3 5 - 50/- 4 3 - 4 0 - 5 3 - 3 0 - 2 5 - 2 0 - . 3 - 1 0 - 5 - 0 - PERCENTAOE RECOVERED BONES EXPECTEO NUMBER 2 5 ioQ. 12 1 2 3 6 HEAD ? 1 / 0 6 J a A ' \ 6 n 3 TRUNK 5 L7 . 1 6 C E R V IC A L 3 24 < 2 25 84 16 36 S A C R A L 1 6 6 R IB S UPPER FRONT LEG S C A P U LA 2 ^ 0 2 D IS T A L H U M E R U S LOWER FRONT LE6 P R O X IM A L R A O IU S J D M ET A CA RP AL D IS TA L M ET AC AR PA L UPPER REAR LEG P E L V IS P R O X IM A L F E M U R 2s LOWER REAR LEG It i. m 2 a ?- 2 Q 2 FOOT I 63 9 1 144 Rl 10 1 2 R 1 12 17 2 12 17 2 12 75. 3 12 fl 1 12 ?5 3 12 4? 5 12 8 . 1 12 n. I 12 R 1 12 5 0 . 6 12 25 3 12 25 3 12 1 3 . 3 24 2 3 P H A LA N G E S 1 7 . 4 2 4 7 . 10 149 - ? - - - - - " TRUNK J_ 0 3 C E R V IC A L / S 12 1 2 V^ 1 42 1 18 s N/ 0 3 5 UPPER 0 2 ? 2 LOWER FRONT LEG 2=) |1 Si is 2 5 S 2 ? 2 UPPER REAR LEG 2 ? x 2 2 j LOWER REAR LEG 2 ., |i is 2 3j Q 2 FOOT | k / l\ / l l / 1 ll / l A / 1 11 A / 11 / \ / 11 / V 1 I I / 11 / 1 / 1 A A / \ / \ , \ / 14 72 1 6 n 0 6 0 0 6 ? 6 i 6 ? 6 50 1 6 81 5 6 0 0 6 0 0 6 0 0 f ?7 4 6 1 7 . ] 6 17 1 6 16 2 12 2 3 1 0 0 u 3, 2 72 - - - - - - - - - " b LEVEL A-6 100- 95 - 9 0 - 65 - 8 0 - 7 3 - 7 0 - 6 5 - 6 0 - 3 3 - 5 0 - 4 5 - 40 J 35 J 3 0 - 25 H 2 0 . , 5 - 5 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAD ? 17. 1 6 d LEVEL B-1 ANIMALS 3 95 - 9 0 - 8 5 - 8 0 - 7 5 - 7 0 - 6 3 - 6 0 - 5 5 - 3 0 - 4 5 - 4 0 - 3 5 - 3 0 - 2 3 - 2 0 - . 3 - 1 0 - " PERCENTAGE RECOVEREO BONES NUMBER 5 1 1 17. 1 6 1 3 67. 4 6 0 0 3 HEAD \ s 1 1 M 1 \ 33. 2 6 33. 1 .1 TRUNK S J 0 0 3 9 \ \ V 0 0 3 1 A 8. 1 12 l rN / 8. 1 12 | \ 2. 1 42 | J SJ 5. 2 42 2 ^ 0 0 18 ? 2 A A M 33. 6 18 ( 0 0 3 j I 0 0 3 1 35. 25 72 1 j 4. 3 72 UPPER i li o2 ? * LOWER FRONT LEG f| ^2 ii 2 * ? 2 UPPER REAR LEG u 2 % K s LOWER REAR LEG 2 < a A 5 O IS T A L M ET A TA R SA L J 0 0 6 3 3 . 2 6 50. 3 6 33. 2 6 50. 3 6 50. 3 6 83. 5 6 1 0 0 6 6 17. 1 6 0 0 6 0 0 6 100. 6 6 17. 1 6 50. 3 6 1 7 . 2 12 FOOT 2 0 0 12 I 1 . 1 72 " - - " - - - " " - UPPER ? .- FRONT L^ G ? H U M E R U S - if 1 LOWER " FRONT LEG 25 ?? ii 2 a ?i UPPER REAR LEG J x 2 ^ 2 LOWER REAR LEG 3< ii 2 ? 11 2 ? 2 1 A I \ / \ \ h 1 '\ \ A M \ A \ 1 \l l / \ / I 33. 2 6 0 0 6 0 0 6 50. 3 6 33. 2 6 0 0 6 67. 4 6 0 0 6 17, 1 6 0 0 6 33. 2 6 inn 6 6 17 1 (j sn 3 6 a l i : DOT 2 0 0 1. ' S l 1 72 " - " " FIGURE 18.?The attained percentage of the expected number of bones for the minimal number of recovered caribou bones from the Walakpa site: a, level A-5; b, level A-6; c, level A-7; d, level B-1. 84 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a L E V E L B - 2 M I N I M U M N U M B E R OF - A N I M A L S 2 1 0 0 - 9 5 - 9 0 - as - 8 0 - 75 - 7 0 - 6 5 - 6 0 - 5 3 - 3 0 - 4 5 - 4 0 - 35 - 3 0 - 2 5 - 15 - 1 0 - 5 - P E H C E N T A G E R E C O V E R E O BONES E X P E C T E O N U M B E R H E A D 5 2 I T R U N K 5 I | 1 | ? U P P E R 5 a o 2 0 I D - J - is 2 3 LOWER F R O N T 1 EG 2 5 || 2 ^ 0 2 U P P E R REAR LEG tlj 2? x 2 if 1?1 1?1 n n j . 2 50 1 2 sn 1 2 2 2 50 2 2 2 2 LOWER REAR LEG 2 fl i - ? j | 0 2 FOOT | 3 ! . - - - - - - - - - I 1 n - ? - ? ? ? - ?1 ' * 2 2 2 2 5(1 4 4 0 24 FIGURE 19.?The attained percentage of the expected number of bones for the minimal number of recovered caribou bones from the Walakpa site: a, level B-2; b, level B-3; c, level B-4; d, level B-5. FAUNAL ANALYSIS 85 a LEVEL 8-6 MINIMUM NUMBER OF ANIMALS 2 9 5 - 9 0 - 6 5 - 8 0 - 7 5 - 7 0 - 6 5 - 6 0 - 3 3 - 3 0 - 4 3 - 4 0 - 3 3 - 3 0 - 2 5 - 20- 15- 1 0 - 3 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER 2 0 0 4 C LEVEL B-8 NUMBER OF ANIMALS 4 100- 9 5 - 9 0 - 8 5 - 8 0 - 7 5 - 7 0 - 6 5 - 6 0 - 3 5 - 5 0 - 4 3 - 4 0 - 3 5 - 3 0 - 2 3 - 2 0 - 15 - 5 - PEftCENTAGE RECOVEREO BONES EXPECTED NUMBER HEAD 2 3 0 0 4 0 0 2 HEAD 2 \ \ \ 25. 2 8 2 3 A A \r V 0 0 8 25. 1 4 9 0 0 2 C E R V IC A L / 0 0 8 TRUNK 1 2 A 7. 2 28 \ 0 0 12 f / 0 0 2 1 A 8. 4 48 TRUNK 5 \ M 0 0 4 C E R V IC A L / \ 6. 1 16 j 2 / s / 0 0 56 8. 2 24 | A A 25. 1 4 5 \ 13. 15 96 UPPER FRONT LEG I \ 0 0 4 FR l 13. 1 8 ? 3 LOWER FRONT LEO si - a ii 2 5 2 IS ? 2 UPPER REAR LEG 5J x 2 . i i K 1 LOWER REAR LEG 2 4 i- ? it S -< -'5 FOOT A ST R A G A LU S t A h ft \ \ \ I \ r\ \A \ \ i 0 0 4 50 2 4 0 0 4 0 0 4 0 0 4 50. 2 4 2 5 . 1 4 0 0 4 2 5 . 1 4 2 5 . 1 4 0 0 4 0 0 4 / 0 0 4 A 13. 1 8 PPER NT LEG o 2 V, 2 LOWER FRONT LEG 2 = ^ t t ii - % ? 2 UPPER REAR LEG j j U x 2 H S LOWER REAR LEG 2< ii 2 * ?- 2 Q 2 2 3 P H A LA N G E S \ 0 0 8 0 0 4 8 - - - - - - - - - - - - - - - " - - I 13. 1 8 50. 4 8 25. 2 8 0 0 8 25. 2 8 25. 2 8 13. 1 8 0 0 8 0 0 8 0 0 8 63. 5 8 38. 3 8 0 0 8 13. 2 16 3 1 0 0 16 4. 4 96 - - 1- - - - - - b LEVEL B-7 NUMBER OF ANIMALS 4 100- 9 5 - 9 0 - 8 5 - 8 0 - 75 - 7 0 - G 5 - 8 0 - 35 - 5 0 - 4 5 - 4 0 - 35 - 3 0 - 2 5 - 2 0 - , 0 - 5 - PERCENTAGE RECOVERED BONES EXPECTED NUMBER HEAO j 50 4 8 d LEVEL B-9 MINIMUM NUMBER OF j ANIMALS 11 9 5 - 9 0 - 85 - 8 0 - 7 3 - 7 0 - 6 5 - 6 0 - 5 5 - , 0 - 4 3 - 4 0 - 3 3 - 3 0 - 2 5 - 2 0 - 1 0 - 5 - PERCENTAGE RECOVERED 80NES EXPECTEO NUMBER 2 3 TRUNK a I ? 2 13. 1 8 25. 1 4 1 0 0 4 4 63. 10 16 20. 11 56 4. 1 24 I 5 UPPER FRONT EG 3 H o 2 S ' LOWER FRONT LEG 2 D |1 ii 2 g O IS TA L M ET A CA RP AL UPPER REAR LEG = 2? x 2 2 j 1 LOWER REAR LEG 2 C i\ If 1 FOOT i 2 3 P H A LA N G E S t A 1 I 1 . . j \ l\ h l \ l\ / 1 / \ I \ 1 \ \ I \ I \ fs. / \ / \ I I 1?\ l\/ \ \ \ \ \ 1 \ 1 \ 1 \ 25. , 4 22. 21 96 13. 1 8 25. 2 8 50. 4 8 38. 3 8 13. 1 8 0 0 8 50. 4 8 25. 2 8 0 0 B 0 0 8 13. 1 8 75. 6 8 0 0 8 25. 2 8 25. 4 16 0 0 16 1 . 1 96 - - - - - - - - - - " HEAO ^ 14. 3 22 1 3 TRUNK 2 C E R V IC A L | 2 . 5. 1 22 45. 5 11 36. 4 11 32. 14 44 31 . 46 154 68. 45 66 I 5 UPPER 1 3 t t o 2 !jj 3 LOWER FRONT LEG 2 = ->2 is 2 ^ ^ 2 i ? 2 UPPER REAR LEG UJ ?5 x 2 LOWER REAR LEG 3 ? s i\ I I ? 2 FOOT l 3 I 18. 2 11 4 1 . 108 264 50. 11 22 50. 11 22 86. 19 22 9. 2 22 59. 13 22 50 11 22 14. 3 22 50. 11 22 100 22 22 50. 11 22 95. 21 22 95. 21 22 27. 6 22 73. 16 22 8. 3 44 32. 14 44 9. 24 264 - - " FIGURE 20.?The attained percentage of the expected number of bones for the minimal number of recovered caribou bones from the Walakpa site: a, level B-6; b, level B-7; c, level B-8; d, level B-9. 86 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 LEVEL B-10 NUMBER OF ANIMALS 12 6 5 - 2 5 - PERCENTAGE RECOVERED 80NES NUMBER HEAD 3 \ 17. 4 24 2 4. 1 24 3 100 12 12 TRUNK 5 67. 8 12 1 23. 11 48 I 36. 61 168 2 58. 42 72 | 3 3 . 4 12 g 53. 153 288 FRONT LEG s 46. 11 24 0 2 29. 7 24 S! 50. 12 24 LOWER ii 7 1 . 17 24 i\ 42. 10 24 it 58. 14 24 ?3 D 2 58. 14 24 REAR LEG J 25. 6 24 2 = x 2 29. 7 24 ? 2 58. 14 24 LOWER REAR LEG 2 < 71. 17 24 il 71. 17 24 iS, ^ 2 25. 6 24 i \ 25. 6 24 FOOT l 6. 3 48 3 25. 12 48 i 3. 9 288 - FIGURE 21.?The attained percentage of the expected number of bones for the minimal number of recovered caribou bones from the Walakpa site in level B-10. ribs, scapulae, pelves, and lower leg bones. Foot bones remain proportionately high or low to head bones which, as mentioned above, may indicate the distance the animal was killed from camp. However these, as well as missing lower leg bones, may indicate dog feeding. In the levels B-9 and B-10 these bones are relatively scarce, indicafing^ that most of the caribou meat utilized in these two levels came from places other than Walakpa. The occurrence of a few of these bones may indicate the occasional local kills. The upper limb bones and trunk bones are relatively low to completely absent in the hunting camp levels, while in the base camp levels they are represented by 50 to 100 percent of the expected bones. It can be ob? served that the occurrence of the scapulae and pelves is quite erratic. It is felt that their presence or absence has to do with whether or not the meat was frozen when butchered. WHALE The rare occurrence of whale bones in the Birnirk levels at Walakpa is significant. Although the sample is extremely small, I believe that it represents a true picture of the whales hunted by the Birnirk in? habitants. Considering the combined length of oc? cupation of both the B-10 and B-9 levels (page 92 for details) and the probability that the time spans cover two fall seasons, one or two spring seasons, and one summer season, it would seem reasonable to as? sume that during this time the larger baleen whales would have migrated past the site at least three or four times, unless, of course, the migratory habits of these mammals were different than those we observe today. If these whales were present in these waters during the Birnirk occupation, their presence does not appear in the Walakpa faunal sample. This may indicate that the Birnirk Eskimos were not getting these animals regularly, that the bones were left at the kill sites, or possibly they were not even hunting the large whales. However, by late Birnirk times, which does not occur at Walakpa, whale-hunting equipment is found and is probably imported from western Alaska (Ford, 1959:103). The evidence seems to indicate that dur? ing the early Birnirk phase, large whales were not hunted regularly, but their importance increased dur? ing middle Birnirk and may have been an important element in the late Birnirk economy. POLAR BEARS The distribution of polar bear bones is unique (Table 8). Bear meat was considered an excellent food by the historic Eskimos (Spencer, 1959:374). Most of the bear meat, other than the liver, which is extremely toxic, is eaten. In the hunting camp levels where bear elements did occur, they were exclusively head or feet elements. It is presumed that the choice cuts of meat were taken back to the base camp with the claws and teeth being extracted at the hunting camp. However, in the base camp levels there were again fewer choice meat elements and more head and feet bones. It is possible also that the claws and skulls were brought back still attached to the hides, result? ing in this bone distribution. DOG, Fox, AND W O L F By the breaks in the dog skulls, it appears that a small variety of dog was eaten. Coastal Eskimos are not now inclined to eat dog, but have no feelings against it (Spencer, 1959:374). T h e inland Eskimos raise dogs for food, and wolves and foxes are also boiled and eaten. Dog and fox may have made up a small portion of the diet of the people living on levels B-9 and B-10, but are infrequently found in the later levels. FAUNAL A N A L Y S I S 87 TABLE 8.?Distribution of selected bone elements for polar bear, bearded seal, and walrus from the various occupation levels of each culture phase, Walakpa site CD CO rd ? cu i-i 3 4-> r H 3 U CD rH 3 $ CD 4-> r H Th u >! rH H ca H irk c -rH m c o ?H 4-> rd ?0 ftrH C 3 CD 3 O > o O (D E O J A - 1 A - 2 A - 3 A - 4 A - 5 A - 6 A - 7 B - 1 B - 2 B - 3 B - 4 B - 5 B - 6 B - 7 B - 8 B - 9 B-10 P o l a r TJ id & X 2 - _ - 2 - 1 - - 2 - - - - _ 6 7 | 3 SH E-i 1 8 Bear tj> CD tTi r-l CD ?H O 0 u id fc, CQ - - _ - - - - - _ - - - - - 1 4 1 2 2 -P O o fc, 1 1 _ - 3 3 1 1 2 3 1 8 3 0 Bearded -a CO CD a 1 - 1 - l - 1 1 _ 3 5 ti 3 rH EH 3 1 1 2 1 2 1 1 1 1 4 5 2 4 2 9 CTl CD r H 4-1 c 0 rH fc. 2 1 2 - - - 1 1 3 1 6 5 Seal CTl CD r H X u id HI 2 - 1 1 1 _ - - 1 5 2 +i 0 0 tu 8 5 2 4 2 1 1 1 5 1 - 1 6 2 0 ?0 rd CD a l - - - _ 2 3 3 Walrus 3 3 h EH 1 1 2 3 2 1 2 4 1 2 Cn V f-H 4-1 c o l-l CK 1 1 " 1 1 1 1 2 1 Cn CI x o id m l _ - - 1 - 1 2 2 2 2 0 4-1 o o Cu - 2 _ 2 2 3 1 0 6 OTHER MAMMALS AND BIRDS Except in levels B-9 and B-10, the bones of ani? mals other than seal and caribou make up an ex? tremely insignificant number. Those bones, as men? tioned above, are thought to have been introduced from outside the site in the form of supplemental foods. This is particularly apparent in the case of birds where the majority of bones found were those on which larger portions of meat occurred. There were few heads and no feet, with the humeri, sterna and furcula making up over 90 percent of the bone count. This is also apparent with bearded seal and walrus bones. T h e bones found in the hunting camps repre? sent, for the most part, leg bones, foot bones, and occasionally trunk bones (Table 8). From these ani? mals the choice portions of meat are the heart, liver, and ribs (Spencer, 1959:374). The rest of the meat is less highly regarded for human consumption and is fed to the dogs. The flesh of the flippers and legs are sewn into skins and allowed to ferment. When the flipper is boiled it is regarded as an excellent food (Spencer, 1959:374). Skull bones of these animals are very dense and may have also been carried to hunting sites or collected on the beach for materials to manufacture artifacts. In summary, the faunal analysis has pointed out several important aspects from which to interpret archeological evidence of the Walakpa site. The first and most important inference that can be made is that there has been a change in the seal populations between the Birnirk and Early Thule periods and the possibility of larger sizes of ring seals in the Birnirk and Late Thule periods as compared to variations in size of the seals from the Early Thule period. This is thought to have occurred because of the nutritional defects of a shorter pupping season, caused by the earlier ice break-up in the spring during the possibly warmer, Early Thule cultural phase. The second important inference which can be made from the faunal analysis is that there are two basic patterns of site utilization. The first is that of brief hunting camps represented by small numbers of seal and caribou kills. The elements that were left be? hind in these camps are butchered and discarded bones and a relative absence of choice meat-packet bones. It is thought that the better cuts or portions of meat were taken back to a base camp while the rest was left at the hunting site. The second pattern of site utilization is that of a base camp site. Here we found that the reverse distribution of bones occurred. The choice or large meat packet bones made up the majority of the bone count, while the less desirable cuts made up sig? nificantly fewer bones. This would indicate meat in? put from the hunting camps. Another important consideration is that during Birnirk times few whale bones were being introduced into the faunal inventory. The possibility that the Birnirk Eskimos were not systematically hunting large whales is quite significant for the origin of Thule culture and is discussed in greater detail on pages 97 and 113. Archeological Analysis of the Walakpa Cultural Units In order to clarify the cultural processes repre? sented by the archeological assemblages unearthed at Walakpa Bay, it may be fruitful to examine the seasonal round of the historic Point Barrow Eskimos (Murdoch, 1892; Spencer, 1959), which can be viewed as a complex scheduling of various seasonal economic activities, which correlate with the maximum poten? tials of the various ecosystems. The seasonal round is manifested in a series of economic options so that when one ecosystem is at a low point, another eco? system may be exploited. The seasonal options are ranked in order of their economic potentials, so that the primary option has the greatest energy yield, while the secondary and tertiary options are re? spectively less. If the primary option activity is suc? cessful, the secondary and tertiary options are less important and are used only to provide a variation in diet or activity. However, if the primary option should fail, the secondary and tertiary options be? come extremely important for replacing the major food resource. The scheduling of options will depend seasonally on the success of the primary option. These options are summarized in the discussion of the sea? sonal round and in Figure 1. Historic Point Barrow Eskimo The seasonal round of the historic Point Barrow Eskimo begins in the spring (Figure 22). During mid- April most people are gathered in the main village (Utkiavik). If the weather is favorable the whaling season will soon commence. The major activities revolve around preparations for whaling (spring, option 1). New gear is made and old gear is repaired and cleaned. A small amount of duck and caribou hunting (spring, option 2) can be done at this time. The whaling season continues through May and tapers off during the first week of June. The whaling camps are situated on the ocean ice near ice leads. Camps are located near each other, and there is frequent intercrew cooperation. The remains of these camps will not, in most instances, FIGURE 22.?Seasonal options and their general geographic location. survive the summer, and broken and discarded tools usually end up in the ocean leaving no evidence for the archeologist. During June the offshore ice usually breaks up and drifts out to sea. Cooperative hunting may continue, concentrating on the bearded seal. After the whaling season, several subsistence op? tions exist. A hunter may remain in the village and participate in cooperative hunting (summer, option 1), or if he desires he may take his family to individual hunting camps (summer, option 2). These camps are 88 ARCHEOLOGICAL ANALYSIS OF CULTURAL UNITS 89 oriented toward hunting of seals, caribou, water birds, and fishing. In July, the main caribou herds arrive, and they are hunted by cooperative hunting parties composed of the same people who make up the whaling crew. Some men will return to the village from the small summer camps to participate in the hunt, others prefer to stay in the summer camps and hunt caribou on their own. If a man returns, his family generally stays at a summer camp to concentrate on the fish resources (option 3). Although the caribou remain on the coast for a short time, they can be obtained in large numbers when hunted on a cooperative basis. By mid-August, the walrus herds are moving along the coast with the drifting pack ice. Walrus hunting is another cooperative activity. The walrus is es? sential for the manufacture of many items needed to sustain life. The walrus herds have mostly left the area by the end of August at which time more effort is expended on fishing and duck hunting. The sum? mer camps then become the primary center of activity. In the fall a secondary whaling season might occur in the Barrow area (fall, option 1). At this time families start returning to the main village; but many individuals elect to continue hunting caribou and seal (fall, option 2). Most hunters are generally back in the main village by November (winter, option 1). Sometimes during the winter, a few families will leave the community to ice fish on the inland lakes and hunt caribou (winter, option 2) or venture down the coast to hunt seals and caribou (winter, option 3). However, the attractions of the village are great dur? ing the winter months of December through March, when one can visit with friends and participate in group activities, such as dances, games, and religious activities. This is also a period for repairing and manufacturing tools. This brief sketch of the seasonal round of the historic Point Barrow Eskimo indicates that there are certain limitations placed on the Eskimo by the cyclical nature of the major food sources. In order to insure enough food for the long winter, it is important that a hunter participate, in cooperative economic associations. It is also clear that several options are available for a variation in the pattern of existence if a primary energy source should fail. Archeological Considerations In order for the archeologist to reconstruct the economic round of the prehistoric Eskimo culture, the investigator must define basic sets of interrelated variables associated with specific seasonal economic activities. The variables that should define economic pursuits are (1) settlement types and locations, (2) tool kit types, and (3) faunal analysis. SETTLEMENT TYPES AND LOCATIONS.?The settle? ment types to be used in this treatment are the type i, II, and in settlements as defined by Campbell (1968) for the Tuluaqmiut Eskimo. Type i: The central base camp. This is the largest local group forming a political unit and a village. At Point Barrow, it is the village of Utkiavik. Type ii: Smaller, less permanent camps occupied for several months during the year. These camps are not necessarily occupied every year. The summer hunting or fishing camps belong here. Many of these camps exist along the coast, inland lakes, and rivers in the Point Barrow area, but none have been ex? cavated. Type HI: Camps set up for short durations for various hunting and fishing activities. These are lo? cated in the geographical areas of exploitation. T O O L KIT TYPES.?Tool kit types have been de? fined on the basis of implied activities all of which are confirmed by research on the historical Eskimo. The tool kits utilized by an Eskimo group would include those for hunting whales, seals, caribou, birds, and fish; and also those used for processing, mainte? nance, and manufacturing; transportation; culinary; environmental protection; and community activities. Of this inventory of tool kits, it is presumed that only hunting tool kits, environmental protection tool kits (dwelling type and associated artifacts) and com? munity activity tool kits (drums, games, etc., used during the winter), and possibly transportation tool kits (boats, sleds, and related gear) would be good indicators of seasonality. The other tool kits are less specific in their functions and can be expected to occur on almost every type of camp site at all times of the year. FAUNAL ANALYSIS.?Faunal and floral analysis are essential in understanding the seasonal round. In many cases, this type of data is not available to the archeologist. In the Arctic, animal and plant remains are usually well preserved; however, because of the remoteness of most sites and the tremendous cost of transportation, few investigators bring this material back to the laboratory and do not capitalize on this type of analysis. This is unfortunate because the age of animals at the time of death can be determined from tooth eruption and wear, growth ring patterns, and examination of the stages of epiphyseal closure. It also appears that butchering and meat-handling processes among the Eskimo may fluctuate during the year, depending on the type of camp site, the weather, 90 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 and if the meat is intended for immediate consump? tion or storage. This type of data is an extremely important tool for determining the seasonal aspect of an archeological site. The Chronology of the W a l a k p a Site The chronological sequence of the occupation levels at Walakpa represent one of the most complete series showing cultural development in northern Alaska. The superimposed nature of the occupation levels have yielded new data that will help clarify these cultural developments. The chronology presented here is based primarily on the occurrence in the stratigraphic sequence of certain cultural features that seem to be significant. Only two wood-charcoal C-14 samples were recovered, and the dates fit the sequence reasonably well: 970 ? 90 B.P. (Gak 2298)=A.D. 980 for the upper Birnirk level in the A area test trench, and 840 ?90 B.P. (Gak 2297)=A.D. 1110 for Early Thule in the A area test trench. Additional dates were .analyzed using seal bones; the results are, as usual, unsatisfactory. These dates are: (1) level B-10, 1240 ?95 B.P. ( S I - 2 1 5 8 ) = A . D . 710; (2) level B-8, 1695?95 B.P. (S I -2159 ) = A.D. 255; (3) level B-7, 1765 + 90 B.P. (S I -2160) = A.D. 185; (4) level B-6, 1070?90 B.P. ( S I - 2 1 6 1 ) = A . D . 880; (5) level A-5, 1555 ?95 B.P. (S I -2162) = A.D. 395. Of these dates, only one falls close to the expected age, that is, sample 1 (SI-2158), A.D. 710 for Early Birnirk. Birnirk is divided into three phases: Early, Middle and Late. It is suggested that these phases date as follows: Early Birnirk, A.D. 500-700; Middle Birnirk, A.D. 700-800; and Late Birnirk, A.D. 800-900. These dates are arrived at by (1) the approximate date of A.D. 500 for the Kugusugaruk site, which is here considered to be the earliest Birnirk date, and (2) the terminal date of A.D. 900 for Birnirk, because the Early Thule occupation of Nunagiak must have occurred between A.D. 900-1000, as indicated by the rise in the sea level elevation. The break for Middle Birnirk is purely arbitrary, and further dating is being awaited. Early Birnirk is characterized by having harpoon heads with medially placed trifurcated spurs. This type of spur is found exclusively in the lowest level at Walakpa and from two burials at Point Hope. The spur is also thought to have developed from Old Bering Sea, where many harpoon heads have trifur? cated spurs. The major differences between Birnirk and the earlier Eskimo cultures seem to be (1) that the Birnirk open-socket harpoon head only occurs in Birnirk, and (2) there is a decreased emphasis on decorative motifs in Birnirk. Middle Birnirk is a relatively subjective phase. It is typified as having harpoon heads with bifurcated spurs. This type of spur is found exclusively in two levels (B9 and B8) at Walakpa, as well as at other sites. There are variations and overlaps in spur styles; the sequence is not as clear-cut as it should be. How? ever, it seems that the vast majority of harpoon heads from this phase have the bifurcated spurs. Many others have ornamental third spurs, but it can usually be demonstrated that these are strictly orna? mental with the basic spur being bifurcated. These ornamental spurs may represent a transitional phase between the trifurcated and bifurcated spurs. Late Birnirk did not appear at Walakpa but was found at mound A in the Birnirk site. The harpoon heads from this phase have single-pointed spurs with decorative medial lines recalling the bifurcated spur. Sometimes these medial lines separate tiny points, which also give the impression of an ornamental bifurcated spur. A decorative single point placed laterally at the top of the spur also recalls the bi? furcated spur. This last feature occurs in the Early Thule phase. The Thule sequence is divided into two phases, Early and Late. Early Thule dates from A.D. 900 to A.D. 1400, and Late Thule lasts up to historic contact. The major difference between Late Birnirk and Early Thule is the dropping of the decorative medial line on the harpoon head spur and the disuse of many of the classic Birnirk harpoon heads, particu? larly the Birnirk open socket variety. The major differences between Early and Late Thule are that Early Thule has the classic Thule harpoon types and antler arrowheads with rounded shoulders. The Late Thule harpoon heads are obviously related to the Early Thule types, with many modifications and pro? liferations of the Early Thule types. T h e major dif? ference is the change from round-shouldered antler arrowheads to the square-shouldered types and the use of ownership marks in the Late Thule phase. The date of A.D. 1400 is used for the transition because of the cooling of the climatic conditions and the lower? ing of the sea level, which occurred at this time. These two events probably triggered the beginning of communal whale hunting and the development of large villages?both of which are characteristic of Late Thule in northern Alaska. C u l t u r a l Analys i s BIRNIRK The Birnirk phase at Walakpa was found in both middens and underlays Early Thule cultural levels. ARCHEOLOGICAL ANALYSIS OF CULTURAL UNITS 91 In the A area of excavation, a Birnirk level was en? countered in the 1968 test trenches. Only the upper Thule levels of A area were examined in 1969, but the evidence indicates that it may contain several distinct levels of Birnirk occupation. The lower por? tion of the unit contained the earlier Naulock and Birnirk types. No structural features were excavated in the A area. The B area contained three Birnirk-occupation levels. These consisted of a permanent house, a pos? sible tent structure and three levels of trash middens. Estimation of population size is impossible without further excavations in the A area. However, the size of the A midden would indicate that it contains no more than one additional house, and it is doubtful that more than two households could have camped here at any one time. The lowest Birnirk level was B-10. The harpoon heads from this level are all early types and may date this level as being one of the earliest Birnirk occupa? tions in the Point Barrow area. This unit consisted of a trash midden and several upright posts. T o the southwest of the trash midden there was an activity area defined by a large area of matted caribou hair. Associated with the caribou hair were several two- handed scrapers and a log. The insides of the two- handed scrapers were matted with caribou hair as was one end of the log. Caribou skins were placed over the log, and the scrapers were used to pull the hair away from the hide. The concave portion of the scrapers matched the upper diameter of the log. From the faunal analysis of this level it appears that level B-10 represents a continuous habitation of several months duration. The dentaries on the juvenile caribou mandibles and the epiphyseal closure of other bones indicate that the animals were killed during the winter. The exceptionally high number of seals points to a long period of habitation. The remains of 162 seals were recovered from this level which represents a minimal number. This would average out to represent about a seal a day during the winter. This can be compared to the 18 seals re? covered from level B-8, which apparently represents a short-duration Birnirk hunting camp. Many of the seals and caribou may have been killed locally. How? ever, from the occurrence and absence of the animal elements it appears that as high as 37 percent of the seals and 42 percent of the caribou may have been killed at other places and their meat transported to this camp site. The number of walrus and bearded seal bones is significantly lower than would be ex? pected. Level B-9 appears to represent a spring-through- fall habitation. This is indicated by the eruption and wear on caribou dentaries along with an extremely high proportion of immature animal bones. More? over, the wear pattern on the seal teeth found in this level is significantly different from those of the pre? sumed winter levels. A high number of walrus, bearded seal, and polar bear bones may also indicate that this level was occupied from spring through fall. A possible tent structure was discovered in this level (Figure 23). Three logs were found placed in a roughly U-shaped rectangle. Perhaps they had been placed over the lower edges of a tent-covering to hold them down. Eight horizontal planks make up an entrance platform near the center of the open end. Upright legs were placed at the corners where the logs come together. Between the entrance platform and the sides, there were several upright logs along the front edge. These uprights were probably used for the tent framework. Two upright logs in front of the structure and slightly to the west possibly served as drying racks. These logs were placed about 1.2 meters apart. All of the upright logs were cut off in the upper levels, and these portions of the logs were used for other purposes or by later inhabitants. Several other miscellaneous upright logs were found, but it is tsr^l = Horizontal Logs \(^)\ = Vertical Logs I I = Bones FIGURE 23.?B-9 occupation floor showing logs possibly used to hold down the tent walls, and logs used for entrance way. 92 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 [gag 'Horizontal Logs |, M rd M X ? H u ? H CO r H A) > OJ ? 3 c 0 ? r l 4 J (0 t> ft C 3 3 0 0 0 S O A-1 A-2 A-3 A-4 A-5 A-6 A-7 B-1 B-2 B-3 B-4 B-5 B-6 B-7 B-8 B-9 B-10 01 c ? H i d E OJ 1 3 ? C 0 a o i d 0 ELJ ? Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Whale Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Seal Caribou Ugruk Walrus Polar Bea Whale Bird Fish Seal Caribou Ugruk Walrus Polar Bea Whale Bird Fish Seal Caribou Ugruk Walrus m >44 r H o (d e V 4 - H rss B '-i i s id 3 E T I 3 - H ? > - H - H C 7 3 ? H C S H 13 4 11 6 6 4 10 1 12 6 13 3 7 3 7 3 0 6 2 5 3 17 2 6 1 12 2 13 4 IS 4 2 1 r 1 1 + + 68 11 3 3 - 3 1 + + 162 12 7 3 Polar Bear 3 Whale Muskox Bird Fish 1 1 + + i n r C -P t j i OJ ? H X X CJ i d 01 P J r H r Q r P ? H i d 01 OJ o i ? o 6 J T J r H r H 0) i d - H JJ >H 0 EH > , t j l t - l V4 0 0) c ?a? w 27. 16. 35. 28. 21. 13. 35. 8. 26. 26. 33. 34. 27 15 22. 08. 24. 00. 35. 16. 18. 25. 21. 16. 32. 16. 46. 19. 33. 15. - 65. 64. - - 67. 40. 01 4-> S 0 o x J u i d 0) n r H r Q "K ? H rd 0) 0) o) ? 0 r H r H 0) i d - H ? P >H 0 E H > , 0 ! 4-1 , J 0 0) c ^ w 67. 32. 74. 67. 57. 33. 54. 51. 41 . 23. 52. 48. 56. 34. 31. 44. - 32. 29. 35. 16. 57. 25. 30. 19. 40. 09. 34. 27. 61 . 22. - 53. 53. 52. 53. - - - 0) - H A ? H 01 01 0 0) CL, 4-> OJ r H X i d o p rd 0 ft. EH OJ UH 4-> 0 01 rd ^ 2 09. 21. 21. 36. 15. 10. 05. 11. 05. 44. 08. 16. 05. 09. 06. 16. - 12. 12. 18. 08. 06. 12. 11. 26. 09. 25. 15. 23. 09. 14. 20. 24. 15. 44. - ? d 0J V i d ? H 0 0 01 01 < r H IB c 3 oi rdrH lu 0 0 i n E-i 0 tn u c rj j-H fl+J @ c 3 3 2 K 2 5 7 29 8 16 5 8 11 47 4 32 2 11 6 2 2 0 1 2 6 4 2 3 5 5 7 15 13 21 4 . -0 5 2 5 1 - -0 7 2 27 2 0 0 0 0 0 5 1 c 0 ? H 4 J i d ? P n 0 ft 01 c rd l-i e-i " 0 4 J OJ rd r H 0 in m 0 0 - ~ 0 0 1 0 0 0 1 0 0 0 - - 0 0 0 0 - 0 0 0 0 0 0 0 3 0 0 0 0 _ 3 0 - - ~ 3 0 3 5 - _ _ r? V ? H > ? H ? P O <: > i 4-1 4-1 ? H U c id 3 144 e 4-i 0 b u < 0 ? 0 0 0 0 0 _ 0 0 0 0 0 - 0 - 0 0 _ 2 - 2 5 - _ _ - OJ & EH r4 OJ P r H OJ r C in T ? T - T T T T - T T T T T T T T _ S _ - T S? - > i r Q T l 0) C ? H e i\1 ?. r-0J CO C 4-1 - H 0 OJ CO - H a > i 4 J r H ft o i rd O 3 rH ? rd C r H C O n ) 0 o i c in ii 3 m cu id OJ CO fc. CO Late Fal l 2-Fal l ~ Late Fal l 2-Fal l - Late Fal l 2-Fal l Early Winter 2-Fal l Late Fal l 2-Fal l - - - Spring Winter 3-Winter - ?_ Winter 3-Winter - _ - _ Winter 3-Winter _ _ Spring-Fal l 1-Summer - - Fal l -Winter 1-Winter - + Present in undetermined quantities. Missing data, T Tent, S Sod. ARCHEOLOGICAL ANALYSIS OF CULTURAL UNITS 95 there were over two tents here at any one time. The tent remains were similar to that described for the B-9 level. They consisted of several logs placed horizontally in subrectangular patterns with hori? zontal planks making up entrance porches. The interiors around the logs were remarkably devoid of cultural debris. Only the central portions of the tents contained debris accumulations. In each tent, the clear area between the logs and the debris is believed to have been storage or sleeping areas, with either con? tainers or skins having been placed on the floor. Most of these floors were covered with scattered cut sod. The Late Thule phase at Walakpa is defined on the basis of ownership marks on the antler arrowheads. Ownership marks are identification signs that are carved on the body of the arrowheads. These marks occasionally occur on other objects but are found primarily on antler arrowheads and whaling harpoon heads?both communal activity tools with the marks serving to identify the killer or owner of specific animals. Ownership marks, according to Spencer (1959:150) are not inherited, but a man designs his own mark when he feels he has sufficient equipment and prestige to merit having a mark. His mark is related to his father's in that it is similar in the use of the same motifs but with slight variations. Brothers have closely related marks. When examining the marks from Walakpa (Figure 27), we see that there is a close relationship between the marks from the earliest level until the historic period. These marks are variations of slanted lines around a single straight line. Although it is not con? clusive, I would suggest that these marks may indicate that the same family used Walakpa for their fall- hunting camp over the generations from perhaps before A.D. 1400 until the historic period. The same historic ownership mark occurs at Nuwuk, and it may not be too far out of line to suggest that the family who camped at Walakpa were Nuwukmuit. Table 9 is a summary of the artifact and faunal remains from each of the levels at the Walakpa site. Comparison of Walakpa and Other Birnirk and Thule Sites Most archeological reports on Eskimo cultures contain detailed comparisons of trait lists. These lists are designed to show similarities and differences be? tween various phases of Eskimo prehistory. As more sites are found and reported, various phases become less distinctive and often they may now be grouped together into larger cultural complexes. T o illustrate this point, one can examine Ford's (1959:241) trait comparisons of Birnirk and Canadian Thule. He described 46 traits that tie Birnirk to Thule . How? ever, from Walakpa and other excavations in Alaska and Canada, his trait list could now be expanded to include Thule II, Thule m, Thule iv, Sicco, and Tasik harpoon heads, tapering tang antler arrow? heads, and seal scratchers as occurring in both Birnirk and Thule. In another list, Ford (1959:241) described eight traits that first appeared at Nunagiak and were shared by Canadian Thule. These traits are (1) Thule II harpoon heads, (2) Tasik harpoon heads, (3) arrowheads with knobbed tangs, (4) harpoon socket pieces with wedge-shaped tangs, (5) bird dart side prongs, (6) bladder dart socket pieces with tubular and split tangs, (7) double-crescent thimble holders, and (8) gull hooks. Of this list, which was designed to prove that* Nunagiak is the cultural link between Birnirk and Thule, the first four traits are found in collections from the Birnirk sites. As Taylor (1963:460) points out, these finds seriously weaken Ford's argument that Nunagiak was the transitional phase and thus brings Birnirk and Canadian Thule closer together. Ford (1959:241-242) in another comparative list, notes 34 traits that first appeared in Canadian Thule and are found in later Point Barrow phases. But six of these traits are now found in either Birnirk or Northern Alaskan Early Thule levels at Walakpa. These are (1) single-edged, slate knife blades, (2) snow probes, (3) men's knife handles, flattened, oval in cross-section, (4) picks and mattocks with lashing notches, (5) crooked knives, and (6) cylindrical ulu handles. Traits that are now found in Thule but not in Birnirk are (1) loose lance heads with open sockets, (2) fixed lance heads with wedge-shaped tangs and side blades, (3) weapon points of baleen, (4) square- shouldered antler arrowheads with spurred tangs, (5) Late Thule harpoon heads, (6) Thule bird darts, (7) barbs for salmon spears, (8) broad snow knives, (9) whale bone shaves, (10) winged needle cases, (11) lamps, and (12) soap-stone cooking pots. T h e present writer believes that many of the items in the above list will be dropped as exclusive to one or the other cultural complexes when more Birnirk sites are excavated and described. T o show that there was continuing influence from the Bering Straits region to the Point Barrow area, Ford (1959:242) pointed out traits shared between Punuk and later Point Barrow cultures. In this list he included the seal scratcher and bladder-dart mouth? pieces. Both of these items were found in earlier cultural units at Walakpa. T h e seal scratcher, as mentioned above, was found in the Birnirk unit, while an excellent example of a bladder-dart mouth? piece was found in an Early Thule level. Four traits supposedly link Okvik of the Bering Sea area and Nuwuk-Utkiavik (Ford, 1959:242). Among these are blunt, bird arrowpoints and three-piece adzes with handles lashed to the socket at right angles. The three-piece adz occurs in both Birnirk and Early Thule and the blunt, bird arrowpoint is found in Early Thule occupation at Walakpa. The use of trait lists was of value in early days of Eskimo archeology. At the time the exact relation? ships between far distant cultures was not known. These lists, during the 1920s and 1930s, helped to show the relationships of these cultures from Green? land to Siberia. Now, however, we should turn to more sensitive diagnostic comparisons. It is felt here that the most diagnostic traits that can be used to indi? cate Eskimo cultural development and relationships are harpoon heads and, possibly, antler arrowheads. While their known distribution and chronological positions are still somewhat confusing, these artifacts, nevertheless, show general style changes through time. Few sites have been found in which the stratigraphy 96 COMPARISON OF WALAKPA AND OTHER SITES 97 indicates temporal positions of various Birnirk and Thule harpoon heads and arrowpoints. In fact, in most sites, these artifacts have been discovered in separate middens and it is unfortunately impossible to tell which midden accumulated first, as C-14 dating is not that sensitive. Therefore, the stratigraphic continuum of the 17 levels of the Walakpa midden is important in that it precisely indicates the relative temporal positions of these artifact types. Therefore, the harpoon heads and antler arrowheads found in these levels will be compared with other sites of Birnirk and Thule age. A second type of comparison will be made between Birnirk and Thule cultures on the basis of settlement size for the purpose of indicating possible group size and showing any significant change through time. This latter comparison is tenuous and is based largely on circumstantial evidence because many investigators generally fail to seriate habitations accurately into horizontally comtemporaneous units. N o r t h e r n Alaska Most Birnirk sites occur in Northern and Western Alaska, and none have been found east of Point Barrow. Few of these sites have been excavated, and others, for various reasons to be explained, are not believed to be Birnirk by the present author. BIRNIRK (Figure 28:9).?The Birnirk type-site, located midway between Utkiavik and Point Barrow, consists of 16 mounds. Seven of these mounds were excavated in part by Ford (1959). Three were ex? cavated by Wilbert Carter of Tufts University be? tween 1951 and 1953. Other excavations were con? ducted at Birnirk by Eskimos who were employed by Stefansson (1914:393), but the exact location of where they worked is uncertain. The mounds represent a long period of occupation, which may be as early as Old Bering Sea in age (Carter, pers. comm.), with occupations during all the Birnirk phases and includ? ing a Thule occupation. The Birnirk occupation of the site can be grouped into three periods based on harpoon head types (Table 10). The earliest Birnirk occupations are found in mounds H, Q, and K. This phase, as at Walakpa, is marked by harpoon heads with medially placed, tri? furcated spurs. They include the types of Birnirk, Naulock, Oopik, and Tuquok. Middle Birnirk arti? facts, similar to those in levels B-9 and B-8 at Wa? lakpa, are also found in mounds H and Q. These harpoon heads have bifurcate, laterally placed spurs. These were also found in mound B and C of Ford's excavations. The types represented here are Naulock, Birnirk, Natchuk, and Tuquok. Late Birnirk is well represented in the various structures of mound A, but also occurs in mounds J, R, H, and Q. This phase has the simple lateral spur, with a decorative, medial line. These spurs are found on Naulock, Tasik, Birnirk, Tuquok, Oopik, Katoktok, Tipiruk, Thule II, Natchuk, Sicco, and Alilu. Late Birnirk was not found at Walakpa. The harpoon heads that were found at Birnirk, but were not found at Walakpa, are (1) Oopik, an open socket type with either lashing slots or grooves, barbless blades placed at right angles to the line hole, inset-side blades placed in the same plane as the line hole, and spurs with either single or multiple points placed either directly below or at right angles to the open socket (Ford, 1959:77, fig. 27b-4); (2) Katoktok, which is very similar to Oopik, except that it does not have the inset-side blades (Ford, 1959:80, fig. 28g), and (3) Tipiruk, which has an open socket, lashing slots and blades without either barbs or end blades (Ford, 1959:80, fig. 28h-l) . The Thule occupation of the Birnirk site is found primarily in mound H. It appears to be late in the Thule sequence and has only Thule II harpoon heads. Carter (pers. comm.) feels that the elevation of mound H is high enough to have remained dry during periods of high sea levels. However, it can be pointed out that this Thule occupation could have coincided with the drop in sea level of A.D. 1400/1500. The antler arrowheads that I have been able to examine from Birnirk tend to follow the same se? quence as is seen at Walakpa, with the rounded shoulder, straight tang varieties being common in the early phase structures, while the rounded shoulder, conical tang variety is typical of the late structures in mound A. Two whaling harpoon heads should be noted as having been found at Birnirk. Ford (1959:103) reported one from mound A, which has a round Thule type of line hole, but the angular shape is suggestive of Punuk-whaling harpoon heads. Carter (pers. comm.) found a whaling harpoon head, along with other whaling gear in mound Q. The harpoon head from mound Q is much more Punuk looking than is the example from mound A. It is made of ivory, with an elongated, oval line hole, which is placed near the lower margin of the head. Of the two harpoon heads, I would say that Carter's specimen is much earlier than is Ford's, which was found with Late Birnirk harpoon head types. The only piece of Birnirk whaling gear found at Walakpa, was a large harpoon socket piece (Plate 16??). Little evidence is found at Walakpa of whale hunting, and Carter and I feel that the Birnirk hunters did not regularly hunt whales. 98 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 FIGURE 28.?Polar projection of the locations of Birnirk, and Early and Late Thule sites. Although the sites are on the coast, for emphasis they are shown as large dots and are designated by numbers with their names and cultural affiliation as follows: 1, Cape Baranov (Late Birnirk); 2, East Cape (Early and Middle Birnirk, Late Thule); 3, St. Lawrence Island (Early Birnirk); 4, Cape Denbigh (Early and Late Thule); 5, Cape Prince of Wales (Birnirk, Early and Late Thule-Punuk); 6, Cape Krusenstern (Birnirk, Early and Late Thule); 7, Point Hope (Early Birnirk, Early and Late Thule); 8, Nunagiak (Early and Late Thule); 9, Point Barrow sites: Birnirk site (Early, Middle, and Late Birnirk, Late Thule); Utkiavik (Late Thule); Koguk (Early and Middle Birnirk); Nunavak (Early, Middle, and Late Birnirk); Walakpa (Early and Middle Birnirk, Early and Late Thule); 10, Hershel Island (Early Thule); 11, Vaughn (Early Thule); 12, Jackson (Early Thule); 13, Memorana (Early Thule); 14, Lady Franklin Point (Early Thule); 15, Pembroke (Early Thule); 16, Clare (Early Thule); 11, Bell (Early Thule); 18, Ml (Early Thule); 19, Ruin Island (Early Thule); 20, Umanag (Early Thule); 21, Nugdlit (Early Thule). Carter (pers. comm.) estimates that the occupation of the Birnirk site during any one phase was rather small, consisting of possibly two to three houses. This fits the estimates of the size of the occupation at Walakpa. NUVUWARUK (Figure 28:9).?Nuvuwaruk, located at Browerville, near Barrow Village, consists of three house mounds (Ford, 1959:55-56). Ford made several minor tests at the site finding only one Thule n harpoon head and no antler arrowheads. T h e place? ment of Nuvuwaruk into the Thule sequence is impossible on the bases of the evidence at hand, and it seems unlikely that any further work will be done at this site, as Browerville is now built up. UTKIAVIK (Figure 28:9).?The Kugok burial mounds excavated by Ford are located at Utkiavik (1959:25-32). These mounds contained Birnirk har? poon heads of the trifurcate spur, Birnirk open socket, and the Naulock bifurcated spur types, as well as artifacts that have Old Bering Sea styles of decora? tion. Ford (1959:31) thought that these burials may be slightly earlier than most of the material from the Birnirk site. This was because of the Old Bering Sea decoration. Now, with the stratigraphic placement at Walakpa of the trifurcate-spurred harpoon heads as being early Birnirk, it seems that Ford's assumption was correct. Most of the house mounds at Utkiavik represent Late Thule occupations. As the present village of Barrow is built over the site it is nearly impossible to COMPARISON OF WALAKPA AND OTHER SITES TABLE 10.?Summary of occurrence of various Birnirk and Early Thule harpoon types in the Point Barrow area 99 Birnirk and Early S i t e s Walakpa Birnirk Kugok Nunagiak Kugusugaruk Nunavak M o u n d A B B B B B A B C D J R H Q C Thule St ru ct ur e L e v e l Early Thule Test Trench Early Thule Test Trench B-7 B - 8 B-9 B-10 A B C E F H Punuk House 1 2 - 3 - 4 5 1 - 3 Media l ly Placed Tr i furca ted Spurs Na ul oc k Bi rn ir k Tu qu ok Oo pi k K a to kt ok T ip ir uk - " - - _ _ _ _ 1 5 1 1 - - - - - - - - - - - - - - - - - 1 - 3 2 1 4 4 1 - 1 - - - _ - - - 4 - _ _ _ - _ - - - - - - 3 - - L a t e r a l l y Placed Bi furcated Spurs Na ul oc k Bi rn ir k Tu qu ok Oo pi k K a to kt ok " - " - 4 1 - - - - - - - - - - - - - - - 1 - - 1 1 - - _ _ - " - 2 1 3 2 2 1 - - 5 1 1 1 - 1 1 - 2 - - 1 1 - - 3 u ?H ft ?H EH - - - " " - " Na tc hu k 2 2 - - - - - - - - L a t e r a l l y Placed S ing le Spur Decorat ive Medial Line Na ul oc k Ta si k Bi rn ir k Tu qu ok Oo pi k Ka to kt ok Ti pi ru k Th ul e II Na tc hu k 1 1 - 1 _ i - - - 1 - - - 1 - _ 1 - 2 - 1 - 1 2 1 - - - - 1 1 2 2 - - - 1 2 3 - - - - 7 1 - - 1 2 1 - 2 - - _ 1 - - - 1 - 1 - - 1 - 3 4 5 1 2 1 3 4 2 2 1 - - 1 - - - _ - _ - - - 1 - 2 1 2 1 - - " " 1 - - 1 2 - - 1 - - - 1 - _ _ _ with Si cc o T hu le II I 1 1 1 - - - 1 - - - 1 - - - - - 1 1 - - - - - - - Cl os ed So ck et - - 1 - 1 1 - - - 4 - - - - - 100 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 tell how many houses were once there. In 1968 we mapped over 17 houses. Reports from villagers indicate that many of the modern houses were built over older houses with the older houses being de? stroyed. It is also reported that many houses have been washed into the sea. However, it appears that this village must have been quite large in former times. During 1968, we were able to test several areas of the site as Ford (1959:67-74) had earlier. The harpoon head types that were found here are Utkiavik, Nunagiak, Cape Smyth, Nuwuk, Brower, Kilimatavik, Kuk, Thule II, and Thule n with an extra barb. T o summarize the harpoon heads mentioned above, which have not been previously described: Cape Smyth is a closed socket variety, which resembles the Nuwuk type, but has blunt, end blade prongs, which are grooved for binding a thick end blade (Ford, 1959:89, fig. 32a-d); Brower is a closed socket harpoon head, which is extremely similar to Nunagiak, but is slotted for an end blade (Ford, 1959:90, fig. 32e-g); Kilimatavik has closed sockets, multiple barbs placed in the same plane as the line holes. One side is flat with the other side being round with the spur placed on the round side of the harpoon (Ford, 1959:89, fig. 31k-p) . Barrow is the same as Nuwuk but sig? nificantly larger (Ford, 1959:92, fig. 32k-n). The antler arrowheads have square shoulders and spurred tangs. The harpoon heads found at this site are similar to those from the Late Thule levels at Walakpa. Many of the types listed above were not found at Walakpa except in the historic levels. These are Kuk, Cape Smyth, Brower, and Kilimatavik. These types were also found at the historic village of Nuwuk, which is located at the tip of Point Barrow. It is thought here that these types are either protohistoric or his? toric in this area. The Barrow and Nunagiak types were not found in any of the Walakpa levels. Of the harpoon heads found at Walakpa only one type was not found at Utkiavik. This is the Thule nd as defined in this paper (p. 22). The antler arrowheads from Utkiavik are nearly identical with those from the Late Thule levels at Walakpa. NUNAVAK (Figure 28:9).?The Nunavak site is located about 6.5 kilometers southwest of Barrow Village. This site consisted of two localities, one on the ridge north of Nunavak Lagoon and the other on the south side of South Nunavak Lagoon. W. B. Van Valin, a school teacher from Point Barrow, acting under a contract from the University of Pennsylvania Museum, collected several mounds, but I was unable to determine at which locality he excavated. House mound one contained Early Birnirk-Naulock harpoon heads with trifurcated spurs, as well as the Late Birnirk types of Tasik and Natchuk. It is felt here that there were two occupations of this mound, one in Early Birnirk and one which was near the end of the Late Birnirk phase. House mound three con? tained the Naulock and Birnirk open socket harpoon heads with bifurcated spurs. Alfred Hopson, a resident of Barrow village, con? tinued Van Valin's work at Nunavak for the Univer? sity of Pennsylvania during the 1920s. He excavated a series of burials, but few artifacts were found as the preservation was poor. It is not possible to tie these burials in with the various mounds. KUGUSUGARUK (Figure 28:9).?Dated between A.D. 318 and 804 (Rainey and Ralph, 1959:367), this site consists of 6 mounds about 16 km southwest of Barrow. It was excavated by Van Valin between 1917 and 1919. The results of his work are published in Mason (1930). There is some confusion as to the location of the site, as Mason describes a place that could only be Utkiavik, and the artifacts are described and cataloged at the University of Pennsylvania Mu? seum as having come from Utkiavik (Barrow Village). However, when Ford (1959:19) was in Point Barrow in 1931-1932, Hopsen and others showed him where Van Valin had excavated, which was indeed at the Kugusugaruk site rather than at Utkiavik. The site consists of several house mounds, as well as several charnel houses. Here, as at other Birnirk sites, all the phases of Birnirk culture are represented by the harpoon heads found in the various struc? tures. Structure one seems to be the oldest, with harpoon heads that have trifurcate and bifurcate spurs of the Naulock and Birnirk types. Houses two, three, four, and five contained har? poon heads with bifurcated Birnirk and Tuquok types of the later single-spurred varieties, house four had one Tasik and house 5 had one Birnirk, and two Oopik. The antler arrowheads, which were found with the various harpoon heads, conform to the Birnirk se? quence at Walakpa. T h e rounded shoulder, slightly bulbed, straight tang varieties occurred with the earlier harpoon heads. Conical tang, round shoulder examples occurred with the later varieties. Between Kugusugaruk and Nunagiak, with the exception of Walakpa, no Birnirk sites have been re? ported. There are at least four areas in which house mounds have been reported (Ford, 1959:18). Al? though these are reported as recent, some of these may be Birnirk or Thule in age. One of these, Atanik (Figure 28:8), is indeed a historic village. There are numerous house mounds at Atanik and the situa? tion of the site on Peard Bay is similar to the situation COMPARISON O F WALAKPA AND OTHER SITES 101 of the Birnirk site on Elson Lagoon. I would suspect that several of these house mounds may be of an earlier age. The rest of the sites listed by Ford were not occupied or at least have not been referred to as sites occupied during the historic period. As Ford did not excavate these sites and as historic artifacts are found on the surface of most areas where there is good hunting, it may be possible that some of these sites are Birnirk in age. NUNAGIAK (Figure 28:8).?The Nunagiak site, located 18 miles northwest of Wainwright, Alaska, consists of thirteen mounds (Ford, 1959:56). Ford tested seven of the mounds, but concentrated his work in three of the structures. None of the mounds tested by Ford had any classic Birnirk occupations. In mound A, the only harpoon heads found were the closed socket Nuwuk variety. Mound B had at least two occupations, one was Early Thule in age with Sicco and Thule n harpoon heads. Above the Thule occupation in mound B was a later occupation with five Nunagiak-type harpoon heads. Nunagiak harpoon heads, named by Ford, are similar to Thule II, with two symmetrically placed barbs, but differ in that they have closed sockets. Ford (1959:89) contends that these developed from Thule n, but occurred later than the spread of Thule into Canada. At Walakpa, this type was found only in the historic levels, and it appears that Ford was right in that they developed too late to be part of the Early Thule tool kit. Mound C is of considerable interest in that three Punuk heads were found in association with Early Thule harpoon heads. These included the Punuk types of in (a) x and vx as described by Collins (1937c: 118-120). The Early Thule types which were found were Natchuk, Sicco, Katoktok, and Thule n. It is not clear from Ford's monograph as to what he thought the differences between Punuk m(a)x and Sicco are, as he feels Sicco is a direct copy of m(a)x (Ford, 1959:86). Sicco may better be considered a Punuk-Late Birnirk cross-tie, as are the whaling harpoons from the Birnirk site. The Birnirk types of Natchuk, Katoktok, and Thule II are considered here to be Early Thule, as their spurs in all cases but one are laterally placed and do not have the decorative medial line. I feel that these may represent a very early stage of Early Thule, as the Katoktok specimen does have a medial line and all of the Natchuk examples have ornamental side-blade slots. Furthermore, a Thule n harpoon head, purchased by Ford from an Eskimo who dug at mound C, had a decorative spur similar to the Thule II from level B-8 at Walakpa. However, the Nunagiak specimen did not have the elaborate decoration of the Birnirk example from Walakpa. The antler arrowheads that were found in these mounds were of the rounded shoulder, conical tang variety. A spurred-tang arrowpoint was found near the entrance of the house in mound C, but it may be intrusive as this area was disturbed by relic collectors. Mound C, which is apparently the earliest house at Nunagiak, was built below the present water level of the lagoon (Ford, 1959:58). Ford's conclusion was that the area has subsided since the house was con? structed. However, if the house were built during the period when the sea level was lower than it is now, the subsequent rise in sea level might account for the flooding rather than the subsidence of land. Further, if the dates of A.D. 1000/1100 (Hume, 1965:1166) are correct for the rise in the sea level, one must infer that mound C at Nunagiak was built before that time. As pointed out above, the harpoon heads found at this mound appear to be extremely Early Thule in age. This would date the beginning or transition of Birnirk to Thule at some time prior to A.D. 1000/ 1100. POINT H O P E (Figure 28:7).?There is evidence of an Early Birnirk occupation at Point Hope. Three Birnirk burials were excavated by Larsen and Rainey (1948, pi. 87:1-4), which contained the Early Birnirk, trifurcated spur harpoon heads of the Tuquok and Naulock types. The Tuquok harpoon head from burial T B 171 is of an extraordinary size and may have been an early whaling harpoon head. The antler arrowheads from the Birnirk burials are the straight tang variety and the knobbed tang variety. Both of these types were found in the Walakpa- Birnirk levels and are considered early. No house mounds were excavated that contained Birnirk arti? facts at Point Hope. TIGARA (Figure 28:7).?The Western Thule finds from the Tigara locality at Point Hope, burials 253 and 186, are probably very early Early Thule in age. Two examples of the Birnirk-Natchuk type (Larsen and Rainey, 1948, pi. 88:4-5) and one Thule II with a lateral spur, similar to the spur on the Walakpa specimen from Level B-9 (Larsen and Rainey, 1948, pi. 88:3), were found. The antler arrowheads from these burials were of the rounded shoulder, knobbed tang variety, which is common in the Walakpa- Birnirk and Early Thule levels. JABBERTOWN (Figure 28:7).?The Western Thule house found at Jabbertown is also Early Thule in age. The Thule n harpoon head (Larsen and Rainey, 1948, pi. 95:1) is closely related to the Birnirk-Thule II from Walakpa with a vestigial spur, but this speci? men has an end-blade slot. Plate 95:2 (Larsen and 102 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 Rainey, 1948:175) is an Early Thule Natchuk type harpoon head. The antler arrowheads from this house have conical tangs similar to the Early Thule arrowheads from Walakpa. OLD TIGARA (Figure 28:7).?The village of Old Tigara at Point Hope is probably a Late Thule historic village (Larsen and Rainey, 1948). In 1924 there were 122 house mounds, but by 1940 only 70 had not been washed into the ocean (Larsen and Rainey, 1948:20). The artifacts collected by Larsen and Rainey included the harpoon types of Kilimata? vik and Nunagiak. These were found along with square-shouldered antler arrowheads with spurred tangs. Rassmussen, as reported in Mathiassen (1930a: 154), made a collection from Old Tigara. His col? lection included two Thule nb, as defined in this paper, a variant of Thule 11 with an extra side barb, a typical Thule 11, Nuwuk, Kilimatavik and a variant of Thule iv (Mathiassen, 1930a, pi. 12:1-7). It can be seen that this village has a long period of occupation; in fact, the village of New Tigara is simply an extension of this older site. The artifacts collected by Rasmussen are more like those from the Late Thule occupation at Walakpa than those collected by Larsen and Rainey. As it appears that this village is growing by accretion, and as the older houses are being washed away, it would also seem that some of Rasmussen's artifacts may date earlier than others, particularly those collected by Larsen and Rainey. Rasmussen's Thule nb is identical to those from the Late Thule levels at Walakpa. His Thule 11 variant is very similar to one we collected from a Late Thule house at Utkiavik. The Thule m and Thule iv may fit either into Birnirk or Early Thule. The rest of the harpoon heads found by Rasmussen, and Larsen and Rainey are considered to be either proto- historic or historic in age. Western Alaska CAPE KRUSENSTERN (Figure 28:6).?Giddings (1965) reported both Birnirk- and Thule-age sites from Cape Krusenstern. These sites now have been analyzed, and a manuscript is being written by Douglas Anderson of Brown University. The Birnirk remains from the beach ridges at Cape Krusenstern consisted mostly of stone artifacts, as the wood and bone preservation was poor. Only one harpoon head was recovered, and it was not typical of Birnirk from Point Barrow. This harpoon head had a closed socket and an end blade slit that was situated in the same place as the line hole. Morphologically, it is similar to the Nuwuk variety at Point Barrow, but it is much longer from the line hole to the end blade than are the Nuwuk types. The chipped stone is much finer in technique than are the Birnirk examples from Walakpa. There also seems to be a greater proportion of stone artifacts at Cape Krusenstern, as fewer than two dozen flaked stone artifacts were found at Walakpa. The Early Thule houses at Cape Krusenstern had better preservation than the Birnirk houses. They produced harpoon heads of Natchuk, Thule 11, Thule in, Thule iv, and Sicco-Punuk ma(x) types. These appear to be quite early in the Thule sequence as several examples have bifurcated spurs. The antler arrowheads also appear to be very early with round shoulders and straight, bulbous, or coni? cal tangs. Several other houses appear to be later in the Thule sequence having only Thule 11 harpoon heads, and one has a Nunagiak-type harpoon head. KOTZEBUE (Figure 28:6).?The archeology of the Kotzebue area has been defined by Giddings (1952) and VanStone (1955). From various sites on Kotzebue Sound and the Kobuk River, a fairly good chronology for Late Thule occupation has been worked out. The sequence includes: Ekseavik, A.D. 1400; Old Kotzebue, A.D. 1400; Intermediate Kotzebue, A.D. 1550; and Ambler Island, A.D. 1730-1760. These sites have a slightly different aspect from those of the Point Hope area and the west. In this area there was continued strong contact with the Punuk culture of the Bering Straits, with emphasis on inland fishing. However, all phases contain harpoon heads in the cultural inventory, which can be compared to the Late Thule cultures of the northern coast. The harpoon heads that occur in the Ekseavik collection are Thule 11, Thule in, and a local variety of Nuwuk. The Nuwuk harpoon heads differ from the northern varieties in that they are constricted both above and below the line holes. They also have a local decorational motif that does not occur at Walakpa. The antler arrowheads have square shoulders and spurred tangs. Other than the local embellishment of the Nuwuk harpoon heads, this culture phase is represented by the harpoon heads and antler arrowpoints, which are very similar to those from the Walakpa Late Thule . Old Kotzebue harpoon heads are identical to the Nuwuk type found in the Early and Late Thule levels at Walakpa. The type of decoration found on the Nuwuk harpoon heads here is quite similar to the decoration on those from the Point Barrow area. The antler arrowheads found here are also square shoul? dered with spurred tangs. The Intermediate Kotzebue is probably very closely related to the protohistoric periods of the northern coast. The harpoon heads found here are Kilima- COMPARISON O F WALAKPA AND OTHER SITES 103 tavik, Nuwuk, and a possible variant of Nunagiak. The antler arrowheads have square shoulders and spurred tangs. This culture phase compares very favorably to the historic and protohistoric levels at Walakpa, Nunagiak, and Nuwuk. AHTEUT (Figure 28:6).?Ahteut, originally exca? vated by Giddings (1952), is an Early Thule occupa? tion in the Kotzebue area. It has harpoon heads, which include the types of Thule II and Thule m, Punuk ma(x), which have design elements typical of Punuk. The antler arrowheads have rounded shoulders and knobbed tangs. With its date of A.D. 1250, this phase fits in the Early Thule levels at Walakpa. CAPE PRINCE OF WALES (Figure 28:5).?Two Birnirk sites are reported at Cape Prince of Wales (Collins, 1940:558). One of these sites is Kurigitavik and the other, located 3.2 kilometers north of Wales, is known as the Beach Midden. Neither site has been described in detail. One Birnirk open socket harpoon head was found at the Beach Midden site. However, the head was poorly preserved, and the spur is no longer present. A carbon-14 date was taken from a nearby cut at the same level as the Birnirk harpoon head, and yielded a date of 1480 ?240 B.P. (P-63) (Rainey and Ralph, 1959:368). KURIGITAVIK (Figure 28:5).?Kurigitavik is thought to be a single midden containing several occupations (Collins, 1940:562). An Ipiutak type 1 harpoon head was found in the lowest level. This type of harpoon point is very close to the Oopik open socket harpoon head of the Early Birnirk. It has a medially placed, trifurcated spur. A side blade slot, which is in the same plane as the round line hole, is cut completely through the harpoon head. A carbon-14 date, taken from nine small pieces of wooden shafts, is dated 1320 ?230 B.P. (P-65) (Rainey and Ralph, 1959:368). This date of A.D. 637, indicates that Ipiutak lasted longer in Western Alaska than it did in Northern Alaska. Above the Ipiutak level at Kurigitavik, there oc? curred artifacts that Collins (1940, pi. 16:8, 9) de? scribed as containing Birnirk-like harpoon heads that were transitional into Thule. These Birnirk-like harpoon heads were Natchuk and Thule II with an ornamental bifurcated spur. These Thule n harpoon heads have more of a Thule appearance than Birnirk, and I would place them as being similar to the Early Thule from Point Hope and Nunagiak. The upper levels of Kurigitavik produced harpoon heads that were of the later Thule n types, as well as Punuk harpoon heads. The antler arrowheads from the Early Thule levels at Kurigitavik are, in general, similar to those from Walakpa, in that they have rounded shoulders and conical tangs. The later examples have square shoulders and spurred tangs. NUKLEET (Figure 28:6).?At Cape Denbigh, Gid? dings (1964) described the Nukleet culture. This culture is divided into three phases: Early, Middle, and Late. The earliest phase, according to Giddings (1964:116), dates prior to A.D. 1250 and is related to Punuk with Birnirk cross-ties. The harpoon heads found in this level are Sicco or Thule in and a spurred variety of Thule II. The antler arrowheads from this level are of the Birnirk and Early Thule varieties and not Punuk. These arrowheads are round to oval in cross-section, while, as mentioned above, Punuk arrowheads are triangular in cross-section (Collins, 1937c:222). They have rounded shoulders and both conical and knobbed tangs. I believe that this unit can be equated with the Early Thule occu? pations described above. The Punuk harpoon head type is either Sicco or Thule in, a type found in both Late Birnirk and Early Thule, and other traits such as the antler arrowheads also appear in Early Thule and Late Birnirk. St. Lawrence Island Birnirk-related harpoon heads were found at vari? ous sites in the St. Lawrence Island area (Figure 28:3). These harpoon heads occurred at Seklowaghyaget, Ievoghiyok, Miyowagh, Hillside, Cape Kialegak, Kukulik, and the S'keliyuk sites. All of these sites are of Old Bering Sea or Early Punuk ages. The occurrences of the earliest Birnirk-like types were at Miyowagh and S'keliyuk. The harpoon head that has been named Type nix by Collins (1937c: 104) is very similar to Naulock, but it is decorated with Old Bering Sea art styles. The Type n(a)y harpoon head found at Miyowagh is similar to the Tuquok variety of harpoon head, which has a trifurcated, medially placed spur and was found at Point Hope in the low? est level at Walakpa. Type n(a)y found at Miyowagh, S'keliyuk, and one on Cape Kialegak of St. Lawrence Island, is apparently Early Punuk in this area of the Arctic. Another Birnirk harpoon head found here, Type nx of Collins' (1937c: 102) classification, is similar to Oopik. This type of harpoon head is Old Bering Sea and Early Punuk in age. Type m(a)x in Collins' (1937c: 118) classification, as pointed out above, is identical to the Tasik type. It is Early Punuk in age and lasts in various forms through the Punuk cultural phase. It occurred at the Seklowaghyaget, Ievoghiyok, Miyowagh, and Hillside sites. 104 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 Antler arrowheads from these sites have been classified into at least seven types (Collins 1937c: 135). The Old Bering Sea types range from little or no shoulders to rather square shoulders with tapering tangs. Some have end-blade slots and barbs, while others are just barbed. A Walakpa specimen (Plate 19a) from the lowest Birnirk level is similar to the Old Bering Sea types with end blade slits, but differs in that its shoulder is not as rounded and it has two small asymmetrical barbs. The later Old Bering Sea-Early Punuk types are extremely important, as they show the relationship of antler arrowheads from Old Bering Sea to Birnirk. These are types 5 and 6 in Collins' classification. These have rounded shoulders with slight knobs (Collins, 1937c, pi. 34). They range from single barbs to no barbs. In general these types are identical to those found at Walakpa and the Birnirk site. A Birnirk specimen from Walakpa (Plate 5lp) has the deeply cut barbs and is nearly identical to the one pictured by Collins (1937c, pi. 34:6). The antler arrowpoints from S'keliyuk are in gen? eral of the same types as described above (Ackerman, 1962). However, Ackerman (1962, fig. 3) describes an example of the conical tang variety. The conical tang was found in the Birnirk levels at Walakpa. The Walakpa and the S'keliyuk finds indicate an early date for the appearance of this arrowhead-tang type. The later Punuk type of arrowheads are different from either those of the Old Bering Sea age or of Birnirk. They have as their most characteristic feature triangular cross-sections and most are made of ivory (Collins, 1937c:222). None of these types of arrowheads have been reported from any Birnirk site. Siberia EAST CAPE (Figure 28:2).?At the Uelen and Ekven cemeteries on East Cape Siberia, there were burials that contained Birnirk or Birnirk-related harpoon heads (Arutiunov, Levin, and Sergeev, 1964; Aruti- unov and Sergeev, 1968; Levin 1964a, 1964b; Oklad- nikov and Beregovaia, 1971). The burials, based on Walakpa artifacts, would fall into the Early and Middle Birnirk periods. Burial 17 would be the earliest, with harpoon heads of the trifurcated-spur variety of a Birnirk and Tuquok-Naulock cross. This harpoon head has the inset side blades of the Tuquok, and the end-blade slit of the Naulock types. A closed socket harpoon head was found that fits Collins' (1937c: 109) type nix, with an end-blade slit, which is in the same plane as the line hole. It has a trifurcated spur, with the center point being longer than the two lateral points. A Naulock harpoon head from burial 17 has a single lateral spur with a secondary, decorative point. The harpoon heads from burial 6 are a Birnirk open socket harpoon head that has a bifurcated, lateral spur and a closed socket harpoon head. The spur of the Birnirk type has several decorative points on the outside margin of the spur, which is similar to those from Walakpa level B-8. T h e closed socket harpoon head is broken, and it is impossible to tell from the illustration if the end-blade slit is at right angles to the line hole or in the same plane. Harpoon-head examples of simple bifurcated, spurred Birnirk, Naulock, and Natchuk types were found in burial 5. An Oopik harpoon head was found that had a trifurcated spur without the decora? tive line on the middle point. A closed socket Thule iv harpoon head was also found. This burial appears to be the youngest of the three burials from Uelen. Mathiassen (1930a, pi. 18) pictures several later Thule types from East Cape. These are the harpoon type of Thule 1, Thule in, and a variant of Thule n. Also included in this collection is one square shoul? dered, spurred tang antler arrowpoint. This collection was acquired by Knud Rassmussen from local natives, and it is impossible to tell their exact proveniences (Mathiassen, 1930a:72). CAPE BARANOV (Figure 28:1).?At Cape Baranov near the Kolyma River, several examples of Birnirk harpoon heads have been found (Beregovaia, 1953, Okladnikov and Beregovaia, 1971). Three houses have been excavated at Cape Baranov that are Late Birnirk in age (Okladnikov and Berego? vaia, 1971). All of the harpoon heads from these houses have laterally placed, single spurs with decora? tive medial lines. Only two types were reported, Tuquok and Birnirk. House one contained eight Birnirk open socket harpoon heads and five Tuquok examples. House five produced three Birnirk open socket harpoon heads and three Tuquoks. A house in another nearby site yielded one Tuquok with the same type of spur. These types of harpoon heads did not occur at Walakpa but were found in mound A of the Birnirk site and should be considered Late Birnirk in age. Canadian A r c t i c HERSHEL ISLAND (Figure 28:10).?The first site east of Point Barrow that has an early cultural component is the Hershel Island site (MacNeish, 1959). The size of this site is not reported. T h e harpoon COMPARISON OF WALAKPA AND OTHER SITES 105 types found here are Thule i, Thule n, and Nuwuk. The antler arrowpoints have rounded shoulders with the knobbed tangs. This site also contained Barrow curvilinear pot sherds. The artifact types from this site are mostly comparable to the Early Thule levels at Walakpa. The major difference is that the Thule I harpoon head was not found at Walakpa. This is probably a matter of luck rather than its absence at Walakpa. As it has been pointed out above, the harpoon-head sample size from these levels is very small. It is felt that when further investigations are conducted in the A area of Walakpa, the Early Thule harpoon-head types, which did not occur in the B area, will be found. Barrow curvilinear pottery has been found in Canada and is used to point to Birnirk origins. How? ever, it has been found in all three cultural units at Walakpa and should not be used as a temporal indicator. VAUGHN (Figure 28:11).?The Vaughn site, located at Cape Parry, has two house middens (Taylor, 1972:8). The harpoon head types are Tasik, Thule II, and Nuwuk. These harpoon heads have lateral spurs that do not have a decorative medial line. The antler arrowpoints have rounded, nearly square shoulders and knobbed tangs. Many traits that have been listed by Taylor give this site a nearly western flavor. These traits are the absence of soapstone, quartzite scrapers (which are common in Birnirk and Early Thule levels at Walakpa), the form of the adz handle, and the incised barb on antler arrowheads. The only feature that is found here and is not found at Walakpa is the tendency of the antler arrow? heads to have square shoulders. This type of shoulder occurs at an early age from this point eastward and is considered by Collins (1951:563) to have developed in the Canadian Arctic during the Thule cultural period. This is substantiated since the square shoulder does not occur at Walakpa until later in the Thule phase. Taylor (1972:12) felt that this site is well-developed Canadian Thule and dates the same as Naujan and the Ml site. I see no reason to date it as Late Thule, as most of the taits listed by Taylor are found in Birnirk and Early Thule levels at Walakpa. It seems logical to place this site chronologically, as being related to the Early Thule levels at Walakpa; but because of the squarish shoulders of the antler arrowheads, it appears to be slightly later than the occupation at Hershel Island. JACKSON (Figure 28:12).?This is a small site ad? jacent to the Vaughn site on Cape Parry (Taylor, 1972:14). It consists of at least three house mounds. The harpoon head types found here are Thule i, Thule II, Thule m (which has side grooves), possibly Alilu, Nuwuk, and two late Dorset culture harpoon heads. The arrowpoints have square shoulders with knobbed or spurred tangs. Here again Barrow cur? vilinear pottery and pebble scrapers were found. Of interest is the occurrence of the Eastern Thule- type mattock handle (Taylor, 1972, pi. 3c) and the spurred tang antler arrowhead (Taylor 1972, pi. 3e, f). Taylor (1972:25) felt that this site represents an established stage of Thule, not proto-Thule, but is a contemporary of Nunagiak. Carbon-14 dates of this site are A.D. 744?100 (M-1509) and A.D. 774?100 (M-1508), but Taylor estimated its age as approxi? mately A.D. 1200. The occurrence of the Thule m with ornamental side-blade grooves and an Alilu harpoon head, along with the pebble scrapers, argues for an earlier age. However, the mattock handle and spurred-tang antler arrowheads suggest that it is probably younger than the Hershel Island and the Vaughn sites. MEMORANA (Figure 28:13).?On Victoria Island there are several sites that probably date the same as the Jackson site. The first discussed is the Memorana site (McGhee, 1971:166). This site consists of four oval-shaped houses. The harpoon types are Thule II and Thule in. Several of the Thule n harpoon heads have a single barb (McGhee, 1971:185, pi. 6b) and may be of the Late Birnirk-Early Thule-Natchuk type. The antler arrowheads have square shoulders and spurred tangs. McGhee felt that this site is fairly old and is contemporaneous to Nunagiak, because the harpoon heads have lashing slots rather than drilled holes (McGhee, 1971:167). Thule n harpoon heads from Early Thule levels at Walakpa have both lashing slots and drilled holes. For this reason and because of the antler arrowheads, I would place this site closer to the Jackson site in age. LADY FRANKLIN POINT (Figure 28:14).?Sicco and Alilu types of harpoon heads were reported from a collection from Lady Franklin Point (Taylor, 1963: 458). Taylor (1972:35) in search of a possible site with these harpoon head types, excavated a site at Lady Franklin Point. This site had 21 house mounds. It produced harpoon heads of the Thule n, the end- blade, slotted variant of Thule n, and the Nuwuk closed socket type (Taylor, 1972:39). These were associated with antler arrowheads with rounded shoulders and knobbed tangs, tapering tangs and square shoulders with spurred tangs. Taylor (1972:39) felt that the artifacts from this site would suggest an occupation no older, and per? haps younger, than Nunagiak. I would concur with his evaluation, but would suggest it may be older than 106 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 the Memorana or Jackson sites, as the rounded shoulder, knobbed tang variety of antler arrowheads arid the lack of soapstone point to a close relationship with both Hershel Island and the Vaughn site. It would be interesting to know the exact provenience of the arrowheads and see more of the house mounds excavated, as there is a possibility that not all of the house mounds represent the same stage of cultural development. This site may have earlier components than is indicated by the brief testing. PEMBROKE (Figure 28:15).?The Pembroke site, located at Cambridge Bay on Victoria Island, has twelve structures (Taylor, 1972:44). Among the artifacts found here are a single-barbed harpoon head and a square-shouldered antler arrowhead. Taylor felt that the harpoon head is of the Natchuk type and hints an early age. Close examination of this harpoon head indicates that it is probably a broken Thule type n which has had the edges refaceted, giving it a Natchuk appearance. This would indicate that this site dates more closely in time to the Jackson and Memorana sites. CLARE (Figure 28:16).?The Clare site, located near the Pembroke site, is apparently a caribou hunt? ing station. It consists of as many as 11 tent rings (Taylor, 1972:46). The predominant artifacts are the same type of antler arrowheads as were found at the Pembroke site. No harpoon heads were reported from here, save a preform which could be either Nuwuk or Thule iv. It is felt that the Clare site dates about the same age as the Pembroke site. BELL (Figure 28:17).?The Bell site on the Ekalluk River on Wellington Bay of Victoria Island has416 houses (Taylor, 1972:54). Harpoon heads found here include both Thule n and Thule in. An antler arrow? head has a square shoulder and offset spurs. This site probably dates the same as the Pembroke and Clare sites. Ml (Figure 28:18).?At Resolute Bay, Collins (1955:50, 51) found three small sites, the Lake site with nine houses, the Ml site with twelve houses, the M2 site with nine houses, and the M3 site with five houses. The Ml site has Sicco and Natchuk har? poon heads in association with Thule n and in. The Natchuk harpoon head has an ornamental side-blade slot. The other sites are contemporaneous to Ml and have Thule type iv harpoon heads. The antler arrow? heads that are found here have rounded to nearly square shoulders with conical knobs and symmet? rically placed spurs. With the presence of the Sicco and Natchuk har? poon heads with the ornamental side-blade slots, the occupation at Resolute Bay appears to be one of the earliest Thule sites in Canada. However, the spurs of the harpoon are not decorated with the medial line, which would place the occupation as similar to level B-7 at Walakpa. Greenland RUIN ISLAND (Figure 28:19).?This small island, located in Marshall Bay of the Thule district in northwest Greenland, has seven house ruins on it (Holtved, 1944:70). T h e artifacts from these houses include several Thule n harpoon heads with the lateral spur, as noted on the Thule n harpoon heads from the Birnirk level at Walakpa (Holtved, 1944, pi. 3:5). Plate 3:8 also illustrated the ornamentation above the lashing slot that occurred on the Birnirk specimen from Walakpa. No antler arrowheads were illustrated from this site. It is felt here that the strong resemblances between the Thule n harpoon heads found at Ruin Island and those from the Western Arctic would date this site as very early Early Thule. UMANAG (Figure 28:20).?This site and Comer's midden are also located on Marshall Bay (Holtved, 1944:112). The site includes 27 houses and 2 midden areas, one known as Comer's midden. From Holtved's charts it is very apparent that all of the houses do not represent contemporaneous occupations. Holtved (1944:146-149) indicates that there are probably four levels of occupations. The harpoon heads found here include Thule i, Thule n, Thule n with end-blade slots, Thule in, Thule iv, Nuwuk, and Sicco. The antler arrowheads range from rounded shoulders with knobbed tangs to rather square shoulders with offset spurs. The presence of the Sicco-type harpoon head and the round shouldered antler arrowheads would indicate that the lower components of this site may be of the same age as Ruin Island. NUGDLIT (Figure 28:21).?The site of Nugdlit is located in the Thule district of northwest Greenland (Holtved, 1954). This site has 62 ruins; they are divided into three groups with two being older than, and one dating from, the 16th century. However, once again it is impossible to determine within these groups if the houses were contemporaneously occu? pied. The harpoon heads found here include Thule n with end-blade slots, Thule n, Thule iv, Nuwuk, and Sicco. The antler arrowheads have rather square shoulders and spurred tangs. The site has been radio? carbon-dated to the first half of the tenth century A.D. (McGhee, 1970:187). I would concur with this early date for this site because of the presence of the Sicco and Nuwuk-type harpoon heads. Other Thule sites in Canada and Greenland (Mathiassen, 1927a, 1930b, 1931, 1933, 1934, 1936; COMPARISON OF WALAKPA AND OTHER SITES 107 Mathiassen and Holtved, 1936; Collins 1951b, 1955; Holtved, 1944, 1954; Larsen, 1934; Manning, 1950; McGhee, 1971; Taylor, 1964, 1967) are considered as well-established Canadian Thule and do not contain the harpoon heads of Natchuk and Sicco, which mark the Early Thule development in Northern Alaska. As these sites are rather numerous and repre? sent a fairly uniform cultural phase, I will not attempt to describe each of them, but I will point out the types of harpoon heads and arrowpoints that are representative of the Late Thule culture in the East? ern Arctic. The harpoon heads found in these Late Thule sites are Thule types i, n, iv, vi, vn, and a closed socket variety with a bifurcated spur. The bifurcated spur is placed directly below the socket. This last variety appears to be a local invention and does not occur in the Western Arctic. The antler arrowheads have rather square shoulders and spurred tangs. Occa? sionally an antler arrowhead is reported with rounded shoulders and knobbed tangs. The size of these sites range from one house mound to more than twenty. The above comparison of the harpoon heads and antler arrowheads from various related culture phases in the Arctic to those from the levels at Walakpa, point to several major trends in the development of Birnirk and Thule cultures. These trends are re? viewed below and summarized in Figure 29. S u m m a r y EARLY BIRNIRK It is increasingly obvious that Birnirk had its roots in the Old Bering Sea-related cultures of St. Lawrence Island. This is reflected in the Birnirk harpoon heads of the Naulock, Tuquok, and Oopik types, which have analogous forms with shared features of medially placed trifurcated spurs with the type n harpoon heads from the Old Bering Sea and Early Punuk phases. The closed socket harpoon heads found in Early Birnirk have as their proto? types the nix and my harpoon heads from St. Law? rence Island. 4 These Old Bering Sea-Birnirk-related harpoon heads turn up along the coasts of both Siberia and Northern Alaska. This is evidenced by the finds at the Uelen and Ekven cemeteries in Siberia and the Early Birnirk finds at Point Hope, the lowest level at Walakpa, and the Birnirk site. The Early Birnirk harpoon heads have as their most characteristic feature, the medially placed tri? furcated spurs. Usually, the center prong of the spur has a decorative medial line, but this is not always the case. The outside lateral edges of the spur and the body of the harpoon heads are sometimes decorated with thin, raised, scalloped lines which have notches cut at their apex. Sometimes the margins of the spurs are serrated with a series of small barbs. Early Birnirk is found at the Birnirk type site mounds H and Q, Kugok, Nunavak structure 1, Kugusugaruk, level B-10 at Walakpa, the Point Hope-Birnirk burials and burial 17 at Uelen, Siberia. MIDDLE BIRNIRK There is a gradual transition from the trifurcated spurred harpoon heads to harpoon heads with bi? furcated spurs. The transition occurs with various combinations, but most commonly there are bifurcated spurs with small decorative third barbs. During this phase of Birnirk the decorative features continue, with ornamental side-blade slots becoming common. However, there is a general decreasing of decorative features. This phase, which should be properly called "Middle Birnirk," is found at the Birnirk site mounds R, H, and Q, Kugok, Nunavak structure 2, Kugusugaruk, and Walakpa levels B-9 and B-8. In Siberia, burial 6 at Uelen is Middle Birnirk in age. During this phase in Northern Alaska, the Nat? chuk and Thule n harpoon heads appear. These early forms of Natchuk at Walakpa tend to be heavier harpoon heads with broad, shorter blades than those of the Early Thule-Natchuk varieties. The Birnirk-Thule n harpoon head from Walakpa has Birnirk decorative features typical of this phase, but has a single-pointed spur. The spur is decorated with a pronounced ornamental barb, which is found on harpoon heads as late as Early Thule, at which time it is finally discontinued. LATE BIRNIRK This final phase of Birnirk is marked by decreas? ing ornamentation of the spur and a higher incidence of ornamental side-blade slots and spur prongs. The spurs are totally lateral by this time and are decorated with medial grooves, but usually have only one point. The late phase of Birnirk was not found at Walakpa, but the .harpoon types from mound A at the Birnirk site were almost exclusively Late Birnirk. Late Bir? nirk also occurred in mounds H and Q at the Birnirk site, mound C at Nunagiak and houses four and five at Kugusugaruk. The only other reported occurrence of Late Birnirk is at Cape Baranov in Siberia where harpoon heads of this description were found in three 108 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 UJ HU L i? UJ 1- < _i UJ _l 3 I > _l EA R * tr z a: m UJ H- < _l ^ tr z cr m DL E O 5 *: C? Z f-'A .-.,?< . X en.?,;/-., *. ' ' * ? ??ft'- -?;&??' ? ? ' m i-f V^m - . ; ? i 35" ? ..r ' . ; > ? " ? , r j* / ? ? _ . . j t ^ ? & . - '"-?'i '? ' ? I T ' S ;*.y ^i-ik -^ >) --C ,; *?r ,^ ??%&& 1? PLATE 4.?Early Thule levels in B area: a, Early Thule level B-4 (sign board indicates this as an Eastern Thule level; stratigraphic units are visible in the wall behind the sign board); b, Early Thule level B-6; c, another view of the B-6 occupation level; d, close-up of square J 20, level B-7 Early Thule. (b, c, d, incorrectly labeled "Birnirk"; note concentration of artifacts). PLATES 123 7 ? ?:' J ^ l H^^^S^k ^ PLATE 5.?Birnirk occupation level B-8: a, house structure (the stratigraphy is visible in the wall); b, detail of square J 17, Birnirk house floor; c, detail of square K 17, Birnirk house floor; d, another view of square K 17, Birnirk house floor. 124 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 PLATE 6.?Birnirk artifacts from the mound A test trench: a, Tuquok open socket harpoon head; b, d, Naulock open socket harpoon heads; c, Thule iv closed socket harpoon head; e, fragment of open socket harpoon head; /, g, fragments of harpoon foreshafts; h-j, harpoon head preforms; k, antler arrowhead; I, n, bone awls; m, gauged drill; o, unidentified chert effigy; to. chert knife blade. PLATES 125 PLATE 7.?Birnirk artifacts from the mound B test trench: a, Birnirk open socket harpoon head; b, c, Naulock open socket harpoon heads; d, Tasik open socket harpoon head; e, Thule iv closed socket harpoon head; /, g, harpoon head fragments; h, baleen snow goggles; i, seal float plug; /, k, ivory wound pins; I, m, wood wound plugs. 126 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 I 5 CM. PLATE 8.?Birnirk artifacts from the mound B test trench: a-c antler arrowheads! d, unidentified bone artifact; e, sinew twister; /, flaking tool. PLATES 127 PLATE 9.?Birnirk artifacts from the mound B test trench: a, wooden spoon; b, adz blade; c, unidentified bone tool; d, paddle tip; e, drilled bear tooth; f-l, bola weights. 128 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 I 1! 5 C M . 2 I N. PLATE 10.?Birnirk artifacts from the mound B test trench: a, b, d, ulu blades; c, men's knife blade; e-g, chert knife blade; h, knife handle. PLATES 129 PLATE 11.?Birnirk artifacts from the mound B test trench: a-c, bone awls; d, e, engraving tools; /, unidentified wooden object; g, h, belt toggles; i, cup and pin game. 130 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 if" 5 CM. 2 I N. K J PLATE 12.?Birnirk artifacts from the mound B test trench: a-d, wedges; e, f, bone scrapers; g-j, gaming pieces. PLATE 13.?Small bow from the mound B Birnirk level test trench. PLATES 131 PLATE 14.?Birnirk artifacts from the mound B test trench: a-e, unidentified wood and baleen objects. 132 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 15.?Birnirk artifacts, level B-10: a, Naulock open socket harpoon head; b, Tuquok open socket harpoon head; c, Thule iv closed socket harpoon head; d-h, Birnirk open socket harpoon heads; i, small harpoon foreshaft; j-n, large harpoon foreshafts. PLATES 133 PLATE 16.?Birnirk artifacts, level B-10: a, harpoon shaft ice pick fragment; b-d, harpoon shaft ice picks; e, harpoon socket piece. 134 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 17.?Birnirk artifacts, level B-10: a, float bar; b, c, wound plugs; d, wound pin; e-g, ice scoop rims. 5 CM. 2 IN . PLATE 18.?Birnirk seal rattle, level B-10. 136 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 19.?Birnirk artifacts, level B-10: a, antler arrowhead; b, antler arrowhead fragment; c-f, bola weights; g, foreshaft fragment; h, slate knife blade; i, composite knife handle; 7, knife handle. PLATES 137 PLATE 20.?Wooden hunting hat, level B-10. PLATE 21.?Birnirk artifacts, level B-10: a, boat seat; b, harpoon shaft. 138 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 22.?Birnirk artifacts, level B-10: a, paddle tip; b, wooden shaft; c, d, paddle blades. PLATES 139 PLATE 23.?Birnirk artifacts, level B-10: a-e, whetstones. 140 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 24.?Birnirk artifacts, level B-10: a-c, whetstones. PLATES 141 PLATE 25.?Birnirk artifacts, level B-10: ?, awl handle; b-g, awls; h, i, bow drill bearings. 142 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 26.?Birnirk artifacts, level B-10: a-c, whale rib mattock heads; d, whale rib pick. PLATES 143 PLATE 27.?Birnirk artifacts, level B-IO: a, ulu; b, ulu handle; c-i, ulu blades. c i PLATE 28.?Birnirk artifacts, level B-10: a, b, d-g, k, pebble end scrapers; c, f, h, i, side scrapers;;', flint flaker point. 144 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 29.?Birnirk artifacts, level B-10: a, flaked pebble; b, notched quartz pebble; c, Birnirk split pebbles scraper; d, whetstone; e, ground hematite nodule; /, ground slate nodule. f PLATE 30.?Birnirk artifacts, level B-10: a-e, bone scrapers. PLATES 145 *m VmmYWmm*ummma H B d mmmmm PLATE 31.?Birnirk artifacts, level B-10: a-f, two-handed metapodial scrapers. 146 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 32.?Wooden meat tray, level B-10. PLATES 147 PLATE 33.?Birnirk artifacts, level B-10: a-e, wooden bucket bottoms. 148 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 34.?Birnirk artifacts, level B-fO: a, b, spoons; c, spoon-like artifact; d, flaker hammerhead; e, kayak rib. PLATE 35.?Drum rim, level B-10. PLATES 149 PLATE 36.?Birnirk artifacts, level B-10: a, ivory chain; b, ivory object; c, baleen saw(?); d, toy bow; e, organic doll; /, wooden object; g, cup and pin game; h, baleen object. 150 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 37.?Birnirk artifacts, level B-10: a, Birnirk open socket harpoon head; b, harpoon head fragment; c, d, harpoon head preforms; e, harpoon foreshaft; /, antler arrowhead fragment; g, baleen snow goggles; h, snow probe ferrule; i, dart butt piece; ;', fish arrow prong; k-q, bola weights; r-t, composite knife handles; u, v, composite awls; w, engraving tool; x, awl; y, chert drill bit; z, toggle fragment; aa, ivory pendant; bb, ivory awl. PLATES 151 PLATE 38.?Birnirk artifacts, level B-9: a-c, sled arches; d, kayak rib. 152 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a 5 CM. PLATE 39.?Birnirk artifacts, level B-9: a, b, adz blades; c, ground hematite; d-h, whetstones. PLATES 153 PLATE 40.?Birnirk artifacts, level B-9: a, b, whale rib mattock heads; c, mattock handle. 154 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 41.?Birnirk artifacts, level B-9: a, wedge; b, whale rib ice pick; c, whale rib mattock blade. PLATES 155 PLATE 42.?Birnirk artifacts, level B-9: a, b, ulu handles; c-k, ulu blades. 156 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 ??or- e 2 IN . PLATE 43.?Birnirk artifacts, level B-9: a, stone object; b, c, i, chert bifaces; d, slate knife blade; e, pebble side scrapers; f?h, pebble end scrapers; k, I, retouched flakes. PLATES 157 i ^:m^m - ^ - j * ? ' ? - * ? '????'? ' ' >*." , - ? ? ?:?'?. : r * l V % ; ^ *tr** ? < > ~-?grjflE tf PLATE 44.?Birnirk artifacts, level B-9: ?, flaked and ground slate object; b, c, large, bifacially flaked objects. 158 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 45.?Birnirk artifacts, level B-9: a, b, hammerstones; c, drilled stone object; d, ground slate nodule; e-g, Birnirk split pebble scrapers. PLATES 159 - o ft ? PLATE 46.?Birnirk artifacts, level B-9: a, d, f, g, j , bone scrapers; b, c, e, bone awls; h, i, gauged drills. 160 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a PLATE 47.?Birnirk artifacts, level B-9: a, b, small wedges; c, seal float mouthpiece; d, net weight; e, wooden shaft; /, baleen object; g, wooden bucket bottom; h, baleen bucket. PLATES 161 ? r f^td^^i^^^i^TT-TN - PLATE 48.?Birnirk artifacts, level B-9: a, b, toy bow fragments; c, arrowshaft; d, toy bow. PLATE 49.?Birnirk artifacts, level B-8: a-d, Birnirk open socket harpoon heads; e, f, Natchuk open socket harpoon heads; g, Thule n open socket harpoon head; h, i, harpoon foreshaft frag? ments; ;-/, harpoon shaft ice picks. 162 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 50.?Birnirk artifacts, level B-8: a, wooden snow goggles; b, d, seal scratchers; c, ice staff ring; e, snow probe ferrule. PLATES 163 i; m n 5 CM. 2 I N. % PLATE 51.?Birnirk artifacts, level B-8: a, b, bird-dart side prong; c, d, seal float mouthpiece stoppers; e, f, fish arrow prongs; g, wound pin; h, wound plug; i-l, bola weights; m, ivory ring; n-q, antler arrowheads; r, knapping tool; s, engraving tool. 164 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 52.?Birnirk artifacts, level B-8: a, adz blade and antler sleeve; b, scraper handle; c, knife handle; d, chert knife blade; e, gauge drill; /, g, pebble side scrapers; h, planoconvex end scraper; i, bifacial tool. PLATES 165 PLATE 53.?Birnirk artifacts, level B-8: a, b, ulu blades; c, ground hematite nodule; d, e, g, whetstones; /, ground slate nodule. 166 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 Lix.fe 5 CM. PLATE 54.?Birnirk artifacts, level B-8: a, humerus scraper; b, scapula scraper; c, two-handed bone scraper; d, e, kayak ribs. PLATES 167 5 CM. 2 I N. PLATE 55.?Baleen bucket, level B-8. 168 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 2 IN, PLATE 56.?Polar bear mandible tied with baleen cord, level B-8. PLATES 169 PLATE 57.?Birnirk artifacts, level B-8: a, rawhide cord; b, toy bow; c-e, unidentified wooden artifacts. 170 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 58.?Early Thule artifacts from the mound A test trench: a, Natchuk open socket harpoon head; b, Sicco open socket harpoon head; c, Thule 11 open socket harpoon head; d-g, antler arrowheads; h, planoconvex end scraper; i, toy bow fragment. 5 CM. < ^ J m PLATE 59.?Early Thule artifacts from the mound B test trench: a, b, Thule n open socket harpoon heads; c, d, slate end blades; e, harpoon end blade preforms; /, inflation nozzle; g, ivory labret; h-m, antler arrowheads; n, chert arrowhead; o, bow fragment. PLATES 171 5 CM. 2 IN. PLATE 60.?Early Thule polar bear carving from the mound B test trench. 172 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 ? e Pi 1-4S3 5 CM. PLATE 61.?Early Thule artifacts from the mound B test trench: a, wooden staff with human face; b, c, d, bola weights; e, engraving tool; /, bone awl; g, chert drill bit; h, chert knife blade; i, ulu blade; /, drilled stone. PLATES 173 ? & n B H r a. ? v * C W ^ O * ^ PLATE 62.?Early Thule artifacts from the mound B test trench: a, antler sled arch; b, cut drift wood; c, d, whale rib mattocks. 174 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 ?&? m n 5 CM. PLATE 63.?Early Thule artifacts, level B-7: a, Punuk n (f)(x) harpoon head; b, Thule in open socket harpoon head; c, Sicco open socket harpoon head; d, Thule iv closed socket harpoon head; e, Thule n open socket harpoon head; /, harpoon head fragment; g, harpoon foreshaft; h, antler arrowhead; i, j , antler arrowheads; k, snow goggle fragment; I, n, atlatl fragments; m, fish arrow prong. PLATES 175 vv b 5 CM. PLATE 64.?Early Thule artifacts, level B-7: a-f, bola weights; g, small ivory wedge; h, men's knife blade; i, j , adz blades; k, drilled bear tooth. 176 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 < S . \ J&0E2, iPp*^ ? 5 CM. 2 I N. PLATE 65.?Early Thule artifacts, level B-7: a, bifacial preform; b-d, pebble end scrapers; e-g, i, pebble side scrapers; h, hafted flake knife. PLATES 177 5 CM. 2 I N. PLATE 66.?Early Thule artifacts, level B-7: a, c, d, bone awls; b, bone scraper; e, baleen object; /, pointed shaft with cord; g, polished stone; h, baleen knot; i-j, whetstones; k, wooden shaft. 178 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a hi PLATE 67.?Early Thule artifacts, level B-7: a-c, boat or sled fragments; d, ground slate hammerstone. PLATES 179 c d PLATE 68.?Early Thule artifacts, level B-6: a, harpoon foreshaft fragment; b, harpoon ice pick fragment; c-e, antler arrowheads; /, fish arrow prong; g, h, arrowshafts; i-k, bow fragments; /, harpoon shaft fragment. 180 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 ' % & * ? ? ? * , *?: <* -?wfauL~.: S C M . 2 I N. PLATE 69.?Early Thule artifacts, level B-6: a, bola weight; b, baleen and wood object; c, seal float mouthpiece; d, cup and pin game; e, kayak model; /, h, unidentified baleen objects; g, baleen whale effigy. PLATES 181 / PLATE 70.?Early Thule artifacts, level B-6: a, b, seal float bars (?); c, adz handle; d, ulu blade; e, baleen object; f, h, j , chert knife blades; g, awl handle; i, k, knife handles. 182 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 5 CM. 2 IN. PLATE 71.?Early Thule artifacts, level B-6: a, b, two-handed bone scrapers; c, unidentified wooden object; d-g, gauged drills; h, small pointed shaft; i, cut wood with shave spalls. PLATES 183 PLATE 72.?Early Thule artifacts, level B-6: a, flint flaker point; b, bone awl; c, small bucket bottom; d, ground and flaked discoid; e, pebble end scraper; /, ground stone. 184 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 i a b J I n 5 CM. 2 I N. PLATE 73.?Early Thule artifacts, level B-6: a, slate end blade; b, chert end blade; c, d, harpoon foreshaft fragments; e, f, antler arrowheads; g, antler arrowhead preform; h, arrowshaft; i, bola weight; ;', ulu blade; k, o, chert knife blades; /, grooved seal tooth; m, bead; n, slate knife blade; p, drilled antler; q, side scraper. PLATES 185 * f J y.'i-f' / 5 CM. 2 I N. m n PLATE 74.?Early Thule artifacts, level B-5: a, b, bone awls; c-f, gauged drills; g, h, unidentified wooden objects; i-k, kayak ribs; I, m, unidentified antler tools; n, small arrow or drill shaft. 186 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 \ 5 CM. 2 IN. PLATE 75.?Early Thule artifacts, level B-4: a, Nuwuk closed socket harpoon head; b, Thule n open socket harpoon head; c, d, antler arrowheads; e, arrowshaft; /, wound plug; g-i, knife handles; /, bola weight; k, drilled tooth; I, snow goggle fragment. PLATES 187 * | f J 5 CM. PLATE 76.?Early Thule artifacts, level B-4: a, b, ground slate nodules; c, ground stone; d, pot? sherd; e, retouched flake; /, small knife blade; g, ulu blade; h, unidentified pointed stick; i, gauge drill; /, hide peg. 188 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a f PLATE 77.?Early Thule artifacts, level B-3: a, harpoon head fragment; b, c, slate end blades; d, small foreshaft fragment; e, bird dart point; /, g, antler arrowheads; h, marlin spike; i, j , chert arrowheads; k, arrowshaft; l-n, baleen knots. PLATES 189 5 CM, 2 I N PLATE 78.?Early Thule artifacts, level B-3: a, slate knife blade; b, knife handle; c, notched bear tooth; d-k, small pointed wooden pegs. 190 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY a NUMBER 2 0 I 5 CM. 2 I N. PLATE 79.?Early Thule artifacts, level B-3: a, composite knife handle; b, drilled wooden object; c, stone drill bit; d, gauged drill; e, bone awl; /, adz blade; g, large quartzite retouched flake; h, scapula scraper. PLATES 191 H PLATE 80.?Early Thule artifacts, level B-2: a, Thule in open socket harpoon head; b, harpoon foreshaft; c, d, antler arrowheads; e, bird arrowhead; /, fish arrow prong; g, pointed shaft; h, adz handle; i, bone scraper; /, k, hide pegs; /, drilled wooden object. 192 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 9 5 CM PLATE 81.?Late Thule artifacts from the area A test trench: a, d, Thule nd harpoon heads; b, c, Thule nc harpoon heads; e, harpoon head preform; f, g, harpoon foreshafts; h, slate end blade; i, seal float mouthpiece stopper; ;', antler arrowhead preform. PLATES 193 V ?2 m b o , * "a * ? o rt C 03 ^ S e s S s - - - jO? ? -?-! < } / c * C o a a. u JZ E o !-4-1 ~ ffi g in CD X 194 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 20 PLATE 83.?Late Thule artifacts from the area A test trench: a, b, gauged drills; c, engraving tool; d, bone awl; e, needle case(?); /, marrow extractor or rib scraper(?); g, marlin spike(?); h, fire drill platform; i, unidentified wooden object. a , \ j i 5 CM. PLATE 84.?Late Thule artifacts, level B-1: a, whaling harpoon head; b, Nuwuk closed socket harpoon head; c, d, slate end blades; e, antler arrowhead; /, wound pin; g, whetstone; h, end scraper; i, carved wooden object; ;', drill spindle; k, unidentified wooden object; /, ulu handle. 196 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 V f . \ v J PLATE 85.?Late Thule artifacts, level A-1: a, harpoon head preform; b, c, slate harpoon end blades; d, harpoon ice pick fragment; e, antler arrowhead; /, g, antler arrowshafts; h, chert arrowhead; i, bow fragment; /, feather-cutting board; k, marrow extractor; I, bola weight; m, drilled tooth; n, atlatl hook; o, labret(?); p, small needle case; q, r, baleen cords. PLATES 197 1 - ' jd ffm ' U I fcSOOW- PLATE 86.?Late Thule artifacts, level A-1: ?, drill shaft; b, fishing pole; c, d, f, retouched flakes; e, pebble end scraper; g, ulu blade; h, bone scraper; z, whale rib mattock fragment. 198 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 / ? ft* J 5 CM. PLATE 87.?Late Thule artifacts, level A-2: a, e, Thule nc open socket harpoon heads; b, harpoon head fragment; c, d, f, g, slate end blades; h, harpoon foreshaft fragment; i, harpoon ice pick fragment; ;', wound pin; k, fish arrow prong. PLATES 199 a b d e f g h J m I n PLATE 88.?Late Thule artifacts, level A-2: a-n, antler arrowheads; o-r, antler arrowhead preforms. 200 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 i J I 5 CM. 2 I N. PLATE 89.?Late Thule artifacts, level A-2: a, wooden bird arrow; b, c, arrowshaft fragments; d-g, chert arrowheads; h, i, bola weights; /, marlin spike; k, feather-cutting board; I, harpoon dart head. PLATES 201 a 5 CM. PLATE 90.?Late Thule artifacts, level A-2: a, ulu; b, ulu blade; c, d, g, retouched flake knives; e, slate knife blade; /, notched pebble; h, i, whetstones;;', bifacial core. 202 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 I I ftj PLATE 91.?Late Thule artifacts, level A-2: a, paddleshaped wooden object; whale rib ice pick; c, snow shovel blade; d, e, marrow extractors or rib scrapers(P). PLATES 203 ' * ? ? ' * * , - * , * / m 5 CM. 2 I N. PLATE 92.?Late Thule artifacts, level A-2: a, antler wedge; b, amber pendant; c, j , potsherds; d, wooden bowl fragment; e, ivory awl; /, g, bone scrapers; h, pointed wooden shaft; i, baleen object; k, I, unidentified wooden squares; m, net float bar. 204 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 44 ? 5 CM. 2 I N. PLATE 93.?Late Thule artifacts, level A-3: a, Nuwuk closed socket .harpoon head; b, Thule nb harpoon head; c, harpoon head preform; d-f, slate end blades; g, wooden snow goggle frag? ment; h, i, harpoon shaft ice picks. PLATES 205 fi V 73 ' 1 f 9 m rt J 5 CM. 2 I N. PLATE 94.?Late Thule artifacts, level A-3: a-g, antler arrowheads; h-j, arrowshaft fragments; k, I, chert arrowheads; m, seal float mouthpiece stopper; n, labret. 206 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 m\ 1 JMIfc * ^ * W s < ? & \ ICT'-V 5 CM. 2 I N PLATE 95.?Late Thule artifacts, level A-3: a, ulu blade; b, d, chert knife blades; c, e, g, knife handles; /, slate knife blade; h, whetstone. PLATES 207 ? i i? >-^ajii IT iijrtm,!. y 2 I N. PLATE 96.?Late Thule artifacts, level A-3: a, hide peg; b, fire drill platform fragment; c, drilled ivory object; d-g, miscellaneous wooden objects; h, lump of coal; i, bone crusher; /, bone hammerhead. 208 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 J 5 C M . 2 I N. PLATE 97.?Late Thule artifacts, level A-4: a, Thule na harpoon head; b, harpoon head fragment; c, d, harpoon foreshaft fragments; e, harpoon ice pick fragment; /-/, antler arrowheads. PLATES 209 m.M 5 CM. PLATE 98.?Late Thule artifacts, level A-4: a, knife handle; b, man's knife. e f g h ?dk* i *&>-? F""Hr? . ' c ' ? - PLATE 100.?Late Thule artifacts, level A-4: a, unidentified handle; b, awl handle; c, baleen knife; d, knife handle; e, adz handle; /, whale rib ice pick. PLATES 211 ? v f &E&Qfif^ &&&Z& PLATE 101.?Late Thule artifacts, level A-5; a, Utkiavik closed socket harpoon head; 6, slate end blade; c, baleen snow goggles; d, harpoon head fragment; e, seal drag handle; /, g, harpoon ice pick fragments; h, harpoon shaft fragment. 212 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 m K J PLATE 102.?Late Thule artifacts, level A-5: a-e, antler arrowheads; /, antler arrowhead preform; g-i, arrowshaft fragments;;', k, chert arrowheads. PLATES 213 MM!;: y 5 CM. 2 I N. PLATE 103.?Late Thule artifacts, level A-5: a, fat cup; b, scraper; c, wood shaves bundle; d, fire drill spindle; e, f, bola weights; g, h, ulu blades; i, knife handle; /, snow shovel blade. 214 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 I '??-:?. .^s"-" ? > 3 i' UfA W PLATE 104.?Late Thule artifacts, level'A-5: a, &, large whale rib wedges; c-e, whale rib ice picks. PLATES 215 f 9 PLATE 105.?Late Thule artifacts, level A-5: a-d, hide pegs; e-g, bone scrapers; h, gaming piece; i, chert drill bit; j , k, toy bow fragments; I, kayak rib; m, n, unidentified wooden objects. PLATE 106.?Late Thule artifacts, level A-5: a, b, marrow extractors or rib scrapers (?); c, wooden shaft; d, sled shoe; e, sled shoe preform. 216 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a 'mj&mmiti&(t--'\ 5 CM. 2 IN . PLATE 107.?Late Thule artifacts, level A-5: a, grass caribou effigy; b-d, arrow effigies. 9 PLATES 217 y / 5 CM. 2 I N . PLATE 108.?Late Thule artifacts, level A-6: a, Nuwuk closed socket harpoon head; b, darting harpoon head; c-e, antler arrowheads; /, i-l, arrowshaft fragments; g, h, chert arrowheads. 218 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 PLATE 109.?Late Thule artifacts, level A-6: a-d, bola weights; e, bone weight; /, snow shovel blade. PLATE 110.?Late Thule artifacts, level A-6: a, b, ulu blades; c, slate knife blade; d, knife handle; e, chert knife blade; /, adz handle. PLATES 219 *'" f S* f-f *"-\f J 5 CM. 2 I N. PLATE 111.?Late Thule artifacts, level A-6: a-d, f, chert scrapers; e, pebble end scraper; g, bone scraper; h, two-piece ice scoop; i, bone scraper;;, k, scapula scrapers. 220 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 e f 5 CM. PLATE 112.?Late Thule artifacts, level A-6: a, unidentified baleen object; b, tied rib; c, toy bow fragment; d, unidentified ivory object; e, drilled bear tooth; /, unidentified wooden object; g, lead ore nodule; h, toy arrowshaft. PLATES 221 .*%* > . - . * , ? 9 h 5 CM. 2 I N. PLATE 113.?Late Thule artifacts, level A-6: a, c, d, bone awls; b, e, f, gauged drills; g, h, chert drill bits; i, cut antler. 222 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 ''V-> PLATE 114.?Late Thule artifacts, level A-6: a, ground slate nodule; c-f, miscellaneous stone artifacts; b, bucket bottom; g, ice scoop rim; h, wooden handle; i, bone maul. PLATES 223 t j 5 CM PLATE 115.?Late Thule artifacts, level A-7: a, harpoon ice pick; b, wooden wound pin; c, d, Leister barbs; e, antler arrowhead preform; /, g, antler arrowhead fragments; h, chert arrowhead; i, j , arrowshaft fragments. 224 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 a e J 5 CM. 5 CM. PLATE 116.?Late Thule artifacts, level A-7: a, b, stone net weights. PLATE 117.?Late Thule artifacts, level A-7: a-c, gauged drills; d, slate knife blade; e, f, drilled teeth; g, engraving tool; h, unidentified bone tool; i, j , knife handles PLATES 225 ^ r -"#~- ? 5 CM. 2 I N. PLATE 118.?Late Thule artifacts, level A-8: a, Thule II open socket harpoon head; b, slate end blade; c, d, antler arrowheads; e, arrowshaft fragment; /, side scraper; g, bone wedge; h, toy bow fragment(?); i, adz handle. 226 SMITHSONIAN CONTRIBUTIONS TO ANTHROPOLOGY NUMBER 2 0 1 If C b a f J \ 5 CM. PLATE 119.?Late Thule artifacts, level A-9: a, harpoon foreshaft; b, c, harpoon ice pick frag? ments; d, e, antler arrowheads; /, fire drill shaft; g, chert arrowhead; h, wooden arrowhead; i, antler arrowhead preform; j , ulu blade; k, scapula scraper.