"Arts, Faculty of"@en . "Non UBC"@en . "Anthropology, Department of"@en . "DSpace"@en . "Matson, R. G. (Richard Ghia), 1944-"@en . "Magne, Martin Paul Robert, 1954-"@en . "2018-02-08T19:29:57Z"@en . "2003-05"@en . "This document represents the first complete draft of Athapaskan Migrations: The Archaeology of Eagle Lake, British Columbia, published by University of Arizona Press in 2007. The Press required a substantial reduction in the size of the MS and about 30% of the text and about 50% of the figures were removed. Some of the principal arguments were abbreviated and fuller discussion can be found in this document. Other space reductions were in part mitigated by removing six sections and hosting them as on-line appendices at the University of British Columbia Library cIRcle and these can be seen at: http://hdl.handle.net/2429/39217.\r\n\r\nThis 2003 MS was the basis of Matson\u00E2\u0080\u0099s testimony as an expert witness in the Roger William v British Columbia case and cited numerous times in Judge Vickers\u00E2\u0080\u0099 decision (Tsilhqot\u00E2\u0080\u0099in Nation v British Columbia, 2007 BCSC 1700) granting the Tsilhqot\u00E2\u0080\u0099in land title for the rst time in Canada, which was upheld in a 8 to 0 decision by the Canadian Supreme Court in 2014 (Tsilhqot\u00E2\u0080\u0099in Nation v British Columbia, 2014 SCC44). The required reductions to the length of the manuscript meant that any discussion of the court case was not possible in the published version, although we made certain changes as a result of our participation in the case. Magne\u00E2\u0080\u0099s participation is essentially unacknowledged in the court proceedings because his position as a Federal civil servant meant that his formal involvement could only have been on the Government\u00E2\u0080\u0099s side, for which he declined to appear. Matson therefore spoke largely for Magne and himself in his testimony."@en . "https://circle.library.ubc.ca/rest/handle/2429/64600?expand=metadata"@en . "Athapaskans and Migrations: The Archaeology of Eagle Lake, British Columbia R.G. MATSON University of British Columbia MARTIN P.R. MAGNE Parks Canada, Calgary May,2003 This document represents the first complete draft of Athapaskan Migrations: The Archaeology of EagleLake, British Columbia, published by University of Arizona Press in 2007. The Press required asubstantial reduction in the size of the MS and about 30% of the text and about 50% of the figureswere removed. Some of the principal arguments were abbreviated and fuller discussion can be foundin this document. Other space reductions were in part mitigated by removing six sections and hostingthem as on-line appendices at the University of British Columbia Library cIRcle and these can be seen at: http://hdl.handle.net/2429/39217This 2003 MS was the basis of Matson\u00E2\u0080\u0099s testimony as an expert witness in the Roger William v BritishColumbia case and cited numerous times in Judge Vickers\u00E2\u0080\u0099 decision (Tsilhqot\u00E2\u0080\u0099in Nation v BritishColumbia, 2007 BCSC 1700) granting the Tsilhqot\u00E2\u0080\u0099in land title for the rst time in Canada, which wasupheld in a 8 to 0 decision by the Canadian Supreme Court in 2014 (Tsilhqot\u00E2\u0080\u0099in Nation v BritishColumbia, 2014 SCC44). The required reductions to the length of the manuscript meant that anydiscussion of the court case was not possible in the published version, although we made certainchanges as a result of our participation in the case. Magne\u00E2\u0080\u0099s participation is essentiallyunacknowledged in the court proceedings because his position as a Federal civil servant meant thathis formal involvement could only have been on the Government\u00E2\u0080\u0099s side, for which he declined toappear. Matson therefore spoke largely for Magne and himself in his testimony.R. G. Matsonrgmatson@shaw.caMartin Magnemagnem@rogers.comNovember, 2017ABSTRACT This report presents archaeological research undertaken in the region of Eagle Lake in the west-central Interior Plateau of British Columbia. The central problem is the archaeological recognition and dating of Athapaskan material culture, to shed light on the arrival of the Chilcotin to the region. This report describes the research problem and research methods, and also presents several analyses of ethnographic, lithic, faunal, wood, environmental and general archaeological cultural data. Major field investigations were undertaken in 1979 and 1983, although subsidiary ones took place in later years. In the dry lodgepole pine and grassland region around Eagle Lake and the upper Chilko River, randomly drawn 400 m X 400 m quadrats were used to locate and record archaeological sites and their environs in 1979 and 1983. This study addresses several aspects of both qua drat and site data in a statistical fashion and seeks patterns apparent in grassland and forested areas. Archaeological investigations in 1983 focused on one shallow rectangular lodge site (Bear Lake site or EkSa 36) with an associated complex of cultural features, and on two circular housepit sites, the Boyd (EkSa 32) and Shields (EkSa 13) sites. In 1979 one housepit and four lithic scatter areas were excavated. A considerable amount of project time in 1983 was devoted to ethnoarchaeological study of contemporary Chilcotin subsistence and settlement. The project has resulted in several methodological advances in ethnic group identification, settlement pattern analysis, lithic technological analysis, and dendrochronology. May 18, 2003-2 The research indicates that Chilcotin occupation of the Eagle Lake region included winter habitations as early as AD 1650. This prehistoric occupation period is termed the Eagle Lake Phase. The historic period is relatively late in this region, starting at about 1810. The term Lulua Phase is used to identify the historic Athapaskan occupation. Plateau Pithouse Tradition occupation of the region prior to the Athapaskan entry continues to about AD 1500. Analyses of projectile point and lithic assemblage variability in terms of ethnic origins are very successful in sorting Athapaskan from Plateau Pithouse Tradition (PPT) components, and provide support for previous research in the Mouth of the Chilcotin and Anahim Lake regions. Dendrochronological research was successful in building a 600 year living tree chronology for the region, and was able to show that the historic period winter lodge at the Bear Lake site had been built in AD 1877, as had an ancillary facility. Unfortunately no suitable samples were obtained to dendro-date the prehistoric component at this site, but three radiocarbon assays date this occupation relatively securely. The results of this research are put into the context of larger scale Athapaskan movements to the Pacific Coast and U.S. Southwest over the last 1200 years. May 18, 2003-3 ACKNOWLEDGEMENTS This project has been supported by grants from the Social Sciences and Humanities Research Council of Canada, the B.C. Heritage Trust, and the Hampton Research Committee (UBC). A great many people have been directly and indirectly involved and we shall attempt to properly acknowledge their contributions here. We owe a great deal to the persistence of our field crews. The 1979 field crew included Bill Armstrong, Bob Buchko, Mike Cook, Carrie Cooper, Terry Seidel, Janet Snell, James Tirrul-Jones, and Michael Quinn. Deanna Ludowicz was crew and lab supervisor that season. In 1983 it consisted of: Susan Cronkite, crew and field lab supervisor, and members of the UBC archaeological field school: Heather Black, Roxanne Hill, Heather Macleod, Tom Pulchny, Brad Smart, Wendy Unfreed, and Rudy Van den Broek. Also assisting during the 1983 season were Daphne Begg, Chris Hogarth, Quentin Mackie, Carli Nixon, and Robert Tyhurst. In 1984 Magne was assisted at Taseko Lakes bv Diana Alexander, Quentin Mackie, Brad Smart, and Robert ., Tyhurst. Also playing important roles during the two main Eagle Lake project seasons were Julian and Gordon Matson, who did their share of hauling water and firewood, never let bad weather dampen their enthusiasm for Spiderman, and who at early ages clearly explained the differences between numbers and variables. In 1985 Christophe Descantes, Gregory Mumford, Sheila Rowles, Philip Walker, and Dick Woo participated on the field school on Potato Mountain, where Matson was ably assisted by Diana Alexander. Tracy Barker filled the difficult position of cook, and we were resupplied by May 18, 2003-4 Mike King of Whitesaddle Air Services, and by Kevin, Alex, and Gerry Bracewell of the Bracewell Ranch. Susan Matson is due the utmost thanks for undertaking the unforgiving job of camp cook for the 1979 and 1983 seasons, for her editorial assistance and for her numerous illustrations. Other illustrations and photographs were prepared by the late Moira Irvine in her usual efficient manner. Elli Carrie searched for and found a otolith in the Bear L ake site materials, and Dorothy Godfrey-Smith provided source area analyses of obsidian samples. Artifacts from the Anahim Lake region were lent to us courtesy of the National Museum of Man in Ottawa, and those from ElRn 3 and FaRn 3 were courtesy of the British Columbia Provincial Museum. Many thanks to Anne Morrison and Tom Loy, respectively. Knut Fladmark allowed us to analyze projectile points from Punchaw Lake, David Pokotylo let us analyze some from Upper Hat Creek, as well as assisting us in other ways, and Arnoud Stryd lent collections from Lillooet. The assistance of our two Chilcotin workers was invaluable; Diane Lulua and Gilbert Solomon contributed greatly to the research through their diligence and knowledge of the local area. Several Chilcotin people assisted us in many ways; we are grateful to Henry Lulua, Nora Lulua, Edmond Lulua, Peter Lulua, Scotty Lulua, Doris Lulua, Kasimir Lulua, Mabel Solomon, and Henry Solomon for their tolerance of our presence and for their provision of much information. Many thanks to Marvin Baptiste, Adam Williams and Benny Williams, past chiefs of the Nemaiah Valley Indian Band (now known as the \"Xeni Gwet'in First Nations\"), for supporting this research. George Colgate, then band manager of Nemaiah, was helpful during his tenure. Many thanks May 18, 2003-5 to Alan Haig-Brown for smoothing some paths for us. Several local folks helped us out, exchanged information, and provided companionship, especially, the late Doug Boyd, and the late Scotty Shields who generously shared their knowledge of the Chilcotin, and Ted and Cindy Abbott. Also due thanks are David Aberle, the late CE. Borden, Paul Donahue, Knut Fladmark, Michael Kew, Arnaud Stryd, and the late Roscoe Wilmeth for their advice, encouragement, and related journeys into Plateau prehistory. Braxton Alfred and Neil Guppy provided some remedies for computer gremlins. Magne would also like to thank Orysia Luchak for allowing him occasional time to devote to the Eagle Lake Project, Michael Klassen and Darro Stinson for including him in the 2000 work at EjSa 11, and particularly Judy Trumble and Liam Woodroffe for their patience. All artifacts recovered from the Eagle Lake region are curated at the Laboratory of Archaeology, Department of Anthropology and Sociology, University of British Coiumbia, Vancouver. May 18, 2003-6 TABLE OF CONTENTS ABSTRACT ACKNOWLEDGEMENTS LIST OF T ABLES LIST OF FIGURES I INTRODUCTION (R.G. Matson) --------------------------------------------------9 Problem Project History Natural Environment of the Eagle Lake Region (Martin Magne and Deanna L ud owicz )------------------------18II SETTLEMENT PA TTERNS---------------------------------------------------------26 Plateau Pithouse Tradition Mouth of the Chilcotin PPT Settlement Patterns Chil co tin Tradition----- --------------------------------------------------------40 Ethnoarchaeological Investigations in the Chilcotin Region (Linda Burnard-Hogarth) ---------------------------------55 Notes on Cambium-Stripped Lodgepole Pine (Martin Magne )----------------------------------------------81 Chilcotin Settlement Patterns ---------------------------------------86 Distinctive A thapaskan Artifacts--------------------------90 The Parallel Direct Historic Approach (R. G. Matson)------------------------------------------------ 94 III REGIONAL SURVEYS Quadrat Survey------------------------------------------------------------------105 Research Design Chilko River Survey Design--------------------------------------------------122 Quadrat Survey Results Mouth of the Chilcotin------------------------------------------------125 Eagle Lake Quadrat Survey-----------------------------------------135 Taseko Lakes Quadrat Survey--------------------------------------166 Chilko River Survey --------------------------------------------------170 Potato Mountain Survey---------------------------------------------186 Summary--------------------------------------------------------------------------196 DESCRIPTIVE ARCHAEOLOGY IV Archaeological Site Excavations in the Eagle Lake Region --------------203 R.G. Matson and Martin Magne The Bear Lake Si te---------------------------------------------------------------204 Excavations at Plateau Pithouse Tradition Sites; May 18, 2003-7 The Boyd and Shields Sites--------------------------------------------221 V ARTIFACTS, DATING, AND FAUN AL REMAINS Artifact Descriptions (Patricia Ormerod )-----------------------------------227 Radiocarbon and Dendrochronological Dating---------------------------278 Obsidian Source Analysis -----------------------------------------------------302 Faunal Analysis (Linda Roberts and Martin Magne) -------------------304 VI ETHNIC IDENTIFICATION OF MATERIAL CULTURE Martin Magne and R.G. Matson----------------------------335 Projectile Point Analyses ----------------------------------------------------336 Lithic Assemblage Analysis ------------------------------------------------361 Summary of Ethnicity Analyses--------------------------------------------382 VII ATHAP ASKAN MIGRATIONS; The view from Eagle Lake------------------------------------385VIII CON CL U SI ON S --------------------------------------------------------------------442REFERENCES CITED---------------------------------------------------------------------454 APPENDIX I: TREE-RING DATING OF SAMPLES FROM EAGLE LAKE Marion Parker APPENDIX II: ANALYSIS OF FLOTATION SAMPLES FROM EAGLE LAKE Elizabeth Radomski APPENDIX III: FIELD RECORDING FORMS AND SELECTED DAT A T ABLES FOR EAGLE LAKE AND T ASEKO Cultural Form Physiographic Form Botanical Form Table 1 (MOC IX) MOC Cultural Data from Quadrats. Table 2 (MOC XXIII) MOC Cultural Data from Excavations. Table 3 (84-5a,5b) Eagle Lake Botanical and Physiographic codes. Table 4 (84-6) Eagle Lake and Taseko Quadrat Environmental Data. Table 5 (84-7) Eagle Lake and Taseko Site Environmental Data. Table 6 (84-8) Eagle Lake and Taseko Site Cultural Data. May 18, 2003-8 Chapter I 1-1 (84-1)I-2 (New)1-3 (84-2)1-4 (New)1-5 (New)1-6 (New)1-7 (New)1-8 (New)Chapter II II-1 (New)II-2 (84-12)II-3 (84-3)Il-4 (New)II-5 (84-4)II-6 (84-5)II-7 (84-6)II-8 (84-7)Il-9 (84-8)II-10 (84-9)II-11 (84-10)II-12 (84-11)II-13 (84-13)Chapter III Location of Eagle Lake and other regions cited .................. 10 View of Niut and Potato Ranges from Eagle Lake ............. 19 The 1979 and 1983 Eagle Lake Sampling Frame ................. 19 Cariboo-Aspen-Lodgepole Pine-Douglas Fir ...................... 21 Subalpine Englemann Spruce-Subalpine Fir, Taseko ..... .. 22 Alpine Zone \"Parkland\", Potato Mountain .......................... 23 Chilko River, near Lingfield Ck. ........................................... 26 West End of Eagle Lake from Potato Mountain .................. 26 Teit' s rendition of the Plateau Pithouse ................................. 30 Aboriginal Groups of Interior British Columbia ................. .42 Puntzi Lake Chilcotin Lodge (1951 ) ...................................... .44 Chilcotin Territory (1850) and Reserve Locations ............... 51 Chilcotin Drying Racks near Henry's Crossing .................... 62 Chilcotin Hide Stretching Frame ............................................. 66 Chilcotin Sweat Lodge frame ................................................... 68 Chilcotin Smudge Pit. ................................................................ 70 Drying Rack and Hearth near Henry's Crossing .......... ,. ..... .72 Cambium-Stripped Lodgepole Pine, 1983 .............................. 82 Close-up of Cut Marks on Cambium-stripped Tree ............ 82 Histogram of Orientation of Stripped areas ........................... 83 Parallel Direct Historic Approach ............................................ 98 III-1 (84-17) Eagle Lake Quadrats Surveyed in 1979 and 1983 .................. 113 III-2 (MOC-5) Mouth of Chilcotin Sampling Scheme .................................... 114 1II-3 (A&M-3) Potato Mountain Survey Environs and Eagle Lake ............. 115 III-4 (A&M-6) Potato Mountains Sampling Scheme ...................................... 115 111-5 (New) Taseko Lake Survey Areas ......................................................... 118 111-6 (New) Location of Taseko Lake Quadrats ........................................... 119 III-7 (New) Taseko Lake Quadrat Sketch Maps ......................................... 119 III-8 (MOC-8) MOC Grassland Quadrats 1-6 ................................................ 126 III-9 (MOC-9) MOC Grassland Quadrats 7-12 ............................................... 126 III-10 (MOC-6) MOC Grassland Locations on aerial Photograph ............... .126 111-11 (MOC-10) MOC Forested Quadrat sketch maps .................................. 126 III-12 (MOC-7) MOC Forested Quadrats locations on aerial Photograph .. 127111-13 (MOC-12) Multidimensional Scaling of MOC Quadrats ..................... 128 IIl-14 (MOC-34) Contour Map of EkRo 48 ....................................................... 132 III-15 (MOC-35) EkRo 48 Excavation Floor Plan ............................................. 132 III-16 (MOC-37) Contour Map of EkRo 18 ....................................................... 133 1II-17 (MOC-66) MOC observed Settlement Pattern ....................................... 134 May 21, 2003 IIl-18 (New) Eagle Lake Quadrat Sketch Maps ............................................ .137+ (a-t) IIl-19 (84-17) Location of Eagle Lake Quadrats ............................................... 137 IIl-20 (new) Box-and-Dot plot comparisons of Forest and Grassland ......... .139 IIl-21 (84-18) Sites per Eagle Lake Quadrats ..................................................... 142 111-22 (79-5-1) Pit Features per Quadrat by Distance from River ................... 143 IIl-23 (79-5-2) Pit Features per Quadrat by distance from Lake ..................... 143 IIl-24 (84-19) Multidimensional Scaling of Quadrats by Environment.. ....... 159 111-25 (84-20) MOS of Quadrat Sites by Environment.. ................................... .161 III-26 (84-21) MOS of Sites by Cultural Attributes ........................................... 163 111-27 (79-4-3) Map of EkSa 5, the Canoe Crossing site ................................... 176 111-28 (79-4-4) Location of ElRw 9, the Quiggly Holes site .............................. 178 111-29 (79-4-5) Map of ElRw 9, Quiggly Holes ................................................... 178 IIl-30 (79-4-6) Location of CR 64, 73, and 92, (EkSa 34, 35, and 33) ................ 179 111-31 (A&M 19) Map of CR 92 (EkSa 33), the Brittany Creek site .................. 179 111-32 (79-4-9) Map of CR 64 (EkSa 34) ............................................................... 181 111-33 (A&M 27) Map of CR 73 (EkSa 35) ............................................................ 182 111-34 (A&M-21) Map of Fish Trap Lake Site (EkSb 27; 84-27) ....................... .183 111-35 (A&M 14) Potato Mountain Lithic Scatters .......................................... ... 187 111-36 (A&M 25) Tested sites on Potato Mountain ........................................ .... 188 III-37 (A&M 15) North End of Potato Mountain .......................................... ..... 188 lII-38 (A&M 23) The Mountain Fan Site, P8-3 (EjSb 39) .................................... 189 IIl-39 (A&M 24) The Mountain Pond Site, P8-1 (EjSb 54) ................................. 190 111-40 (A&M 22) The Middle Mountain Site (EjSb 52) ....................................... 192 111-41 (A&M 20) Potato Mountain Artifacts ........................................................ 192 111-42 (A&M 26) Profile of roasting pit at P2-9 (EjSb 33) ................................... 194 III-43 (M&A 8) Roasting Pit Diameters, Probabilistic surveys ........................ 198 Ill-44 (M&A 7) Cache Pit Diameters, Probabilistic Surveys ............................. 199 Chapter IV Figure Heading Pai;e IV-1 (84-22) Bear Lake Site (EkSa 36) ................................................................. 205 IV-2 (84-23)Bear Lake Site, Features A, H, and B ............................................. 205 IV-3 (84-24)Lithic Scatter, Bear Lake Site .......................................................... 207 IV-4 (84-25)Feature A, Bear Lake Site ................................................................ 207 IV-5 (84-26)Plan View of Feature D ................................................................... 208 IV-6 (84-27)Profile of Feature D .......................................................................... 208 IV-7 (84-28)Excavation of Feature H ................................................................. 208 IV-8 (84-29)Trench Through Feature B, Facing North ..................................... 209 IV-9 (84-30)Two Post Holes, Facing East.. ......................................................... 209 IV-10 (84-31)Profile of Feature G Hearth ............................................................. 209 IV-11 (84-32)Feature J, Cache Pit, Facing North ................................................. 210 IV-12 (84-33)West Side of Feature B. .................................................................... 210 IV-13 (84-34)Plan of Feature B Excavations ......................................................... 212 IV-14 (84-35)North-South Feature B Profile ........................................................ 212 IV-15 (84-36)East-West Feature B Profile ............................................................. 212 May 21, 2003 IV-16 (84-37)North Wall Profile Feature J Cachepit.. ........................................ 212 IV-17 (84-18)Oistribution of Lithics and Historic Artifacts .............................. 213 IV-18 (new) Boundaries of Feature B and Feature I Lodges ........................... 215 IV-19 (84-44)Contour Map of Boyd Site (EkSa 32) ............................................ 222 IV-20 (84-45)Profile of Unit 1, Boyd Site ............................................................. 222 IV-21 (84-39)Contour Map of Shields Site (EkSa 13) ......................................... 224 IV-22 (84-40)Profile of Units 9 & 10, Housepit !, Shields Site .......................... 224 IV-23 (84-41)Excavations of Housepit 5, Shields Site ........................................ 224 IV-24 (84-42)West Wall Profile of Unit 3, Shields Site ....................................... 225 IV-25 (84-43)West Wall Profile of Unit 6, Shields Site ....................................... 225 Chapter V V-1 (84-81)Bone Artifacts ..................................................................................... 268 V-2 (84-82) Relationship between Radiocarbon and Calendric years ........... 278 V-3 (New) Bear Lake Prehistoric Radiocarbon Dates ...................................... 281 V-4 (New) Averaged Radiocarbon Date, by Radiocarbon andCalendric Years ...................................................................... 282 V-5 (New) Distribution of Eagle Lake and Potato Mountain dates ............... 295 V-6 (New) Alternative Eagle Lake Regional Sequences .................................. 300 V-7 (84-86)Bear Lake Otolith ................................................................................ 312 V-8 (84-87)Adams River Sockeye Otolith ........................................................... 312 V-9 (New) Comparison of Historic and Prehistoric Bear Lake Fauna ........... 319 Chapter VI Vl-1 (84-88)Eagle Lake Side-Notched Points ...................................................... 337 Vl-2 (84-1) Locations of B.C. sites and Projectile Point Sources ...................... 337 VI-3 (84-89) MOC Kamloops Points ..................................................................... 338 VI-4 (84-90) Chinlac Side-Notched Points ........................................................... 338 VI-5 (84-91) Potlatch Site Points ............................................................................ 338 VI-6 (84-92) Projectile Point Attributes ................................................................ 343 VI-7 (new) MOS of Projectile Points .................................................................... 353 Vl-8 (new) MOS of Projectile Points, Photographs ........................................... 354 VI-9 (new) MDS of Sites, First 2 Dimensions ..................................................... 366 Vl-10 (new) MOS of Sites, First 3 Dimensions ..................................................... 367 Vl-11 (new) 25 Assemblages by Middle Stage & Shatter Debitage .................. 377 VI-12 (new) \"House\" Proportions of Debitage ..................................................... 378 Vl-13 (new) Tool and Debitage Patterns, Houses ............................................... 378 Vl-14 (new) Artifact Profiles of 5 Eagle Lake Components ............................... 380 Chapter VII Athapaskan Migrations; The view from Eagle Lake. VII-1 (New) Distribution of Athapaskan speakers ............................................. 387 VII-2 Mooney's and Ac ko mok ki's Maps ............................................................ 401 VII-3 Northern Athapaskan Archaeological Sites ............................................... 405 VII-4 Kavik Points from Klo-kut, Atigun, and Dixthada Sites ......................... .405 VII-5 Kavik Points from Chinlac, Ulgatcho, and Potlatch sites .................... \" ... 407 VII-6 \"Athapaskan\" Side-Notched Points (from Chinlac) ................................. .410 May 21, 2003 VII-7 (27-6) Changes in Anasazi occupation in PIII an PIV periods ............... .424 VII-8 (27-7) Rocky Mountain route for Athapaskans ........................................ .425 VII-9 White River Ash Fall area (circa A.D. 800) ................................................ .432 VII-10 Athapaskan migration routes ...................................................................... .438 Chapter VIII Conclusions VIII-1 The Bear Lake site Eagle Lake Component.. ................................ .446 VIII-2 The Boyd Site P.P.T. Component.. ................................................. .447 VIII-3 The Shields Site P.P.T. Component.. .............................................. .447 VIIl-4 The CR 73 Pithouse Component.. ................................................... .449 VIII-5 Summary of Ethnicity and Parallel Sequences ............................. .450 Table Chapter I (None) Chapter II Heading 11-1 Chilcotin Drying Racks (After Table 5.1, Burnard 1987) ...................... 63 11-2 Hearth Characteristics (after Table 5.2, Burnard 1987) ......................... 71 11-3. Contents of a typical abandoned salmon fishing camp ......................... 75 Chapter III 111-1 (MOC III) MOC Grassland Quadrat Summary ...................................... 125 111-2 (MOC II) MOC Quadrat Environments ................................................... 125 III-3 (MOC IV) MOC Forested Quadrat Summary ......................................... 126 III-4 (MOC XI) Final MOC Synthetic site Classes ........................................... 129 III-5 (84-2+) Eagle Lake Quad rat Features ....................................................... 136-7 III-6 (84-3) Eagle Lake Quadrat Mann-Whitney U tests ................................ 139 Ill-7 (84-4) Eagle Lake Quadrat Interquartile Ranges .................................... 140 III-8 (84-8) Eagle Lake Quadrat site Summaries ............................................. 145-6 III-9 (79-4-2) Upper Hat Creek Quadrat Summary ......................................... 146 111-10 (79-4-3) Eagle Lake Open Quadrats; Compare with Hat Creek ........ 146 III-11 (79-4-4) Mouth of Chilcotin Summary ............................................ ........ 149 111-12 (79-4-5) Eagle Lake Open Quadrats near Chilko River ........................ .149 111-13 (MM84-1) Taseko Lake Quadrat and Quadrat site surnmaries ............ 167 111-14 (79-4-6) Chilko River Survey Summary ................................................... 172 111-15 (79-4-8) CR 92, EkSa 33 and Quadrat 19-1 Summaries .......................... 181 III-16 (A&M-1) Potato Mountain Quad rat Summary ...................................... .186 Chapter IV (None) Chapter V V-1 Field Assigned Site Numbers and Borden Designations .................... 229 V-2 Side-Notched Points: Metric Data .......................................................... 231 V-3 Single-side Notched Points: Metric Data ...................................... ......... 232 V-4 Large Side-notched Points: Metric Data ................................................ 233 V-5 Multi-notched Points: Metric Data ................................................ ......... 234 V-6 Kavik Points: Metric Data ....................................................................... 235 V-7 Corner-notched Points: Metric Data ...................................................... 236 V-8 Stemmed Points: Metric Data .................................................................. 237 V-9 Miscellaneous Points: Metric Data ......................................................... 238 V-10 Large Formed Bifaces (tci-tho-like): Metric Oata ......................... ......... 240 V-11 Complete Small Formed Bifaces All Forms: Metric Data ................... 241 V-12 Continuous Scrapers: Metric Data ......................................................... 245 V-13 End Scrapers: Metric Data ........................................................................ 246 V-14 Large Bifacially Retouched Flakes: Metric Data ......................... .......... 247 V-15 Small Bifacially Retouched Flakes: Metric Oata ................................... 248 May 21, 2003 V-16 Utilized Flakes: Metric Data ........................................................... H \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 249 V-17 Mutliple-edged Unifaces: Metric Data ................................................... 251 V-18 Sinuous, Multiple-edged Unifaces: Metric Data by Site ..................... 252 V-19 Gravers: Metric Data ................................................................................. 253 V-20 Alternating Perforators: Metric Data ...................................................... 254 V-21 Non-Alternating Perforators: Metric Data ............................................. 255 V-22 Pieces Esquillees: Metric Data .................................................................. 258 V-23 Bipolar Wedges: Metric Data ................................................................... 259 V-24 Bipolar Cores and Fragrnents ................................................................... 260 V -25 Cortex Spall Tools ...................................................................................... 261 V-26 Core Tools .................................................................................................... 262 V-27 Hammerstones: Metric Data ..................................................................... 263 V-28 Cobble Tools ................................................................................................ 264 V-29 Large Flake Tools: Metric Data ................................................................. 266 V-30 Microblades: Metric Data ........................................................................... 267 V-31 Bone Points: Metric Data ............................................................................ 269 V-32 Harpoon points and Fragments: Metric Data ......................................... 270 V -33 Waisted Stones: Metric Data ...................................................................... 272 V-34 Adze Flakes: Metric Data ........................................................................... 273 V-35 Polished Pebbles: Metric Data ................................................................... 273 V-36 (84-10) Radiocarbon Dates, Major Excavations ...................................... 280 V-37 (84-11) Dendrochronological Dates .......................................................... 289 V-38 (84-13) Radiocarbon Dates, other Eagle Lake Sites ................................. 292 V-39 (A&M 13) Potato Mountain Radiocarbon Dates ..................................... 292 V-40 (MOC+) Mouth of Chilcotin, Taseko Lates Radiocarbon Dates .......... 297 V-41 Bear Lake, Shields, and Boyd Site Fauna ................................................. 307 V-42 Bear Lake Fauna ........................................................................................... 310 V-43 Shields Site Fauna ........................................................................................ 322 V-44 Boyd Site Fauna ............................................................................................ 328 Chapter VI Vl-1 Projectile Point Attributes .......................................................................... 341-2 VI-2 Projectile Point Attribute Definitions ....................................................... 344 VI-3 Stepwise MDA, Regional Homogeniety .................................................. 349 VI-4 Stepwise MDA, Ethnic Homogeniety ...................................................... 350 VI-5 Direct MDA, Regional Homogeniety ....................................................... 351 VI-6 Direct MDA, Ethnic Homogenierty .......................................................... 352 VI-7 Summary of Projectile Point Results ........................................................ 357 VI-8 Assemblage Data, 13 Components ........................................................... 364 VI-9 MDA Site Analysis ...................................................................................... 368-9 Vl-10 MDA Site Analysis, using percentages .................................................... 369-70 VI-11 Debitage Data ............................................................................................... 372 VI-12 MDA, Debitage Analysis ............................................................................ 376-7 Vl-13 Artifact Tabulations of 5 Eagle Lake components .................................. 379 May 21, 2003 Chapter VII (None) Chapter VHI (None) May 21, 2003 Plate l Plate 2 Plate 3 Plate 4 Plate 5 Plate 6 Plate 7 Plate 8 Plate 9 Plate 10 Plate 11 Plate 12 Plate 13 Plate 14 Plate 15 Plate 16 Plate 17 Plate 18 Plate 19 Plate 20 Plate 21 Plate 22 Plate 23 Plate 24 Artifact Plates (All in Chapter V) Title \u00EF\u00BF\u00BD Side-notched and Multi-notched Points .............................................. 232 Kavik, Corner-notched, Stemmed and Miscellaneous Points ............ .235 Point Fragments ................................................................................... 238 Large Formed Bifaces (LFBl) .............................................................. 239 Large Formed Biface Fragments and Small Oval Formed Bifaces .... 242 Small Formed Bifaces (SFBl) ............................................................. 242 Formed Scrapers (FOSC) .................................................................... 244 Large Bifacially Retouched Flakes (BIREl) ....................................... 247 Small Bifacially Retouched Flakes (BIREs) ...................................... 248 Unifacially Retouched Flakes ............................................................ 249 Utilized Flakes (UTIL) ..................................................................... 250 Multiple-edged Unifaces (MUUT) .................................................... 25 I Sinuous, Multiple-edged Unifaces (SINU) ....................................... 252 Gravers, Alternating Perforators, Non-alternating Peii\u00C2\u00B7orators ........ 253 Piece Esquillee (PEEQ) .................................................................. 256 Bipolar Wedges and Cores .............................................................. 258 Cortex Spall Tools (SPTO) .............................................................. 261 Hammerstones and Cobble Tools ................................................... 262 Large Flake Tools (LFLT) ............................................................... 265 Microblade and Ground Stone Assemblages .................................. 266 Incised and Decorated Bone (INCB) ............................................... 267 Bone Fish Spear Points (Leister Prongs) (BPNT) ........................... 269 Bone and Antler: Harpoon Points, Awls, and Beamers ................... 270 Historic Artifact Assemblage (HIST) .............................................. 274 Chapter I INIRODUCTION R.G. Matson The Problem This report is primarily a synthesis of archaeological field research undertaken at Big Eagle or Choelquoit Lake, British Columbia mainly during 1979 and 1983. It provides a discussion of the basic research problem, the methodology used, descriptions of sites and artifacts, and analyses of environmental and cultural variables, as well as the resulting cultural history. The objective of the Eagle Lake Archaeological Project is the archaeological identification and description of the Athapaskan migration into the Eagle Lake or Choelquoit Lake area (Figure I-1). As an aside, the current official name \"Choelquoit\" is an anglicized version of the Chilcotin word for eagle. In the past this lake was generally known as Eagle Lake (Lane 1953:36), particularly among the Chilcotin. Apparently when map makers were faced with two lakes locally called \"Eagle\" they used an anglicized version of the Chilcotin word for eagle for one and called the other \"Eagle'' (25 km -15 miles- northwest of Choelquoit lake). Choelquoit is also locally called \"Big Eagle Lake\" to discriminate it from the \"other\" Eagle lake. However, it is Choelquoit lake that is known as Eagle Lake in history. Locally famous figures such as \"Eagle Lake Henry\" lived at Choelquoit Lake. So we continued to call it the appropriate anthropological English name, \"Eagle Lake.\" FIGURE May 18, 2003-9 Vancouver ISl!ll'ld J_-1O !:------80 km '------ \"50 miles British Columbia Figure 1-1. Location of Eagle Lake, Mouth of the Chilcotin, and other regions cited in text. The archaeological investigation of prehistoric migrations is an undertaking basic to the field and is related to interests in the study of culture change. When significant change through time is apparent in artifact assemblages, the common means of explanation of such change is by reference to external events such as environmental variation or migration of new cultures. There has been much dissatisfaction with such explanations in the past, since they were often ad hoc or offered for lack of other evidence (Adams 1968). As a result, explanations that refer to internal cultural events or processes became more commonplace beginning in the 1960s (Adams et al. 1978). Migration, however, is still considered to have been a major reason for change in the archaeological record, though our methods for recognizing it are in considerable need of improvement. As such, in recent years there has been greater acknowledgement of both of the importance of migration as a mechanism of cultural change in the past and in the critical need to be able to recognize prehistoric migration ( Anthony 1990, 1997; Burmeister 2000; Cameron 1995; Matson and Magne 2001; Towner 1996). At the time the Eagle Lake investigation was proposed (Matson 1978) migration was definitely not a major interest for most North American archaeologists. Yet migrations did occur, although a clear methodology for determining them did not exist, with the exception of the \"site unit intrusion\" (Thompson 1958; Trigger 1968). Even in 1986 Rouse does not clearly provide any clear methodology, except for that of a general hypothesis-testing nature in his book-length treatment about three iikely migrations May 18, 2003-10 (Rouse 1986). Archaeology was thus in a bind, having rejected previous inferences of migration on methodological grounds, but yet recognizing the importance of migration to understand the past, at least in certain clear cases. The migration of Athapaskan speaking ancestors of the Navajo and Apache people from Canada around 1000 years ago is one such well known migratory \"event\" (Gunnerson 1969, 1979; Towner 1996). At around that point in time people with very distinct language and ethnic background started a movement that spanned almost half of a continent. The precise nature of the migration is still obscure, (Wilcox 1981, 1988; Magne 2001; Towner 1996) and we do not know whether it consisted of a single large wave, several smaller ones, whether groups split off along the way, or in what ways it influenced the groups through which the Athapaskans moved, although we later give our evaluation of the most likely process and route. The Chilcotin (who currently prefer the spelling \"Tsi'lhqot'in\") are the southernmost Athapaskans in Canada, located in the south-central Interior Plateau of British Columbia (Farrand 1900; Teit 1909; Lane 1953, 1981). At one time another Athapaskan group, the Nicola, inhabited an area to the south of the Chilcotin, but they were extinct by 1850. The ethnographic record and contemporary informants indicate that Chilcotin are relatively recent inhabitants of their territory (Teit: 1909b; Lane 1953,1981; Tyhurst 1984). Thus there is reason to believe that any knowledge gained about their arrival in the region would benefit the issue of the more extensive migration further south as well as the question of investigating and determining migrations in general. A specific procedure, denoted the \"Parallel Direct Historic Approach\" (Matson 1982, 1991) (described beiow) was developed to approach the migration problem. May 18, 2003-11 Other archaeologists were also interested in the same question, and two in particular, Wilmeth (1978) and Donahue (1977) reported on their investigations further to the north and northwest. In summary, the basic problem was the identification of migrations in the archaeological record and the specific question of identifying when the Chilcotin arrived in the Eagle Lake area. Project History The history of the project is somewhat bizarre, really first appearing as a spark in the campground of the Museum of Northern Arizona in Flagstaff, during the summer of 1971 when Bill Lipe introduced me to David Aberle. I was informed that the University of British Columbia was looking for an archaeologist and that Prof. Aberle hoped that whoever was hired would do some Athapaskan archaeology. I eventually was hired by UBC in 1972 and I considered that one of my obligations was to investigate Athapaskan prehistory. My first field project in the interior of British Columbia was at the Mouth of the Chilcotin River, (Figure I-1) where I tested out settlement pattern methodology developed in the Great Basin (Thomas 1969; Matson 1971) and U.S. Southwest (Lipe and Matson 1971; Matson and Lipe 1975). The Mouth of the Chilcotin was clearly part of the Plateau Pithouse Tradition (Richards and Rousseau 1987), being part of the Alkali Lake band (currently ''Esketmc First Nation\") territory, or what Teit (1909a) referred to as the Canyon Shuswap. I carried out the field work there in 1974 (Harn 1975; Matson et al. 1984) where I surveyed 18 400m by 400 m quadrats and carried out some minor excavations. This area had been previously been the subject of an interesting, but uncompleted, project funded by the Opportunities for May 18, 2003-12 Youth program, directed by Paul Sneed, Grant Keddie and G. Jones (Keddie 1972). If the Chilcotin (people) had replaced the previous Plateau Pithouse Tradition inhabitants in the Chilcotin (geographical area), the Mouth of the Chikotin (MOC) artifactual and settlement pattern material could serve as a model for the expected \"pre\u00C2\u00ADAthapaskan\" occupation. An investigation within Chilcotin territory with similar environmental constraints was next required to compare with the MOC. The two most important factors, were the presence of salmon, the most important resource for the Plateau Pithouse Tradition, and a relatively open understory, so that settlement pattern investigations would be similar to those (and relatively inexpensive) at the MOC. The salmon resources in the Chilcotin are concentrated along the Chilcotin River, up to the branch with the Chilko River, when the Chilko River up to Chilko Lake carries the majority of the salmon. So, in 1978 I looked for similar environments to the Mouth of the Chilcotin along the Chilcotin and Chilko Rivers upstream from Alexis Creek, which was the traditional eastern boundary of Chilcotin Indian territory. I was guided by a complete set of provincial site records and two short publications (Mitchell 1970a, 1970b), all that existed on Chilcotin archaeology, outside of Wilmeth' s (1969, 1971) Anahim Lake's investigations. In this tour, the Eagle Lake area, adjacent to the Chilko River, with several previous known Plateau Pithouse Tradition pithouse village sites, and a number of relatively open areas, was the obvious choice. I then wrote a grant proposal to the Social Science and Humanities Research Council of Canada (Matson 1978) to support this research. Three separate research goals were proposed. First, a regional sampling project to discover the range of sites and to compare with the MOC. Second, excavation of sites May 18, 2003-13 that on the basis of surface investigations appeared to be Athapaskan and others that appeared to be Plateau Pithouse Tradition. The second part was to be carried out in future seasons, with separate proposals to be funded. The third goal was to survey a long segment of the Chilko River, mainly to find good candidates for sites to fill out the e arlier portions of the Plateau Culture History, and to complete the settlement pattern a pproach. It was clear, even at that time, that large sites were concentrated along the rivers. Although we located, mapped, and documented over 100 sites along the Chilko River, including testing a number of them, not a single one exhibited the sought for stratification and culture history attributes, so that portion of the investigations was curtailed, although several of the sites did have relevance to our main goals. It is surprising that twenty years after this investigation, we have still have serious problems understanding the time period before the Plateau Pithouse Tradition (PPT) ( Stryd and Rousseau 1996). After the 1979 field season I applied for funding for 1980, and was informed contrary to what I had been told in 1978, that I should have applied for a multi-year project and did not receive the requested funding. I re-applied in 1982, after making available a relatively complete report (Matson et al. 1980) on the 1979 research which served as one of the bases for Magne and Matson (1984) and the current monograph. This proposal was funded and completed the basic Eagle Lake field research. The regional sampling techniques that were used to locate sites consumed the greater part of the 1979 season and about one-fifth of the 1983 season. The objective of the regional sampling program is to be able to sketch the regional settlement pattern of the area, which may or may not vary with respect to ethnic affinities of the participants. May 18, 2003-14 Excavation of selected sites was a minor part of the 1979 season (Matson et al. 1980), but was the major focus of the 1983 season when three sites and a total of 76 1m X Im units were excavated. It was in excavations that we expected to obtain the major evidence of material culture differences between ethnic groups, as well as firm evidence of the nature and timing of the Chilcotin migration into the Eagle Lake area. In the random sampling process, about 8% of the Eagle Lake area was intensively searched, using quadrats measuring 400 m on a side. These were located by an unbiased method in 1979, and in 1983 a new stratum of grassland area was added to the procedure. In addition, Magne initiated research in the Gunn and Yohetta Valleys of the Taseko Lakes region in 1983 funded by the B.C. Heritage Trust, using complementary data collection techniques. Only regional sampling (and limited testing) was undertaken here with the intent of providing baseline data for future studies. The results of this somewhat separate study are described in Magne (1984, 1985b), and the data gathered during the Taseko Lakes project are summarized and used in some of the analyses that follow. As reviewed above, in 1979, considerable effort was also expended in surveying both sides of the Chilko River in an (unsuccessful) attempt to find stratified, non\u00C2\u00ADhousepit sites for regional culture history work. Some additional collection and excavation on selected riverside sites took place in 1983 and 1985. Although we believed we had a good sample of the lower elevation parts of the Eagle Lake area by 1983, we became aware of the importance of higher elevations through local contacts and ethnographic work, along with the opportunity to carry out ethnoarchaeological investigations beyond that which was carried out in 1983 (Burnard M ay 18, 2003-15 1987). The latter were funded by a grant from the Ethnic Studies program of the Canadian government to Matson and largely carried out in the field by Diana Alexander and Robert Tyhurst (Alexander, Tyhurst, and Matson 1985) in 1984 and concentrated on the alpine areas of Potato Mountain, immediately to the south of Eagle lake (Figure 1-1). This investigation provided the basis for another SSHRC grant which supported minor additional work in the Eagle Lake area, and fairly substantial investigations, focussed on regional sampling of Potato Mountain (Alexander and Matson 1987, Matson and Alexander 1990). Funding by the Hampton Committee (UBC) to Matson supported further analysis, dating and manuscript preparation in 2000 and 2001, although the planned fieldwork turned out to be limited to a field reconnaissance. This report also provides background information on the local environment, summaries of earlier and present day Chilcotin ethnography and discussion of regional prehistory. The analytic sections present results of quadrat and s ite environmental and cultural variables analyses, aimed at providing a firm idea of the kinds of sites that occur in the region and their environmental associations, to allow a general description of regional prehistoric settlement patterns. A major part of the report describes the excavation program, descriptive results of the excavations, and analyses aimed at showing the distinctiveness of Chilcotin Athapaskan material culture in relation to the cultural units association with Plateau Pithouse Traditions such as Plateau and Kamloops Horizon (Richards and Rousseau 1987). In these analyses, comparative data from the Mouth of the Chilcotin (Matson, Harn and Bunyan 1984), Anahim Lake (Wilmeth 1978), and Chinlac (Borden 1952) are used extensively both for general May 18, 2003-16 assemblage comparisons and also for more restricted projectile point style comparisons. May 18, 2003-17 Natural Environment of the Eagle Lake Region Martin Magne, Deanna Ludowicz and R.G. Matson The Eagle Lake area was selected for this study of Chilcotin migration because of its environmental similarity to that of the ethnographic Western Shuswap to the east where the M outh of Chilcotin Settlement Project had been carried out (Matson et al. 1984). In particular the extensive open grasslands, and proximity of a major river with a large salmon run in the Eagle Lake region, are similar to the region at the mouth of the Chilcotin River that had been studied by Matson, Ham and Bunyan in 197 4 (Matson et al. 1984). This environmental control is a necessary feature of the parallel direct historic approach as discussed below. Choelquoit Lake, known locally as Big Eagle Lake or Eagle Lake, is located approximately 290 km north-northwest of Vancouver and 120 km southeast of Anahim Lake along the western periphery of the central Interior Plateau (Figure I-1). The lake is 9 km long, lying approximately west-east, and is 2 km wide from north to south. Eagle Lake at one time drained east into Goosenob Lake, and from there northeasterly into the Chilko River. The Chilko River itself is located about 3 km east of the east end of Eagle Lake. The elevation of Eagle Lake is 1189 m (3805 ft) with the Potato Mountain Range rising to 2206 m (7236 ft) from the southwest shore. The peaks of the Niut Range and Razorback Mountain at 2700 m (8856 ft) are prominent features of the horizon when viewed from the east end of the lake (Figure I-2). FIGURE May 18, 2003-18 0 t ... 0 IWIL[ :r-.3 Figure \u00EF\u00BF\u00BD- The Eagle Lake survey universe and sampling frame. Figure I-2. View of Niut Range (right hand side) and Potato Mountain (left) from Northeast side of Eagle Lake (Quadrat 22) The boundaries of the quadrat surveys were the Chilko River on the east and the Potato Range timberline at approximately the 1525 m (5000 ft) contour on the south. The western boundary followed the western shore of Eagle Lake cutting south of Duck.foot Lake. The lack of a natural boundary in the north resulted in the drawing of an arbitrary line along the 1310 m (4300 ft) contour level. The survey universe totals about 83.5 square kilometers in area (Figure I-3). FIGURE Figure I-3. The 1979 and 1983 Eagle Lake Sampling Frame. The study area is located in the southwest portion of the Fraser Plateau of the Interior Plateau (Holland 1964: 69). Flat and gently dipping late Miocene or Pliocene olivine basalt flows underlie most of this area. Little Tertiary erosion occurred here, leaving the plateau relatively undissected and unincised, with no highlands separating the plateau from the coast mountains on the west (Holland 1964: 75). Maximum ice coverage during the Pleistocene is considered to have reached elevations of 2440 m (8000 ft) but the chronology of glacial events for the plateau is problematic (Tipper 1971: 62 - 63). Stages of two or more deglaciations are evident from the complex patterning of land formations resulting from retreating ice. A northward movement is inferred from the moraines deposited by a tongue of ice which flowed May 18, 2003-19 from Tatlayoko Lake valley, while to the south, evidence for a southwestward flow of ice from the Chilko Lake valley to the Homathko River valley is present. Cirque-like basins were formed either during the Fraser or earlier glaciations, and are present in the low hills north of Eagle Lake. The Kleena-Kleene River, Chilko Lake and Tsuniah Lake valleys served as direct over-flow meltwater channels for the retreating ice, which formed small drumlins which Tipper (1971: 61) refers to as \"indistinct glacial grooves\" and esker complexes. Standing on the high ground at the west end of the lake and looking east, the very streamlined, \"tear-drop\" shape of the hill separating Eagle and Fishtrap lakes is very apparent, and indicates an east-west movement of ice. Glacial drift covers most of the plateau and Holland (1964: 70) suggests that less than five percent of the bedrock is exposed. Presently, a grey luvisol, which develops on parent material derived from basaltic rock under coniferous forest, covers most of the area (Farley 1979: 37). With respect to stone materials available to prehistoric inhabitants for chipped stone technological purposes, no major source of basalt, obsidian or cherts are known to occur within the Eagle Lake region. During the course of surveys, scattered cobbles of vitreous basalt were observed. Most of these were small, on the order of 0.1 to 0.5 kilograms in weight, although larger pieces of workable lithic material were occasionally found. Obsidian was in high probability obtained from the Obsidian Creek source area northwest of Anahim Lake. The material was probably traded over the 140 km distance to Eagle Lake. The climate at Eagle Lake is generally cold and dry, lying on the leeward side of the Coast Range, and is subject to a slight rain-shadow effect. A total of 30 to 60 cm of May 18, 2003-20 rain and snow are deposited annually. Temperatures drop to below -15 degrees Celsius in January and rarely reach 15 degrees Celsius in July, although brief spells of temperatures as high as 25 to 30 degrees Celsius usually occur each year in our experience. Fewer than sixty days of the year are certain to be frost-free (Farley 1979: 45 -47).Krajina (1973) identifies three biogeoclimatic zones in the vicinity of Eagle Lake: L Cariboo Aspen - Lodgepole Pine - Douglas Fir; 2. Subalpine Engelmann Spruce -Subalpine Fir; and 3. Alpine Tundra. The first zone is found along a narrow strip, not much wider that 3 km, following the shores of Eagle Lake, Chilko Lake, and the Chilko River (Figure I-2,4). This zone covers a greater expanse of land at the north end of the Chilko River and along the Chilcotin River to the Fraser River, and is most typical of the Chilcotin Plateau. Most of the quadrats that were surveyed in the Eagle Lake region were located in this zone and numerous edible plants have been identified. FIGURE Figure I-4. Cariboo-Aspen-Lodgepole Pine -Douglas Fir zone. Potato Mountain in the Background. (Figure I-2 and 1-4 taken from the same place in Quadrat 22.) Fruit-bearing shrubs include saskatoon berry (Amelanchier alnifolia), juniper CTuniperus sp.), gooseberry (Ribes sp.), soapberry (Shepherdia canadensis), wild rose (Rosa sp.), wild strawberry (Fragaria virginiana), and kinnikinnik (Arcostaphylos uva\u00C2\u00ADursi). Hog fennel (Lomatium macrocarpum), onion (Allium cernum), yarrow (Archillea milletolium), balsam root (Balsamorhiza sagittata) and prickly pear cactus (Opuntia fragilis) are edible herb and root plants which grow in grasslands along sunny, exposed May 18, 2003-21 slopes which are common on the north and east shores of Eagle Lake. The availability of these plant resources is one of the environmental characteristics which the Eagle Lake area shares with areas to the east and south. Eldridge and Eldridge (1980: 20) note that the Dean River Valley, located about 140 km northwest of Eagle Lake, generally lacks root and berry crops. A rapid rise in elevation from the first zone results in a slightly cooler and more moist climate, characteristic of the Subalpine Engelmann Spruce - Subalpine Fir biogeoclimatic zone (Figure 1-5). This is the zone within which the Taseko Lakes region is contained. Spruce, alder, and willow swamps are much more common in this area. The lower reaches of the Potato Mountains to the southwest of Eagle Lake are also in this zone. Food plants to be found here include spring beauty (Claytonia lanceolata), which is a wild root much like a small potato that can be harvested in the late spring from the mid elevations of the Potato Mountains, the Nemiah Valley and Yohetta Valley and in the Alpine Zone in late July. The mid-elevation zone also contains blueberry (Vaccinium sp.) The white bark pine (Pinus albicaulis), the nuts of which are edible, is found at the upper edge of the forested zone. FIGURE Figure 1-5. Subalpine Engelmann Spruce-Subalpine Fir zone. Taseko Lakes region, Tuzcha Lake. The Alpine is the biogeoclimatic zone above the treeline of the Potato Range, the Taseko Lakes valleys and the other mountains in the general region. In many local areas within the Alpine Tundra zone, active glaciers are present and readily accessible. This May 18, 2003-22 zone was intensively investigated on Potato Mountain (Figure I-6) in 1984 and 1985 (Alexander et al. 1985, Alexander and Matson 1987) where it was divided into \"Parkland\" where some scattered trees -- alpine firs and whitebark pine -- still existed and 11Alpine Tundra\" where any trees present were in stunted, Krumholz form. FIGURE Figure I-6. Alpine Zone, \"Parkland\", Middle Mountain area of Potato Mountain. The alpine tundra (Annas and Coupe 1979) usually occurs at elevations above 8825 m (6000 ft), although northeast facing slopes would sometimes have this zone as low as 1700 m (5600 ft). It is distinguished by the lack of true trees, with the only tree species occurring being the alpine fir (Abies lasiocarpa) which took the flattened krummholz form when it did occur. The shrubs varied greatly, from much bare rock exposed with only lichens present, to lush alpine meadows with a wide variety of herbaceous flora. The culturally important mountain potato is found throughout the study area on Potato Mountain where soil and moisture permit, particularly below melting snowbanks. Snowbanks are present year around in protected areas, as indicated by the snow which was present throughout 1985, one of the hotter and drier summers by all accounts. Animals present today include deer, grizzly bear, black bear, marmot, and mountain goat. Bighorn sheep were found in the past, and the discovery of mountain caribou antlers on nearby Mount Nemaia (Alexander et al. 1985:29) indicates that these too may have been present in the past. Wapiti remains were also May 18, 2003-23 recovered at Eagle Lake in excavation contexts dating to the last 1000 years, which indicates the presence of this species in the general area and possibly seasonally in the alpine zone. Wolf was observed as well. Ptarmigan is quite common in some parts of the alpine zone and is the only important aboriginal bird resource found away from lakes. There are a number of lakes present in the Potato Mountain survey of the alpine zone, from Gillian and Dunlop on the west, to numerous small unnamed ones in the east. A variety of migratory waterfowl were found in some of these, often with young. Evidence of forest fires was apparent in many of the quadrats surveyed around Eagle Lake. Burned areas appeared to be rapidly replenished with mixed forest of trembling aspen (Populus tremuloides) and lodgepole pine (Pinus contorta). Preliminary inspection of several tree cores that were collected suggests that a major burn occurred between 60 and 90 years ago, however cores from all of the burn areas were not obtained. The dendrochronological work described below may eventually provide more definite evidence about this widespread environmental process. The mammal, bird and fish populations are believed to not be as plentiful as they had once been. The area is presently being grazed by cattle, putting pressure on the amount of food available for wild species, and local inhabitants blame an influx of white hunters for depletion of game populations. Moose (Akes alces) have recently moved into the central Interior Plateau (Farley 1979: 53). Elk or wapiti (Cervus elephus), on the other hand, are not present today but were present in the past. This is based on testimony from local inhabitants as well as the occurrence of elk remains in our excavated sites dating to pre-Chilcotin times (see faunal analyses below). Native ungulates include mountain sheep (Ovis canadensis), May 18, 2003-24 m ountain goat (Oreamnos americanus), caribou (Rangifer tarandus), and mule deer (Odocoileus hemionus). Deer is reported to be the most plentiful ungulate in British Columbia, though populations are declining (Farley 1979: 53). Carnivores currently present include grizzly (Ursus arctos ) and black (Ursus americanus) bears, wolf (Canis lupus), coyote (Canis latrans), lynx (Lynx lynx), wolverine (Gulo gulo), mink (Mustelavison), marten (Martes americana), fisher (Martes pennanti), and weasel (Mustela frenata). Other small mammals observed in the field include snowshoe hare (Lepus americanus), marmot (Marmota monax), squirrel (Tamiasciurus hudsonicus), flying squirrel (Claucomys sabrinus), beaver (Castor canadensis), and small brown bat (Myotis lucifus). Land capability maps suggest moderate to severe limitations for waterfowl around Eagle Lake (Canada Land Inventory 1970), although a large gaggle of Canada geese (Branta canadensis) frequented the south shore of the lake for several weeks in the summers of 1979 and 1983. Various other bird species have been identified, including bald eagle (Haliaeetus lencophelus), osprey (Pandion haliaetus), loon (Gavia immer), spruce (Canachites canadensis), and ruffed (Bonasa umbellus ), grouse, yellow-bellied sapsucker (Sphyrapicus varius), common tern (Sterna hirundo), sandpiper (Erolia sp.), raven (Corvus corax), robin (Thordus migratorius), barn swallow (Hirundo rustica), and rufous hummingbird ( Selasphorous rufus). Various fish species are available throughout the year. Rainbow trout [formerly Salmo gairdneri, now most likely Oncorhynchus mykiss (Smith and Steady 1989)] and suckers (Catostomus sp.) spawn in running fresh water that enters lakes in early spring and the adults can be caught in the lakes throughout the summer. Late summer sees May 18, 2003-25 the beginning of the annual sockeye salmon (Oncorhynchus nerka) run up the Chilko River (Figure I-7). This run is most plentiful for two consecutive years out of a four year cycle of abundance. Some fewer Chinook or spring salmon (Oncorhynchus tshawytscha) also spawn in the Chilko River. Kokanee, which are land-locked sockeye salmon, are present in Eagle Lake and spawn in the small creek draining Fishtrap Lake at the western end of Eagle Lake (Figure I-8). The Chilko River also contains Dolly Varden (Salvelinus malma) and mountain whitefish (Propsopium williamsoni). These fish can also be caught through lake ice during the winter season and mountain whitefish are present in Eagle Lake FIGURE Figure I-7. Chilko River, facing north, 2 km north of Lingfield Creek. The Eagle Lake region offers a wide variety of plant, mammal, bird and fish species throughout the year. The animals available vary widely in the ease with which they can be procured as the result of vertical and lateral zonations, and also because of marked seasonality. Chilcotin adaptations to this landscape are the subject of the following section. FIGURE Figure I-8. West end of Eagle Lake from the north end of Potato Mountain. Fish Trap Lake is seen immediately to the left of the west end of Eagle Lake. May 18, 2003-26 Chapter 1I SETTLEMENT PA TIERNS: ETHNOGRAPHIC AND ARCHAEOLOGICAL Introduction Given that the basic goal of our investigations was to identify the timing of the Athapaskan movement into the Eagle Lake area, what are the expected characteristics of the Athapaskans as compared to the previous inhabitants? In this section we summarize the relevant ethnographic characteristics of the Chilcotin and the presumed previous inhabitants -- the Plateau Pithouse Tradition--and review the settlement p attern archaeology of each. We begin with a very brief summary of the ethnographic description of the Plateau Pithouse Tradition, and a more extensive summary of the relevant archaeology, focussing on the settlement pattern research at the Mouth of the Chikotin (Matson et al. 1984), as this material is far better known than B.C. Athapaskan settlement archaeology. We then turn to a more extensive review of the Chilcotin ethnography, followed by a short review of our ethnoarchaeological investigations, and finishing with our archaeological expectations for Chilcotin settlement patterns. Plateau Pithouse Tradition The Plateau Pithouse Tradition has become the favored term of archaeologists following Rousseau and Richards (1985) and Richards and Rousseau (1987), to refer to the ethnographic Plateau Culture area (Walker, 1998, Handbook of North American \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD=,L...!.\u00EF\u00BF\u00BD-\u00EF\u00BF\u00BDi..t:u..2-..Aut.::.::e=.!a:a\u00EF\u00BF\u00BD.) and its immediate antecedents. For the Canadian portion of the Plateau area, Richards and Rousseau (1987) describe three broad culture units, named Shuswap, Plateau, and Kamloops, of which the last is the archaeological May 18, 2003-27 equivalent of the ethnographic people. The Shuswap culture is now dated to more than 3000 but less than 4000 radiocarbon years ago (Stryd and Rousseau 1996), the Plateau 2400 to 1200, and the Kamloops 1200 B.P. to A.D. 1800. The ethnographic Plateau Culture area includes a number of different groups but is best known in B.C. through the ethnographies of Teit on the Thompson (1900), Shuswap (1909) and the Lillooet (1906). All of these groups speak Salish languages, but other language families are spoken elsewhere on the Plateau. A sketch of the principal parts of the Plateau Culture as described mainly by Teit follows. The most important economic feature of Plateau culture is reliance on salmon. As far back as Wissler (1917) the Plateau was seen as part of the Salmon area, which he divided into what today are called the Northwest Culture and Plateau culture areas. Although not all parts of the Plateau have access to significant amounts of salmon, those that did are the places with higher population densities and, where present in numbers, salmon were the most important resource. Along the Fraser River, Sneed (1971) showed that population sizes are highly correlated with a relatively crude measure of salmon abundance. Kew (1992) offers a comprehensive discussion of the abundance, distribution and aboriginal use of salmon in the Fraser River system, including the different technologies involved that supports such a conclusion. Although the main stream of the Fraser has several runs of salmon, away from the lower reaches, the most important are the Sockeye and Chinook runs in late summer and early fall. Salmon fishing camps were positioned to obtain and process the salmon at these times. A wide variety of procedures were used to catch salmon, but two most common techniques were the dip net where waters were murky, such as on May 18, 2003-28 the main stream of the Fraser, and harpooning or spearing, now called \"gaffing\" in clear waters. The latter technique was also carried out at night with torches. In clear waters, the salmon would see, and thus avoid, the dip net. Both of these main techniques were most efficiently carried out adjacent to constrictions and rapids along the rivers. The netting and gaffing were carried out by males, but the all important processing and drying was usually done by women. Drying racks are still seen today, although most fish today are put fresh into freezers. Fishing camps, then, could be extensive settlements, involving a large number of people and lasting for many weeks, although since fishing occurred at low water, remains might be regularly washed away. The winter pithouse village was usually moved to next and was generally located not too far away. These settlements would be usually located in fairly low elevations, apparently to be closer to the location of salmon and for milder winter weather. They would also be positioned close to supplies of wood which often meant being located away from the main rivers, as many lower elevation parts of the Fraser and Thompson river valleys are grassland and do not have quality timber for construction purposes. These settlements were usually also located adjacent to good water sources which was an especially important factor on the Fraser, as that murky water was avoided for drinking purposes. This winter basecamp usually defined the local band which had a modest 25-50 size, and typically was exclusively occupied from November to February, although the camp would used as a base for longer than that. The structure that gives the archaeological name to this tradition, the pithouse, was a relatively robust structure, as idealized in Teit (1900) for the Thompson Indians. It would consist of a pit, 5 to 10 meters in diameter, excavated 40-80 cm deep, and with May 18, 2003-29 Pithouse ideally a four post frame with hip rafters (Figure 11-1). These rafters were covered with smaller pieces of wood and, finally, a layer of earth over the top. A smoke hole was left in the center and a notched log for a ladder was used for the entrance, also through the smoke hole. This structure was very energetically efficient (MacDonald 2001), but likely very smoky and dirty, and people were glad to leave them when better weather came along. Recent experience in building replicas indicates that they need constant upkeep and do not last too long without rotting, etc. It is the group of large depressions that existed after the pithouses had decayed that make this settlement type so visible to the archaeologist and made them the focus of archaeology in the 1960s and 70s, although the constant re-excavations of the houses during re-use leads to difficulties in archaeological interpretation (Wilmeth 1977). F IGURE Figure II-L Teit' s (1909a) Rendition of the Pithouse. The processed salmon were typically stored in cache pits, pits from 70 to 300 cm in diameter and up to more than a meter deep. These could be located in pithouse villages, but were often in hidden localities, and are sometimes found in large numbers particularly close to riverside fishing sites. Resources other than fish were also stored in cachepits; roots were the most important. Most root crops were obtained in the spring with the seasonal availability reflecting elevation, with a given root being available later in higher elevations, and higher elevation root species being available as the snow melted. The two most May 18, 2003-30 important root crops in the northwestern part of the Plateau culture area were balsam root sunflower (Balsamorhiza sagittata) and spring beauty or mountain potato (Claytonia lanceolata). The balsam roots were obtained early in the spring and were most abundant in middle elevation grassy environments (1000-1500 m) while mountain potatoes are most abundant in alpine meadows after the snow melts in late July and August. Marmots, which were important for their fur, were also concentrated in alpine areas. The root crops were typically roasted in pits, near where they were obtained. In some cases, rich root areas were locus of large multi-band camps, with Botanie Valley, near the junction of the Fraser and Thompson, being perhaps the best known. Longer established camps away from the winter village might have \"mat lodge\" structures. These are usually rectangular in nature, and there is some evidence that with the horse these became more common and partially replaced the winter pithouse (Alexander 1992), particularly on the United States portions of the Plateau. Richards and Rousseau (1987:43) report that rectilinear structures show little evidence of extensive roof insulation indicating the use of mat structures rather than the classic pithouse structure illustrated by Teit (1900, 1909a). They further state that these structures are present only in the Kamloops culture. The actual ethnographic evidence about root crops is quite limited, but archaeological investigations (Pokotylo 1978; Pokotylo and Froese 1983: Peacock 1998: Alexander and Matson 1987; Matson and Alexander 1990; V anags 2000) have clarified the importance of this procurement system in Canada and many of the details. We will see that this is an important activity in the Eagle Lake area. Fish besides salmon were also important on the Plateau, but given the variability May 18, 2003-31 in these resources it is difficult to generalize. Spring salmon runs of Chinook, where they existed, were important. Spawning of rainbow trout, and minnow family, soon after the ice came off the lakes, were almost always heavily exploited where available. Fishing other times of the year could also be important. Even sturgeon, on the Fraser River, were exploited. Hunting deer, elk, and mountain sheep were also important activities, but not compared to the fish runs. Hunting large mammals was particularly important after the salmon runs and before the onset of winter. Bow hunting was probably the most important technique, although traps, fences and other methods were also used. Hunting might well occur along side of root collecting. Other animals, including waterfowl, were also obtained and could be seasonally important. Although roots were the most important plants for foods, a very wide variety of other plants (and other roots) were exploited (furner 1978). Perhaps the most important were saskatoon, service or June berries (Amelanchier alnifolia) and buffalo\u00C2\u00ADor soap-berry (Shepherdia canadensis), with both of these becoming available in the summer. To summarize the ethnographic settlement pattern, the major sites would be the fall salmon fishing camps, the winter pithouse village, with other large camps possible at fish spawning locations and rich root gathering grounds. A variety of smaller hunting and collecting locations would also be expected. In addition to these settlements, cache pit sites and root-roasting pits would be archaeologically visible. Teit (1900, 1909a) represents the Interior Salish as relatively egalitarian in nature, with very little political complexity or inherited social status. This particular question May 18, 2003-32 was a focus of Ray1 s (1939) investigation who also found the Plateau was relatively egalitarian. Recently claims of greater complexity on the Plateau have been made (Hayden 1997, Schulting 1995; Hayden and Schulting 1997). There is one clear possibility of greater complexity on the Canadian Plateau, the Lillooet phenomenon, which is a time (circa 900-1500 BP) of very large villages of very large pithouses, best known around Lillooet but apparently also extending up the Fraser into the Chilcotin/Chilko river system. However with the exception of the Lillooet Phenomenon, there appears to be a distinct difference in degree of social complexity between the interior and the coast (Matson and Coupland 1995). If one claims 11complexity11 for the Plateau as a whole -- and they are more \"complex\" than many ethnographic hunters and gathers, the Great Basin Shoshone, for example--what does one consider the Northwest Coast? Relatively speaking, then, in this area, the Plateau Pithouse Tradition, with the possible exception of the Lillooet phenomenon, should not be considered complex. How does the current archaeological information about the Plateau Pithouse Tradition correspond to the ethnographic sketch above? Most of the features that are easily recognized through archaeology are confirmed for the Plateau and Kamloops cultures (Richards and Rousseau 1987). The remnants of winter pithouse villages, sometimes called housepit sites, have been long recognized, and are concentrated along salmon streams. It is unclear if the superstructure described by Teit (1900) is valid for the pre-Kamloops cultures, but certainly pithouses of some sort were used in the Plateau and Shuswap cultures. The occurrence of these structures in groups is most evident along the Fraser River, while more isolated, or more strung out sites are seen May 18, 2003-33 along the Thompson, perhaps the result of potable water being present all along the Thompson, or less concentrated fishing locations along the Thompson. Similarly, cache pit sites have long been recognized (Sanger 1970:17) although they have not been a focus of much attention. Most excavation has focussed on pithouse sites. Although not well described in the ethnographic literature, the mid\u00C2\u00ADelevation root procurement system is well known, thanks to the investigations of Pokotylo (1978; Pokotylo and Froese 1983; Vanags 2000) in Upper Hat Creek valley and subsequent research by Stryd (1995) and Peacock (1998). Investigations at Botanie Valley by Baker (1975) and Rousseau et al. (1991) have confirmed and expanded our understanding of the use of that area (as does the research reported for the Eagle Lake region). The archaeological recognition and understanding of root-roasting pits was almost non-existent in 1975 and they are now well-known. Fishing sites, often used today, have also been recognized by archaeologists, although the archaeological evidence, outside of location, is often not sufficient to confirm that function. The ethnographic or historic use, though, leaves little doubt as to their function in many cases. Alpine root procurement is not as well-known, but is well described for Potato Mountain (Alexander and Matson 1986) and for alpine areas near Hat Creek (Alexander 1992; Rousseau et al. 1991; V anags 2000). These investigations support the ethnographic descriptions in general, and add details in some areas. Beyond these site types, the assignment of other functions to other sites is of a lesser level of specificity, and confidence. Judgements such as \"limited activity site'' with a possible \"hunting'' function is often the only viable interpretation, based on what May 18, 2003-34 is usually a limited set of chipped stone debitage and tools. Which, of the wide range of activities that the ethnographic accounts give us, is actually represented -- or even some other activity-- cannot be determined at this time. In sum, for the Plateau and Kamloops cultures, the most intensive and most archaeologically visible activities recorded by Teit are well-recognized, and in good agreement. Going beyond salmon fishing location, winter pithouse village, cache pit, mid-elevation or alpine root-roasting, though, is not yet a standard part of archaeology. Magne (1985a), however, was able to show that those assemblages associated with housepits and cachepits can be distinguished. The earliest part of the Plateau Pithouse Tradition, the Shuswap culture (3300-2400 B.P.), at this time does not show the presence of root-roasting, indicating, at the very least, that it was less intensive at this time. Furthermore, there are indications that the use of cachepits was different at this time, as they are usually not present in pithouse sites of this age, and may not be present at all. Richards and Rousseau (1987) present an argument that the covering of the Shuswap pithouse was lighter than indicated by ethnographic descriptions and therefore that the superstructure described by Teit (1900) was absent. In sum, the major features of the ethnographic pattern are present in the Plateau and Kamloops cultures, but there appear to be some significant differences in the earlier Shuswap culture. Mouth of Chilcotin The above presentation is very general. What about specific locations that are relevant to the Eagle Lake area? Detailed settlement pattern information was one of the May 18, 2003-35 goals of the research carried out at Mouth of the Chilcotin (MOC) in 197 4 (Matson et al. 1984) which will be delineated after the methodology and field procedures common to the MOC, Eagle Lake, Potato Mountain, and Taseko Projects are described. Even though the MOC project was very modest and carried out a long time ago, it appears still to be the most detailed settlement pattern analysis of the Plateau Pithouse Tradition. What follows is a general summary, without the details in that later report, but much more specific than the ethnographic-based sketch above. The MOC survey was in the Canyon Shuswap territory, one of the most densely populated Shuswap areas. A total of nine pithouse villages were located in the MOC sample survey. All of these had cachepits located either adjacent to the pithouses or nearby, and all consisted of multiple housepits in close proximity, defined as within 100 m. (The sites were mapped, collected, and recorded using a gap of 50 m to indicate a separate site, so 12 separate ''archaeological\" sites make up these nine.) As will be described in more detail later, the lower elevations at MOC are grassland covered from the Fraser River up to about 2500 ft elevation, where a dense, Douglas Fir forest replaces it. Eight of the nine pithouse villages were located close to the ecotone of the grassland and the Douglas fir forest, some out in the grassland, and some in the edge of the trees. As there were patches of wood along the Fraser, we interpreted this location as indicating not only the need for wood for firewood and structural timbers but also the dislike of drinking Fraser River water. The ninth pithouse site was located adjacent to the Fraser, in a location with trees. There was also a stream channel adjacent, which was certainly dry in the summer. Some of the pithouse sites were located near clear sources of water, but we had to search for sources of waters for others, and in some May 18, 2003-36 cases, they were small and meager. As this project (and other visits) occurred in the summer, the actual case in November-February may not have been so bleak--and the stream channel located adjacent to the pithouse village adjacent to the Fraser may have been running then. So the wood and water and avoidance of the Fraser River water appear to be generally confirmed for the MOC pithouse villages. Cache pit sites were found in a number of situations at the MOC, including in hidden areas, close to pithouse villages. Six, having from 3 to 8 cachepits each, were located along the steep ravines that ran between the top of the grassland to the banks of the Fraser River. These would be in hidden locations and half way between the Fraser and the pithouse locations, a good place to store salmon. In three cases, cache pit sites were located adjacent to the Fraser River where they had from 3 to 23 cachepits. Small numbers of them were located in grassland-forest ecotone areas usually slightly higher than the pithouses, where other resources could have been cached. Twelve of these \"ecotone cachepits\" were located and six of these consisted of a single cachepit and two of 2 cachepits. The two largest (7 and 8 cachepits) were located within 200 m of housepit sites. It is clear that the cachepits are concentrated in areas convenient for storing salmon, with smaller numbers located elsewhere. In the survey sample, a single riverside quadrat had evidence of riverside fishing, which was recorded as two separate sites. Other fishing sites were definitely present in that area, some of which were still in use by Shuswap (Secwepemc) people in 1974. The above sites constitute the vast majority of the sites discovered at the MOC. The other significant category were the four \"chert debitage\" sites interpreted as Pre-May 18, 2003-37 Plateau Pithouse Tradition. These are named after a distinct chert, which Vanags (2000) also records at a site on Pavilion Mountain. In addition three other sites were located that could not be forced into any of the above categories. What is missing from the MOC is any evidence of a root procurement system. Only six quadrats were surveyed in elevations above the grassland environment, and only one contained Balsamroot sunflower. No root-roasting pits were located there. Pokotylo' s (1978) work indicates that root roasting pits would be expected in the mid elevation open areas with Balsamroot sunflower present, such as at Eagle Lake. In alpine areas, Alexander's (1992), Pokotylo and Froese' s (1983) and Vanags' (2000) investigations show that roasting pits would be expected near areas with abundant mountain potato. Pokotylo' s (1978) investigations show that within the broader zones roasting pits tend to be near wood and sources of water. These associations make sense in terms of the amount of wood needed to roast roots, compared to the weight of roots, and the use of water to keep the roots from drying out while pit-roasting--or for general camp use. PPT Settlement Pattern Using the MOC results as the expected PPT pattern for Eagle Lake, we predict six kinds of sites for the Plateau Pithouse Tradition in the Eagle Lake Area. First, remains of winter villages of pithouses, consisting of a number of housepits, usually placed dose together, associated with cachepits. These sites are expected to be located relatively near good salmon sources, which in the Eagle Lake case is the Chilko River, and to be immediately adjacent to or in wooded areas with close access to potable May 18, 2003-38 water. In this area, the Chilko River is considered to be potable. Second, riverside fishing sites, adjacent to good fishing locations are also expected. Cachepit sites, with substantial numbers of pits, are expected close to riverside fishing sites (three), and in intermediary areas (four) between fishing sites and pithouse sites, if the latter are located at a substantial distance from the fishing sites. Fifth, smaller numbers of cachepits should be located in areas away from the river, sometimes within a few hundred meters of housepit sites, and other times not near any housepits but not in the area between the river and the housepit sites. This last class of cache pits sites may be used for general storage by nearby winter villages, but in hidden locations, or for storage of resources other than salmon. Finally, root roasting pit sites would be expected in areas near where Balsamroot sunflower was present, and probably close to wood and water. These should approximate those found by Pokotylo (1978; Pokotylo and Froese 1983) in Upper Hat Creek Valley. At the time of the surveys around Eagle Lake in 1979, and 1983, the use of the alpine area was unknown everywhere. In 1983 a little work was done adjacent to the alpine zone in Hat Creek by Pokotylo (Vanags 2000) so prior to our work in 1984 and 1985 on the top of the Potato Mountains, the archaeological use of this zone in Plateau Pithouse Tradition in Canada was unknown. The ethnographic accounts are unclear on what would be expected in terms of archaeological remains (Alexander 1992:101). Only some generalized campsites are suggested by Teit (1900, 1909a). May 18, 2003-39 Chikotin Tradition: Ethnography and Settlement Patterns Introduction The following summary is both more detailed and particular than the previous Plateau Pithouse Tradition summary. The information on the Chilcotin is much more limited and less well-known, so that a more detailed summary is in order. Moreover, it is the Chilcotin entrance into this area that is being investigated, so the particulars relevant to the Eagle Lake environment are emphasized. There are only a limited number of sources for the Chilcotin, and all have various drawbacks for our purposes. The best source, Lane (1953,1981), has the drawbacks that the fieldwork occurred only in 1948-1951, after many post-contact changes, and that it was oriented towards a number of comparisons with neighboring groups, rather than the traditional descriptive ethnography so highly valued by archaeologists. Perhaps the second best source is Teit (1909b) which was based on only two weeks of concentrated work and who came to some conclusions supported by no other recent investigation. Other useful sources include Farrand (1898, 1900, 1910) who concentrated on oral narratives, Morice (1890, 1895, 1906-1910) whose accounts focus on the Carrier and only occasionally refer to the Chilcotin, and Ray (1939,1942) who used only a single informant whom he regarded as not very reliable. Tyhurst (1984; Magne and Tyhurst 1984) carried out dissertation research on the Chilcotin 1975-1983 and has also summarized earlier ethnographic research, but little of his research is directly relevant to the pre-contact situation as it was focussed on recent economic changes. Mav 18, 2003-40 ., Ethnographic Sketch The word \"Chilcotin\" (Tsi'lhqot'in) comes from a rendition of the Chilcotin term for \"people of the Chilko River\", which for the Chikotin meant the combination of the Chilcotin and the Chilko rivers (Lane 1981:442). Their traditional territory at contact time (Figure II-2) extended from the upper Dean River north of Anahim Lake, southeast to the southern end of Chilco Lake, and to Coyote Rock, just east of Alexis Creek. This territory is about 28,000 square kilometers (11,000 square miles) and Lane (1981) estimates precontact population as between 1000 and 1500 giving between 19 and 28 square kilometers per person. Access to the all-important salmon resources are along the Chilko-Chilcotin along the eastern edge of their precontact territory, and at the border with the Bella Coola at their northwest portion; the Chilcotin population was probably concentrated accordingly. Lane (1953,1981) and Teit (1909b) agree that the Chilcotin had their most intensive and peaceful interactions with the Bella Coola and Canyon Shuswap (MOC). It was between the two that they acted as middlemen, trading interior items collected by the Canyon Shuswap, as well as material they collected, for coastal items provided by the Bella Coola. This trade was an important source of dried salmon for the Chilcotin, given their limited access to this resource Oorgenson 1980:125; Teit 1909b). Chilcotin relationships with their other main neighbors, the Southern Carrier, to the north (Gibson 1998; Lane 1953:66-7 4) and Lillooet, to the southeast (Lane 1953:81-86) were less regular and included a number of episodes of conflict (Gibson 1998:70; Teit 1909b ). In contrast their relationships with Bella Coola and Canyon Shuswap were generally peaceful and included substantial amounts of intermarriage. May 18, 2003-41 FIGURE Figure II-2. Historic Aboriginal Groups of Interior British Columbia The main subsistence activities of the Chilcotin were hunting and fishing. According to Lane (1981) few animals were not eaten and the most important game were elk, deer, caribou, mountain goats, and sheep. Black bear, marmots, hares, beaver, muskrats, and porcupines were also procured. Important birds were ducks, geese, ptarmigan, and grouse. Most hunting was carried out by individuals by stalking and ambushing. Traps, snares and deadfalls as well as group hunting were also used. Compared to the Plateau Pithouse Tradition, the Chilcotin relied more extensively on large mammal hunting, in line with their reduced access to salmon and root crops. Teit (1909b:782) adds otter, muskrat, and marten to the list provided above and remarks that women snared rabbits and muskrats. Fishing was also very important, although not limited as much to salmon as in the PPT. Salmon was the most important fish, with sockeye being preeminent, followed by Chinook and land-locked kokanee. Trout, whitefish and suckers were also noteworthy. Certainly the early spring spawning of the lake trout and suckers (minnows) in streams running into lakes was very important as that occurred at a time when few other resources were available and when stored foods were usually exhausted. Further, winter lake fishing, as with many Northern Athapaskans, was very important, with the mountain whitefish being particularly valued in this regard. Lane (1953:43), in fact, indicates that lake fishing was the most important subsistence activity, May 18, 2003-42 followed by river salmon fishing and hunting. Tyhurst (1984), however, thinks that salmon fishing was the most important pre-contact subsistence activity. Both kokanee and whitefish spawn in the fall at streams running into lakes. According to both Teit (1909b) and Lane (1953, 1981) traps were used extensively to obtain spawning fish. Other fish such as dolly varden and sea-run steelhead contributed substantially. Both dip nets and harpooning (now usually called \"gaffing\") were used for salmon, depending on the turbidity of the water. In general the techniques used for salmon were the same as those used elsewhere on the plateau (Kew 1992). Other types of nets were also used for fish, although the use of gillnets in precontact times was questioned by some of Lane's (1981 :405) informants. Fish spears or leisters were also used. In comparison with the PPT, the importance of winter fishing and fish other than salmon were greater. Vegetable resources included both berries and roots. Teit (1909b:780) indicates that the roots were obtained and processed as among the Shuswap, although he indicates that they were of lesser importance. Mountain potato (Claytonia lanceolata), hog fennel, wild onion (Allium cernum), Columbian lily and fern-root are the ones he specifically mentions. (Lane (1953,1981) apparently did not identify roots, only mountain potato is specifically indicated.) Tyhurst (1984) adds avalanche lily, balsamroot sunflower, tiger lily, silverweed cinquefoit and biscuit root to this list. Teit (1909b:780) indicates that principal berries are the saskatoon (Amerlanchier alnifloria) and soap-berries (Sherperdia canadensis) and Teit (1909b) and Magne and Tyhurst (1984) add, among others, raspberries, blueberries, strawberries, and kinnikinnik (Arctostaphylos urvi-ursi) berries. Lane (1981:406) indicates that roots were important May 181 2003-43 11-3 Puntzi Lake 951 in the early spring. The use of roots and berries thus appears to be very similar to that of the PPT as suggested by Teit (1909b:780). Another important resource in the early spring was the cambium layer of trees. When the sap began running the cambium layer becomes sweet and one can peel the bark from a tree and scrap it with a caribou antler scraper (Teit 1909b:781). Lodgepole pine, so ubiquitous in the Chilc\u00C2\u00B7.tin, was the preferred tree. It is likely that this was most important in the resou,\u00C2\u00B7 \u00C2\u00B7arce early springtime. As indicated above, 1\u00EF\u00BF\u00BD ,f the tools used by the Chilcotin differed little from those used elsewhere on the ] ,:au. Decorative style, though, was distinctive, as shown by Teit (1909b) for the baskets. These, although made using similar materials in a similar fashion as the interior Salish, are easily distinguished as the Chilcotin usually divided their basketry decorations into four zones, something not carried out by the Salish. For house structures in the winter, the Chilcotin used two, a small pithouse, said to be identical to and a recent adoption from the Shuswap (Lane 1953:146,160, 1981:403; Teit 1909b:775; Ray 1939:133), and an above ground 11rectangular lodge\" (Figure II-3). This latter is the Chikotin among Northern Athapa\u00EF\u00BF\u00BD, (1953:144-146). FIGURE n of the above ground rectangular structure common rhe best description of the iodge is from Lane Figure II-3. Puntzi Lake Lodge. Photo courtesy of Robert Lane. May 18, 2003-44 \"It was rectangular. The size varied. Most of the estimates centered around twenty feet long by fifteen feet wide. The floor was leveled, but not excavated. There was usually one end-post at each end of the house, eight to twelve feet tall and eight to twelve inches in diameter. These were slightly grooved on top, and supported the ridge-pole. Several poles leaned against this ridge-pole, forming a gable shaped frame. There were at least two of these poles on each side. Bark, poles, or split logs which were usually but not always peeled, were laid horizontally upon this frame, covering both sides almost but not quite up to the top. Thus there was an opening several feet wide the length of the house just under the ridge-pole. The ends were enclosed by closely spaced vertical bark slabs, poles, or split logs. These were supported upon the end frame poles, and their horizontal cover. Only the end-posts were set into the ground. None of the posts were lashed or mortised. At one end, some of the vertical pieces were left off to make the door. This opening was covered with a skin. The house was covered with a layer of grass, sod, or bark to chink the holes.\" Lane (1953:45-46) also describes variants of this and mentions that sometimes a small tree is topped to provide one of the end-posts as appears to be the case in Figure II-2. Teit (1909b:775-776) provides a less detailed description, but adds that the floors oflodges were usually strewn with fir or balsam branches. Thus, one of the two types of winter houses is very distinct from the PPT pithouse, although the remains might not look that different from those of a mat lodge. A less robust form of the above structure was also used in other seasons, where a Mav 18, 2003-45 ., '1,ark\" house might use a rope as a ridge pole. This was basically a lighter and less weather tight version of the above. In the summer, brush shelters were usually erected, but as Lane (1953:46) notes \" ... in both summer and the winter, people often camped in the open with no shelter.\" The Chilcotin were divided into four to six loose groups referred to by Lane (1981:407) as \"bands\". These groups did not regularly meet face-to-face, but did consist of families that did interact over time and were interrelated. The winter settlement consisted of a camp of a very small group of families, with no more than two or three families at any single location. This is a much smaller group than the PPT winter pithouse village, and it also was less permanent, as discussed below. Larger assemblages occurred in three different settings. First, in the spring, the fish spawning settings could be times of larger aggregations. Second, in July and August, members from a number of bands gathered in a single very large camp in alpine areas for mountain potato harvesting and marmot hunting. The Potato Mountains, immediately south of Eagle Lake, was the best known of these settings. Finally, good salmon fishing locations often attracted large groups from a variety of areas. The last two kinds of aggregations still occurred as recently as the late 1940s and the large fishing camp still occurs on a reduced basis at Henry's Crossing and at Si wash Canyon, both along the Chilko River. During much of the year, though, a single nuclear family was the basic economic and social unit, although usually part of a camp of several families (Lane 1981:406). This contrasts with the PPT where a larger, more cohesive \"band\", i.e., winter village group, appears to be much more important. In general, the Chilcotin have a more May 18, 2003-46 flexible and mobile way of life than found in the PPT. Earlier we mentioned that Teit (1909b:786) came to some conclusions that have not been supported by any other later investigators. This is his statement: '' . .it appears certain that the tribe was organized in manner similar to that of the Coast tribes. They seem to have had three classes, --nobles, common people, and slaves.\" He does preface this statement with \"From the assertion of the Shuswap, and from what little information I managed to gather from the Chilcotin themselves ... \" Teit (1909b:786) also reports the presence of clans. Lane (1953:51, 186-188, 1981) explicitly rejects both features, as does Ray (1939:39) at least of being a pre-contact phenomena, as does Tyhurst (1984). Lane (1953:186-193) has quite a long discussion about this matter and we think it can be safely dismissed. Many of the differences between the PPT and the Chilcotin can be understood by examining Athapaskans living further to the north, generally in areas without reliable salmon runs and root crops. In such situations, winter base camps are smaller and less permanent than seen among the PPT, and tend to be focussed around lakes where whitefish could be obtained in the fall and in the summer. In such a setting, a lower population density is present and large mammals have a greater importance than in the PPT. Most Athapaskan groups to the north have some sort of rectangular winter habitation, ranging from relatively impermanent structures such as seen among the Chilcotin, such as the Tsetsaut (Duff 1981 ), to larger versions of the same basic layout (Stuart Lake Carrier), to the much more permanent rectangular structures of the Koyukuk in Alaska (Mcfadyen Clark 1996). In many respects the Chilcotin in adaptation, settlement pattern, and material culture appear to be transitional between May 18, 2003-47 the Athapaskan pattern of further north and the PPT. Chilcotin of the Eagle Lake Region The Eagle Lake region is near the southern limit of Chilcotin territory in 1850 (see maps in Teit (1906, 1909a). It lies on the division between two different environments, the rolling Chilcotin plateau and the back side of the coast range, an area of deep trenches filled with large, long lakes (Tatlayoko, Chilko, Tsuniah, Taseko Lakes) separated by mountains, some with extensive alpine areas. The lakes act as both barriers and transportation corridors, and the rivers, mainly as barriers. One of the interesting traits of the Chilcotin was bridge-buiiding (Lane 1953) to make crossing these rivers easy. Thus this area had access to the three most important pre-contact resources, the salmon on the Chilko River, lake fishing, and alpine areas ( on Potato Mountain.) As indicated earlier, at contact, the population appeared to be concentrated in the Anahim Lake area and along the upper Chilcotin and Chilko Rivers, the two areas within Chilcotin territory that had access to significant amount of salmon. The Eagle Lake area is at the upper end of the Chilko Lake area with access to lakes and the Potato Mountain alpine meadows. The occupied Indian reserves closest to the region are at Nemaiah Valley (current name Xeni Gwet'in First Nations) some 35 km southeast (but two to three hours by road), and the Alexis Creek Reserve at Redstone Flats, about 70 km northeast. Two families in 1979 and 1983 lived within 15 km of Eagle Lake, and raised cattle and crops throughout the area. This region is remote even though it is within 300 km by air from the Lower Mainland of British Columbia. The nearest community of any size, Williams May 18, 2003-48 Lake, though, was 4 to 5 hours by gravel road in 1980. This remoteness has contributed to the endurance of the traditional Chilcotin life-style. Haig-Brown (1983) has noted that the Chilcotin are exceptional in that 90% of the populace retains the language, versus the average of some 5% language retention for other British Columbian Indian groups. Post Contact History Tyhurst (1984) conducted research concentrating on the post-contact history of the Chilcotin. What follows is largely based on his account, with some additional material from Lane (1953) and Wilmeth (1978). The first Euro-Canadian contact with the Chilcotin was via the fur trade in the early 19th century (Simon Fraser, 1808), though some European trade goods had already reached the Chilcotin from the coast before that time. Fort Alexandria established in 1821 on the east bank of the Fraser River, north of Williams Lake and moved to the west bank in 1836 (Gibson 1998:68-79) was the main fur trade contact with the Chilcotin. It was probably near here that on July 26, 1808 that Fraser encountered \" .... a large assembly of Athnahs (Shuswap) and Chilkoetins. The latter are from the Westward and came on purpose to have a sight of us, having never seen any white people before\" (Fraser 1966 [1808):124). The Shuswap and Carrier living at Fort Alexandria were attacked at least once by the Chilcotin (Gibson 1998:70,209,222-225; Lane 1953:72). The Chilcotin were first visited during 1822 and 1825 and a subsidiary outpost, Fort Chilcotin, located near the junction of the Chilko and Chilcotin rivers, was sporadically occupied in the period ] 829-1844. This post never produced profitable returns leading to its abandonment in 1844 (Anonymous n.d.). Connolly who May 18, 2003-49 visited the Chilcotin in 1825 and 1829, found the Chilcotin and Carrier at Fort Alexander in serious dispute, perhaps because the Carrier were obtaining arms via the fur trade. Whatever the reasons, overall Chilcotin participation in the early 19th century fur trade was minimal. It may be that it also conflicted with their established role as middlemen between the Canyon Shuswap and the Bella Coola. The Chilcotin were very heavily impacted by the smallpox epidemics of 1863 and 1864. The Canyon Shuswap were devastated by the disease and the Shuswap largely abandoned the west side of the Fraser (Figure II-4). This resulted in the Chilcotin moving down the Chilcotin River to occupy the mouth with the Fraser. Thus the Toosey Indian Reserve was granted to a group of Chilcotin at Riske Creek in what had been Canyon Shuswap territory, as well as at the Anaham Reserve at Alexis Creek, which was previously an unoccupied zone between the Chilcotin and Shuswap territories. The less sedentary and more highly mobile Chilcotin were not as heavily affected by the smallpox as their neighbors, but this epidemic did help set off the Chilcotin War in 1864. The war began by an attack by Chilcotin on Alfred Waddington' s men in the belief that the Homathko Canyon road builders had deliberately spread smallpox (Hewlett 1973). Waddington was building a road up the Homathco River as an alternative route to the Caribou gold fields which were discovered in 1858. So the Caribou Gold Rush could also be seen as a factor in the Chilcotin War. For some three months the Chilcotin War raged, including one episode at Eagle Lake where Donald McLean, a member of a volunteer party from the Cariboo, was killed. In the end, some six Chilcotin were bound over for trial and five were hanged. May 18, 2003-50 50 { 50 Smallpox was apparently devastating to the Chilcotin at Anahim Lake (Wilmeth 1978:8) and after the war that area was essentially abandoned. The name \"Anaham, Anahim, anaxeim\" was used by a series of 19th century leaders (Wilmeth 1978:5,6) and the present reserve of that name near Alexis Creek is probably named after one of them, and likely consists of Chilcotin who had previous lived near Anahim Lake. The Anahim Lake area was then soon occupied by Ulkatcho Carrier, and a reserve was established for them there in 1916 (Wilmeth 1978:11). The group probably located furthest east prior to the smallpox epidemic was named after Alexis Creek (Figure II-4) and was moved west to Redstone. The combination of the abandonment of the Anahim Lake region and the movement downstream meant a major post-1860 movement toward the east by the Chilcotin to their present distribution. F IGURE Figure 11-4. Chilcotin Territory and Reserve Locations Crown land began to be granted to individuals starting in 1860 with a number of \"Certificates of Improvement\" granted for the Chilcotin territory beginning in 1870 (f yhurst 1984). These were first located in the Riske Creek and Alexis Creek areas, and then moved further west. With these grants ranching began throughout the Chilcotin plateau, and within 50 years had become the major social and economic force in the area. In 1887 three reserves were granted to the Chilcotin, the present day Anaham, Toosey, and Stone Reserves, and in 1909, the Alexis Creek (Redstone) and Nernaiah May 18, 2003-51 Reserves (now Xeni Gwet'in First Nations). As the non-Chilcotin ranching economy developed, the Chilcotin began to shift from the traditional economy, and the identity of many Chilcotin became that of cowboys. In this area, one can not graze cattle through the winter, so a source of natural hay that could be cut and stored to feed the stock was critical. Thus, many of the iand grants included natural wet areas that produced hay commonly called ilhay meadows\" and haying became an important activity. The livelihood of well-known Chilcotin such as Eagle Lake Henry was based on access to hay. In Eagle Lake Henry's case, the lowering of Eagle Lake water level during the 1940s made the western portion a large swampy area and gave him the ability to produce a large amount of hay and relative wealth (Per. Com. Scotty Shields, 1983). In contrast to the participation of male Chilcotin in the ranching economy, both Anglo and Indian, the participation of Chilcotin women in wage work has been minimal (Magne and Tyhurst 1984). A reflection of historically unequal participation by men and women in wage labor may be seen in the fact that whereas almost no Chilcotin men are unilingual, in the 1970s and 80s there were a number of older Chilcotin women who spoke only Chilcotin, or no more than a few words of English. Contact-era division of labor was based on women processing goods such as fish, meat, and skins, and, for the most part, upon men obtaining these things. Women obtained and processed plant foods and fibres, including the gathering and weaving of spruce roots and bark for baskets. There appears to have always been a great deal of overlap in knowledge between men and women in traditional Chilcotin society. Knowledge of the ways of doing things was widely shared, though men might know May 18, 2003-52 more about salmon weir construction, for example, while women knew more about processes such as hide preparation. The remoteness of the area as weli as the lack of female involvement with wage labor has facilitated the retention of traditional life-style. One example is that of hide\u00C2\u00ADworking technology as equipment in Nemiah Valley observed in 1983 included hidc\u00C2\u00ADscrapers, hide-smoking pits, hide-smoking frames and hide-stretching frames. Two of the hide-scrapers bear retouched stone spalls with considerable edge-rounding from use. The following chapter by Burnard-Hogarth describes other aspects of traditional technoiogy observed in her ethnoarchaeological research. The Chilcotin people of the 1980s were ranchers, or would wish to be, but were kept from entering ranching or from expanding their ranching activities by a chronic and serious shortage of pasture and arable land. Chilcotin men have worked, and continue to work, as wage laborers on ranches in the area. There was,- however, relatively little participation in the modern industrial economy of the area, which for the most part, is the forest industry. In the past many Chilcotin men acted as drivers for freight wagons, or owned their own teams which they hired out. The coming of roads within the past 45 years, and the widespread ownership of trucks and automobiles has decreased this role to one of relative unimportance. Ethnographic approaches to the Problem of the Origin of the Chilcotin Very early it was recognize d that a core aspect of understanding the Chilcotin was their origin and separation from other Athapaskan or Dene speakers. This was recognized even by Morice, writing about the full range of Athapaskan speakers: May 18, 2003-53 \"An ethnologic problem which is not yet, and will perhaps never be solved, is the question, How did it come to pass that large portions of the Dene nation detached themselves from the main stock and migrated south? When did this exodus occur? What was the route followed by the adventurous bands? ... Two facts only seem pretty safely established, namely: The separation of the southern from the northern tribes happened centuries ago; and, moreover, the national movement resulting in the d ivision of the nation into two different camps was from north to south\" (Morice 1893: 12). Almost all ethnographic descriptions of the Chilcotin since then have added statements about this issue (Farrand 1898; Teit 1909b: Lane 1953,1981; Tyhurst 1984). All appear to be in agreement that the Chilcotin are relatively recent newcomers in their current location. Also, there appears to be agreement that they share many features with the Carrier and probably separated from them relatively recently. In particular, the Ulkatcho Carrier appear to be very similar, and are dose neighbors (Tyhurst 1984:158-166). If one accepts glottochronology /lexicostatistics (which we discuss later in this volume), Carrier and Chipewyan diverged between 628 and 864 years ago (Hoijer 1956; Hymes 1957), and this diversion likely occurred in or near northern British Columbia. According to Davis (1975: 624), lexicostatistics indicate that the Chilcotin split from the Carrier 601 years ago, and 889 years ago with the Navajo, and a similar date for divergence with Chipewyan. Tyhurst (1984:348-354) shows that the relationship between Ulkatcho and Chilcotin kin terms is closer than that obtained by J. M. Kew for between Southern Carrier and Chilcotin, indicating that the Carrier divergence estimate Mav 18, 2003-54 .I is probably a maximum estimate, overestimating the separation date between the Ulkatcho Carrier and the Chilcotin. Thus 500 to 800 years ago the Chilcotin began to diverge from other Northern Athapaskans. This event probably occurred in northern British Columbia. Dyen and Aberie's (1974) study of Proto-Athapaskan kinship through iexicai reconstruction sheds some light on Ulkatcho relationships to Lower Carrier, but none o n Chilcotin because of the lack of data. Their study makes clear that Ulkatcho kinshipterm innovations are a more recent development than those found in the Bulkley River Carrier (197 4: 348). They also hypothesized that the Ulkatcho terminology and social organization changed because they occupied a region with poor salmon availability, were less sedentary, and thus eventually lost the matrilocality and matrilineality of Proto-Athapaskan (Dyen and Aberle 1974: 413-414). If the Ulkatcho and Chilcotin were until relatively recently one group, as indicated by Tyhurst' s analysis, this likely applies to the Chilcotin as well. All in all, we can be reasonably certain that the Chikotin are recent migrants to south-central B.C. (Lane 1953, 1981; Teit 1909b; Farrand 1898). In sum, Chilcotin life-styles of the present can be seen as a Northern Athapaskan adaptation to the more arid environment of the south-central Plateau, with good access to salmon along the Chilcotin and Chilko rivers. As with many northwestern Athapaskan peoples, seasonal resource use involves considerable distances and fairly reliable resources, but focuses to an unusual extent on the reliable, annual salmon migrations. May 18, 2003-55 Ethnoarchaeological Investigations in the Chilcotin Linda Burnard-Hogarth During the months of July and August of 1983, ethnoarchaeological investigations were carried out in the Chilcotin region of British Columbia. The research, funded by a grant from the B.C. Heritage Trust, was proposed and implemented in conjunction with the Eagle Lake project. The objectives of the study were threefold: L To aid in preserving Chilcotin cultural heritage. 2. To gather information on surviving aspects of traditional Chilcotin subsistence\u00C2\u00ADsettlement strategies which might aid in the interpretation of archaeological data. 3. To use the Chilcotin case as an example of non-territorial land use strategies toidentify material indicators of this form of adaptation in order that similar forms might be more readily identified in archaeological situations. A parallel, comparative study was carried out on the Skeena River for the Gitskan, a territorial group, and the results of the two studies are reported in Burnard (1987). Territoriality, as used here, refers to a concept of land ownership coupled with a practice of restricting access to resources and/ or resource locations. The development of this concept is particularly interesting because it appears to be intimately related to the evolution of complex society. The environmental and social mechanisms facilitating the development of territoriality among hunters and gatherers are a matter of theoretical and methodological controversy. Archaeology, with its temporal and developmental perspective, should be able to contribute to an understanding of this problem, but as yet little consideration has been given to territoriality's material, and hence archaeologically May 18, 2003-56 discernible, consequences. Towards this end a number of hypotheses were advanced which predicted material differences between territorial and non-territorial systems. The predicted characteristics of a non-territorial system were to be tested against data derived from the summer's fieid work. According to ethnographic documents: \"The Chilcotin had a very vague concept of land ownership. All the Chilcotin had right to use all the Chilcotin territory. Bands occupied vaguely defined geographic areas. Thev did not \"own\" such areas ... In most border areas there were indefinite ., zones that were used by both the Chilcotin and neighboring groups\" (Lane 1953: 173 -174). The hypotheses are as follows: 1. The number and elaboration of facilities and site furniture present at a particularsite will be less under non-territorial as opposed to territorial conditions. In non\u00C2\u00ADterritorial situations this is expected to be reflected by evidence of minimal energy investment in the construction of site furniture coupled with the use of materials locally available or easily replaceable. 2. Where access to particular locations is restricted it becomes feasible to storepersonal equipment at, or near, its locale of use. With unrestricted use of an area, as is thought to be the case in the area of the Chilcotin, it is predicted that gear will be May 18, 2003-57 curated or cached away from its immediate locale of use. Equipment left at Chiicotin sites should therefore have a iower replacement cost than might be expected in a territorial situation. 3. Factors influencing site placement will differ between territorial and non\u00C2\u00ADterritorial systems, particularly within the category of residential sites. In a non\u00C2\u00ADterritorial system, locations with abundant or concentrated resources are expected to be exploited by multiband units since these areas wouid permit more efficient resource procurement and access would be unrestricted by concepts of privatization. In such instances, primary residential sites would probably be located where secure, but less concentrated secondary resources could be most efficiently exploited. In a territorial system, where access to particular resource areas is restricted by those with ownership rights, primary residential sites are expected to be located near the most abundant and important resources as a means of exercising control. (vVhile residential patterns in the Chilcotin region were undoubtedly altered by the institution of the reserve system, it is hoped that relevant data on the social units involved in resource exploitation at particular locations can still be obtained.) Study Area Largely for logistical reasons the study focused upon the region around Eagle Lake. The Eagle Lake region continued in the early 1980s, as in the past, to piay an important role in the Chilcotin economy, even though it is home to only two to three native families on a year-round basis. Families from outside the area gather annually for Mav 18, 2003-58 ., the August-September sockeye and Chinook salmon runs on the Chilko River. The antiquity of this practice is attested to by a wealth of archaeological sites located in 1979 and 1983 during regional survey work. A second factor favoring Eagle Lake as the focus of ethnoarchaeological investigations is its former importance as a gateway to one of the major root gathering grounds in the Chilcotin - the Potato Mountain range, named for the beds of Spring Beauty (Claytonia lanceolata) or mountain potato, as they are known locally (Alexander et al. 1985: Alexander and Matson 1987; Matson and Alexander 1990). Such concentrations are rare in the Chilcotin, and consequently the area played a key role in the traditional subsistence round. While the edible corm of this plant is no longer a major food source, a number of local residents still make yearly gathering trips and many older adults remember making similar journeys as chiidren. Theoretical Background and :Methodology Archaeological interpretations of prehistoric lifeways are dependent upon a thorough understanding of cultural processes and the material consequences of a functioning society. It has long been recognized that living cultures provide a valuable foundation on which to develop and test such understanding. Archaeologists have repeatedly turned to ethnographic studies with such hopes in mind. Unfortunately, the material aspects of behavior are commonly overlooked by ethnographers. Part of the explanation for the inadequacies so often manifest in ethnographic descriptions of material culture lies in the fact that detailed quantitative data such as are required by archaeologists, are time consuming and potentially disruptive to acquire. The presence May 18, 2003-59 of an outsider, armed with a tape measure and camera, diligently recording such apparently inconsequential detaiis such as the circumference of support posts, the structure, dimensions and contents of hearth features, and the content and location of trash disposal areas, is bound to be a little unsettling even for the most tolerant of individuals. For this reasE>n, two distinct approaches to data coliection were employed in this summer's work. Qualitative data were sought through informal discussion with local residents and through observation and participation in contemporary resource procurement strategies. Quantitative data on camp structure and organization were acquired separately, and whenever possible, during periods of site abandonment. Camps and resource procurement sites from previous years were located and recorded prior to their use in 1983. Recording procedures entailed the drawing of scale maps showing existing facilities, site furniture, and activity areas. Structural details were sketched and/ or photographed with particular attention being paid to the types of materials used in construction. Copies of this field data are curated at the Laboratory of Archaeology, University of British Columbia. Later in the summer of 1983, when the above locations were in use, many of the sites were revisited and relevant social data were gathered. These included the number of individuals present, location of permanent residences, length of stay, etc. In addition, newly constructed facilities were noted for later recording when the sites were again vacated. Earlier interpretations regarding activity area function were also checked at this time. Results May 18, 2003-60 It had been hoped that hunting and root gathering camps and procurement sites would be recorded in the course of the project. Residents described a number of such locations situated in the alpine areas of the Potato Mountain Range. Unfortunately, exploitation of this region did not occur during the study period, rendering observations on the active use of these areas impossible to obtain. Later, though, RobertTyhurst and Diana Alexander carried out ethnoarchaeological and archaeological investigations on Potato Mountain (Alexander and Tyhurst 1985) and Matson and Alexander extended the archaeological investigations (Alexander and Matson 1987; Matson and Alexander 1990). The following site classes were recorded during the summer: 1. Kokanee fishing camp (1);2. Trout fishing camps (2);3. Salmon fishing camps (8);4. Fishing stations (5);5. Unknown function camps (2).As the list suggests, salmon fishing camps outweighed any other site class. This bias is a result of both the choice of study area and the season in which the research was conducted. The Chilko River, above its confluence with the Taseko River, is an exceptionally clear river. Dip netting under such circumstances is ineffective since fish are able to see the net and thus avoid it. The primary technique for taking salmon on the Chilko River at present is with a gaff. There are a limited number of places along the river where conditions are particularly favorable for the use of gaffs. A number of these May 18, 2003-61 F 1 ying racks near River fall within a two mile radius of our base camp, hence the high density of salmon fishing sites in this area. A variety of structure types and facilities were recorded at these sites. The function and range of variability within each class are described below. Drying Racks Drying racks (Figure II-5) represent the most numerous and complex of all FIGURE Figure II- 5. Chilcotin drying racks at Henry's Crossing. May 18, 2003-62 Rack Length Average Width Average Height Rack Supports 1 3.3m 1.7m ,. '7i.,m 4 living trees 2 1.6 4 living trees 3 1.2 1.5 2 living trees, 2 cut poles 4 3.3 1.6 1.6 4 cut poles 5 4.4 1.6 2 living trees, 2 cut poles 6 3.8 1.4 1.3 1 living tree, 2 cut poles 7 4.5 1.8 1.5 2 living trees, 2 cut poles 8 1.7 3 living trees, 1 cut poles 9 4.6 1.6 2 living trees, 2 cut poles 10 5.1 1.7 1.6 2 living trees, 2 cut poles 11 3.0 2.0 1.6 2 living trees, 2 cut poles 12 1.6 3 living trees, 1 cut pole 13 1.5 3 living trees, 1 cut pole 14 4.2 .9 1.5 1 living tree, 3 cut poles 15 3.9 .7 1.2 2 living trees, 2 cut poles 16 3.6 1.6 1.4 2 living trees, 2 cut poles 60 Table II-1. Chilcotin Drying Racks (After Table 5.1, Burnard 1987) permanent structures in the sample. Examples were recorded at salmon, kokanee and trout fishing camps. These encompassed a considerable range of variability. This May 18, 2003-63 variability is thought to be attributable to environmental considerations rather than functional requirements. Wherever possible living trees were used as supports for the horizontally suspended poles which formed the rack upon which a number of spanning stakes might be rested. Where living trees in the necessary configuration were not available, cut poles were used as alternate supports. A number of drying racks exhibited structural modifications designed to deal with inclement weather conditions, including brush windbreaks and tarp supports. Poles used in drying rack frameworks were most commonly lodgepole pine although the occasional trembling aspen pole might be used if pine was in short supply. Spruce was avoided even where trees of a suitable size were abundant. One individual explained that spruce was undesirable because it is so sticky. In no instance were the poles used in the framework of the rack observed to have been peeied before use. This is in contrast to the spanning stakes, which were almost always peeled. Spanning stakes were sometimes left on the drying rack or piled nearby, but often they were missing entirely. Whether they had been destroyed after use or curated is not known. In one instance a cache of spanning poles was discovered some distance back in the bush away from the campsite. This may well have been the case in other instances as well. Where branches were not available on which to rest structural poles, a range of materials was used to fasten the drying rack framework together. This included wire, bailing twine, strips of rawhide and willow bark. In one case lengths of electrical cord were used. Nails or spikes were present in only two out of 19 drying racks. Dimensions varied from just over 2.5 m to over 5 m in length, and from 65 cm to 1.75 m in width. Height was the least variable dimension. All but two of the 16 drying May 18, 2003-64 racks measured for height fell between 1.5 m and 1.8 m. Often this range of variation was encompassed within a single structure as a result of undulations of the ground surface. The lack of height variability was likely the result of raising the salmon enough off the ground so the smudge fire would protect the salmon from insects, but low enough so the salmon on spreader poies couid easily be moved around. Cross Bars One or more horizontally suspended poles were present at almost every campsite. These were of a variety of lengths and heights above the ground. As with the drying racks, wire and/ or bailing twine were the most common means of attachment, and living trees were used as support posts. Functionally these structures were multipurpose. Recorded uses ranged from food caching, hitching posts, airing bedrolls and clothing, and hide stretching rack supports. Again, the archaeological detection of these structures is unlikely. Shelters The standard shelter used at contemporary Chilcotin fishing camps is the canvas wall tent. While the tent itself is transported after use, presumably back to the residential base, the ridge poles, crossed uprights, side poles and pegs or anchor stones are left behind. The remains of mat flooring constructed of evergreen boughs are another common feature of most tenting areas. While spruce is the preferred material for this, pine boughs are also used. The charred remains of a mat floor recovered from the 1983 excavations at the Bear Lake site suggest that this practice, at least, has May 18, 2003-65 e considerable antiquity. No particular arrangement of tenting areas relative to other activity areas was discernible, nor were the same locations necessarily reserved for tents on successive occupations. Tables /Benches A common feature of most, but not all, campsites were table /benches (the low height of these structures enabled them to function in both capacities). These usually consisted of a plywood or board surface supported by a number of posts which had been hammered into the ground until their tops were level. Table surfaces were sometimes curated and often only the support posts remained after abandonment. Again, living trees were incorporated into the structures as supports, whenever they were available. Hide Processing Structures Structures associated with hide processing activities were present at a number of camps. Most common among these were the hide stretching frames, (Figure II-6) of which a total of four were recorded. These rectangular frames were FIGURE Figure II- 6. Chilcotin hide stretching frame (1983). constructed of unpeeled pine poles between 5 cm and 6 cm in diameter and from 1.5 m May 18, 2003-66 to 2.5 m in length. The structurai poles were notched at each end, fitted together log cabin styie, and nailed at the corners. All frames had been left in their position of use, upright and resting at an angle against two trees or horizontal support poles erected for that purpose. Dehairing/Defleshing Poles Dehairing poles were present at two camps. Both poles were made of aspen, approximately 25 cm in circumference and sharpened to a point at one end. During the dehairing process the point end was leaned against a tree and the hide was draped over the peeled, upward facing surface of the pole. The point of contact between the pole and the support tree was marked in one instance by a slight oblong scar 1.58 m above the ground. In the second instance bark had been deliberately shaved from the supporting tree, leaving a blaze-like scar. The base of each dehairing pole, in the course of use, had sunk some 7 cm into the ground, but once again the resulting surface disturbance is unlikely to be sufficient or distinctive enough to permit archaeological detection. Hide Smoking Frames While two types of hide smoking frames were observed during the summer, only one type was represented in the site sample. Two examples were recorded, both in a disassembled state and were only identified with the assistance of a Chilcotin friend. The remains consisted of 5 - 6 peeled willow stakes. I was told that these stakes were assembled in a miniature \"teepee-like\" structure around which the hide was wrapped for smoking. As the observed structures would not have been large enough to support May 18, 2003-67 swea a moose hide, I surmise that the length of the stakes would vary according to the size of the hide. Their thinness (less than 2 cm in diameter), the brittleness of the wood when dry, together with the casual manner in which the structure was disposed of after use, suggest that these were considered to be temporary, single use structures. A second type of hide smoking frame observed in a residential context was treated quite differently. It had been carefully stored (strung up in a t ree), in a fully a ssembled state. Furthermore, its construction was considerably more complex than that of those at the fish camps. It was shaped like a cylindrically domed bird cage rather than like a teepee, and was circumscribed twice by willow withes which lent further structural support. These observations are of some significance to the archaeologist who wishes to \"anticipate certain regular contrasts between archaeological assemblages deriving from highly organized and curated technologies, versus those that are poorly organized and tend toward the expedient manufacture, use and abandonment of instrumental items in the immediate context of use\" (Binford 1979: 341). The Chilcotin hide smoking frames suggest that both curated and expedient technologies for performing a specific task may coexist within a single cultural system. In such cases, selection of the most suitable technology will depend upon the context of the task. Sweat Lodges Sweat lodge structures were recorded at o nly one site (Figure II- 7). FIGURE May 18, 2003-68 Figure II-7. Chilcotin sweat lodge frame (1983). Although the remains of several older sweat lodges were present at this particular camp, only one was operational at the time of observation. It consisted of a framework of unpeeled wiliow wands covered by a canvas tent, the sides of which had been rolled up to permit air circulation. The framework stood 1.42 m high, was constructed of 16 lengths of willow fashioned into 8 arches and placed in a nested, dome-shaped arrangement. These enclosed an area 1.36 by 1.62 m, the distance between individual wands at the base ranging from 50 to 63 cm. The willow wands (2.3 to 3 cm thick at their bases), were sunk into the ground to a depth of about 7 cm. Pebbles had been wedged around the base of a number of these to further secure them. The floor of the sweat lodge had been covered with a mat of willow leaves and clover flowers. A rockpit was situated inside the structure close to the north wall. At one end of this pit, aligned with the support poles, a peeled pine stake 1.05 m long and 3.5 cm in diameter, was embedded in the ground at an inward-leaning angle. At the other end of the pit a small hole suggested the former presence of another such stake. A likely candidate lay on the ground outside. The sweat lodge was situated about 12 m from the Chiiko River. The main camp area was at least 100 m away, back from the river. It seems likely that a desire for ready access to water following the sweat bath was the primary reason for the physical separation of the two areas. The remains of two other sweat lodges were found nearby, at eight and nine meters from the water. May 18, 2003-69 Fishing Platforms Fishing platforms were constructed at three of the five fishing stations. One of these was washed out during the summer however, and no attempt was made to replace it although the location continued to be heavily used. The platforms rested on log supports pounded into the river bottom. Horizontal support beams stretched from these to the bank to which were nailed board decks. None of these p latforms extended more than three meters into the river. Cultural Depressions Two types of pit features were recorded at the various campsites: pits or basins used to contain heated rocks in the sweat lodges; and hide smoking smudge pits. Each had a number of distinctive characteristics which could aid in their archaeological identification. Although the sample is small, smudge pits (Figure II- 8) appear to be smaller and deeper than the sweat lodge rock pits. FIGURE Figure Il-8. Chilcotin smudge pit (1983). Sweat lodge pits are usually about 60 cm in diameter and about 25 cm deep, while hide smoking pits are usually about 45 cm in diameter and 40 cm deep. Moreover, while two of the smudge pits had been cleaned out after use, all four of them contained evidence of burning. The sides of the smudge pits were sometimes discolored and charred pine May 18, 2003-70 cones or at least traces of them remained in the bottoms. The sweat lodge pits on the other hand, showed no evidence of burning although a number of the rocks contained in them were fire fractured. The hearths used to heat these stones were situated nearby, from 3 m to 9 m distant. Hearths Since hearths are among the most commonly encountered archaeological features it was encouraging to discover that structural and functional variation could be identified within this class of remains. While three functional hearth classes were examined, only two examples of the Sweat Lodge hearths were located. Hearth No. Size Range Average Partial or Complete Preferred Function Sample Size Boulder Perimeter Fuel Drying Rack 53 .21 to 1.69 .74 26% Punky wood or Green Popiar Kitchen 13 .78 to 1.97 1.34 92% Pine Sweat Lodge 2 2.51 to 3.45 2.98 Disorganized Concentration of rock Table II-2. Hearth Characteristics (after Table 5.2, Burnard 1987) Kitchen/ general purpose hearths These hearths functioned not only in the context of food preparation but also as a focus for socializing. They ranged in size from slightly less than 1.0 m2 to 2 m2 with the May 18, 2003-71 average being about 1.34 m2\u00E2\u0080\u00A2 Twelve out of thirteen kitchen/ general purpose hearthsh ad boulder perimeters with a high proportion of the constituent rocks being fire\u00C2\u00ADcracked. Where the fuel source was determinable, it generaliy proved to be pine. Drying rack hearths In contrast to the kitchen hearths, drying rack hearths (Figure II-9) FIGURE Figure II-9. Drying rack and hearth near Henry's Crossing. were generally smaller, averaging about 0.74 meters and usually lacking boulder perimeters. There were exceptions to this rule, however, and 14 out of 53 recorded drying rack hearths had at least a partial boulder perimeter. Significantly, very few of the constituent rocks were fire cracked. Rotten or punky wood and green poplar were the most commonly used fuels although informants say that alder wood is preferred when available, but it is not abundant in the study area. Fireweed stalks are also said to be used on occasion. Many of the observable differences among the three hearth classes are understandable in terms of varying functional requirements. The main purpose of adrying rack hearth is to produce smoke rather than heat. Temperatures sufficient to cause fire fracturing in rocks therefore, rarely occur. Kitchen hearths on the other hand, function primarily as sources of heat, hence large quantities of associated fire cracked May 18, 2003-72 rock Functional differences are also evident with regard to fuel selection. Smoky, slow burning fuels were characteristic of drying rack hearths while cleaner, hotter burning pine was preferred for kitchen hearths. Sweat lodge hearths probably required even higher temperatures and thus were larger. Informal observation and participation in archaeoiogical field crew ''sweats\" supports the higher temperature idea, as large fires (and thus, large \"hearths\") were required to heat the rocks to an orange color for good sweats. Interestingly, almost all of the cobbles used in the ethnoarchaeologicaliy observed sweat lodge pits consisted of vesicular basalt. While this material is not uncommon in the area it is by no means the most readily attainable stone. Informants later explained that this type of rock was preferable because it held the heat better. Observations during archaeological field crew sweats tended to support this statement and also demonstrated that the porous basalt tended to resist fracture due to rapid cooling better than other stones. There has been a tendency among archaeologists to overlook the information potential represented by features such as hearths and to focus their analytic efforts on associated artifact assemblages. In pointing out the above sources of variability, I do not mean to imply that the observed patterns necessarily reflect the archaeological situation. Modern cooking vessels and wood cutting implements have probably had a profound impact on hearth structure and size. The point I do wish to make is t hat structural variation resulting from functional differentiation can be identified within this class of remains. As was mentioned earlier, hearths sometimes provide the only clue as to the former existence of extensive, but archaeologically invisible structures such as drying racks. As such, hearths and other features merit far more attention t han has often been May 18, 2003-73 accorded them. Archaeological Implications I suspect that a number of the structures described above resemble quite closely their prehistoric counterparts and it is with this in mind, in aid of the Direct Historic Approach, that the above descriptive information has been presented. At another level, however, as an illustration of behavioral principies reflected in a set of material remains, the resemblance or lack of resemblance of these camps to their archaeological precursors is inconsequential. Regardless of their similarity to local archaeological remains, the contemporary Eagle Lake sites serve to challenge a number of basic assumptions made by archaeologists. One of these is that the nature of the artifactual remains at a site should in some way reveal its specific function. According to Binford (1980: 10), a field camp, \" ... is where a task group sleeps, eats, and otherwise maintains itself while away from the residential base. Field camps may be expected to be further differentiated according to the nature of the target resource so we may expect sheep hunting field camps, caribou hunting field camps ... \" Table 11-3 below is a list of site furniture at a typical Eagle Lake area salmon fishing camp. Perusal of this list and the above structure descriptions reveals that for these sites at least, there is actually very little in the nature of the remains which might serve to identify their specific function. This is in spite of the fact that most of the fish processing activities were carried out on location at the camps. The presence of drying racks is not May 18, 2003-74 in itseif indicative of fishing activities since these structures are also constructed at hunting camps for drying meat. If identification of function is difficult in situations where the entire range of organic as well as inorganic remains are present for examination, how much more difficult can we expect it to be when the remains are archaeological? Field camps may exhibit considerable homogeneity in spite of their special functions. This is in part because maintenance requirements which are responsible for the bulk of the remains generated at a camp, are likely to vary more with group composition and seasonality than with the specific resource procurement activity. Drying racks Drying rack spanning stakes Cross bar Tent support poles Table /bench structure Log sections used as seats Willow barbecue stakes Roasting spits Sticks for removing pots from fire Bundle of willow branches Pile of pine and spruce boughs for tent flooring Empty 5 gallon oil can Several empty jars Aluminum cooking pot - damaged but functional Inner tube section Bailing twine Several lengths of wire Empty bottles Assorted empty cans, many crushed and burned Table II-3. Contents of a typical abandoned salmon fishing camp. Staging in both manufacturing and processing of materials may produce Mav 18, 2003-75 ., incongruities between the nature of the assemblage and the resource procurement activities actually performed at a particular site. It is particularly disturbing to realize that if site furniture and/ or faunal remains were taken as indicative of site function, at least two of the nine salmon fishing camps might wrongly be identified as hunting camps. Hide processing structures were present at a number of sites while the game they represented had been taken earlier at other locations. The length of time involved in hide processing results in various other resources. In only one of the abandoned camps were any gaff hooks observed and only two had gaff poles. One of these was broken and had been incorporated into the framework of a drying rack, the other was too short to be effective, its tip having been repeatedly snapped off. Gaffs and gaff poles together constituted the primary procurement equipment in use at the fishing sites. While the camps were occupied both items were very much in evidence. The degree of curation observed with regard to procurement and processing equipment demonstrates that the frequency with which a particular artifact occurs in an assemblage is not necessarily indicative of its importance. The latter two points have been made by Binford (1978a; 1979) in another context. The Chilcotin data demonstrate that these are not localized phenomena. Archaeologists commonly make decisions about where to place excavation units on the basis of surface artifact densities. Delineation of site boundaries similady depends on artifact concentrations. It was therefore somewhat disturbing to discover that for most Chilcotin camps the highest densities of inorganic remains rarely corresponded to the areas of intensive use. In fact, there was often an inverse relationship between the two! Heavy concentrations of cultural remains were invariably May 18, 2003-76 found in scattered dumps around the camp perimeters. Contrary to being zones of intensive use, these locations appear to have been selected precisely b ecause they were removed from the main living area. The above observations serve to illustrate the heuristic potential of ethnoarchaeologicai studies for examining methodological and analytical assumptions that are commonly made by archaeoiogists. Conclusions Any conclusive statement regarding the project's third objective of identifying material indicators of a non-territorial land use strategy would be premature at this time. Comparative information of territorial systems is still being gathered. The Chilcotin data may, however, be cautiously examined in the context of the three hypotheses stated earlier. 1. It was predicted that site furniture at Chilcotin sites would reflect the minimal levelof energy investment necessary for those items to adequately perform their intended task, and that locally available or easily replaceable construction materials would be used wherever possibie. The reasoning behind these expectations was that it would make little sense to invest time, energy and valuable resources improving a location if future access to those improvements were uncertain. Chilcotin sites recorded in the study do in general conform to these expectations. It is admittedly difficult to assess what constitutes a minimal level of energy investment, but a number of variables suggest that this was, in fact, the case. Natural features, particularly living trees, were incorporated into the design of a structure May 18, 2003-77 wherever possible so that the amount of construction was kept to a minimum. Poles from old structures were salvaged for use in new ones, and fixed attachments (indicators of permanence) were minimized. While structures might be used for number of successive seasons, extra effort was rarely invested to ensure an extended use-life.As predicted, locally available logs and poles were invariably the primary construction materials. While imported materials were present these were almost always recycled items which were no longer useful in their original context. Empty coffee containers became billy cans, scrap metal from an old car might be used as a campfire grill. Furthermore, the recycled materials were usually items which were continually being brought into the system and were therefore easily replaceable, such as bailing twine, containers, and inner tube fragments. 2. Equipment left at Chilcotin sites was expected to have a iow replacement cost. Itwas reasoned that unrestricted use of an area, characteristic of non-territorial systems, would discourage people from leaving behind valuable or personal equipment. A list of equipment used at the various camps visited during the summer includes tents, tarps, bedrolls, food preparation equipment, clothing, Coleman lamps, knives, guns, hide processing tools and fishing equipment in the form of gaff poles and hooks. None of these items were left at the camps after they had been abandoned. This is not surprising since most of these items are used in other contexts as well. I was, however, puzzled at the lack of functional gaff poles left at the abandoned fish camps. (Archaeological field crews, though did locate a number during survey, but not necessarily in obvious places.) Their absence was surprising for a number of reasons. First, it seemed unlikely that they could serve any useful function away from these May 18, 2003-78 locations and secondly, their structure seemed so rudimentary that I had assumed that their replacement cost would be low. Inquiry into the manufacture of these poles revealed otherwise. It was repeatedly emphasized that shaving a pole down to a suitable thickness took a great deal of time. Furthermore, before a pole was usable the wood had to be seasoned. Green wood, it was explained, was too heavy and flexible. In short, the energy investment involved in the manufacture of a gaff pole was considered sufficient to warrant transporting it back to the residential base. The caching of peeled drying rack spanning stakes and the curation of table tops lend further support to the idea that portable items with more than a minimal labor investment, or with more than a minimal replacement cost, will not be abandoned. 3. The final hypothesis to be considered concerned the nature of the social unitsexploiting particular resource areas. It was predicted that where access to areas of highly concentrated resources was not restricted by concepts of private property, as was believed to be the case among the Chilcotin, such areas would be exploited by large, multiband units. An effort was therefore made to obtain information on the band affiliation of those groups fishing in the Henry's Crossing area during the study period. The results indicated that a number of bands were represented, principally the Nemiah and Redstone bands. At least one Carrier individual from Anahim Lake also fished at this location. Further data were collected on the use of specific campsites and fishing stations. When queried on this subject most individuals said they preferred to return to the same campsite year after year but if another group happened to precede them to that location they simply camped somewhere else. Of a sample of twelve camps, three were May 18, 2003-79 established in locations previously used by the same group, two camps were new, three camps were in locations used earlier by other groups, and four former campsites were not reoccupied at all during our stay. Access to fishing stations was similarly unstructured. Members of almost every camp were observed fishing at the three main sites. A fisherman might divide his efforts among three or more fishing stations in the course of a single day. In short, exploitation patterns in the study area did seem to reflect a principle of unrestricted access to resources, and as predicted, the social units involved in saln1on procurement did exhibit multi-band affiliation. These preliminary attempts to identify material indicators of territoriality do appear encouraging. It is interesting to note that the expectations discussed here regarding the types of material patterning generated by a non-territorial land use system received support from the Chilcotin data, and differed significantly from that found for Gitskan on the upper Skeena River, a territorial group (Burnard 1987). Although not all the factors discussed above aligned themselves as expected in this comparison, there was general support for the propositions reviewed above. Mav 18, 2003-80 .I Notes on Cambium-Stripped Lodgepole Pine Martin Magne Apart from occasional references in the ethnographic literature, I first became aware of the practice of removing cambium from coniferous trees through the report by Anne Eldridge (1982), who described living spruce and pine trees bearing evidence of cambium removal in the Liard River-Lower Post region of northern B.C. On a casual trip to Eagle Lake in the winter of 1982, several such trees were noted on the east end of Eagle Lake near to where the UBC archaeological field school posted camp, and I decided to make an effort to obtain data from these features in the future. Data were obtained on 24 stripped (or culturally-modified) trees in the Eagle Lake region in 1983, not including some 20 trees observed in Quadrats Gl and G6, but are found in five other locations. One stripped tree was observed at Quadrat 19:1 near the Chilko River, three in a small ravine on the east end of Eagle Lake, three at Canoe Crossing at site EkSa 5, three near the Bear Lake site, and 12 at about the midpoint of the east end of Eagle Lake at a location which will be called Henry's Camp. Another two were recorded at site EhRv 2 in the Taseko Lakes region (Magne 1984), and many more are present there. Given the activities of porcupines, bears, birds, disease and other factors, how do we know that these trees have been modified by humans? The most telling characteristic is the presence of cut marks on the inner bark, and often on the outer bark, of the trees. The cut marks are most often at breast height, usually form an upside down \"V\" at the top of the cut, are often cut around small branches on the tree, and often terminate the strip at the bottom of the cut (Figures II- 10 and II-11). Two trees May 18, 2003-81 e '*'\u00C2\u00B7 A ium-1 l Clo cut marks ium-str were observed in the small ravine area that had been stripped starting at a height of 4.5 meters above the ground. This is unusual, most strips starting at about 1 meter to 1.5 meters above ground level. FIGURE Figure II-10. A typical cambium-stripped lodgepole pine tree (1983). FIGURE Figure II-lL Close-up of cut marks on cambium-stripped tree (1983). Several variables were recorded for each tree, including cardinal orientation of the strip, length of the strip, width of the strip, elevation above ground of top of strip, elevation of bottom of strip, tree diameter, and number of visible cut marks. Two 3/8 inch increment cores were taken of each tree: one from the strip through the center, and another from the opposite side, through the outer bark to the center. It was intended to mount, sand and read these cores to provide age of the tree, age at stripping and thus age of the stripping activity, but this was not done. Marion Parker took one-inch diameter cores from the three trees near the Bear Lake site, to attempt seasonality readings from the changes in growth due to the trauma of the stripping, but this does not appear feasible due to the changes in growth as the tree attempts to heal. May 18, 2003-82 8 6 .. \u00E2\u0080\u00A2 \"\"-o -\"'-\u00C2\u00B7 '\u00C2\u00B0\u00E2\u0080\u00A2o\u00E2\u0080\u00A2 .... \"b 83 19:1 N= I Fl ELAP Comp RavineLJ N=3 Canoe Crossing N=3 N NE E SE S SW W NW magnetic orientation \u00E2\u0080\u00A2 Bear Lake N=3 Henry's Comp N= 12 EhRv 2 N=2 circumferential n:-\e-Figure Ant->. Histogram of orientations of stripped areas on lodgepole pine trees. All of the recorded trees are lodgepole pine (Pinus contorta), although one stripped trembling aspen was observed on the east side of Eagle Lake. Apart from the two exceptional examples noted above, the strips are between 40 cm and 140 cm in length, and between 8 cm and 17 cm in width (Maximum dimensions). Tree diameters (at a point half-way down the strips) range from 14 cm to 35 cm, and there is slight bimodality here, with a group of trees between 14 and 17 cm in diameter, only two between 17 and 23 cm in diameter, and 16 trees between 24 and 35 cm in diameter. Eleven of the 24 trees have strips oriented due north, another five are oriented northeast, two are oriented east, six are oriented southwest through west to northwest, and the two tall-stripped trees (which are each only 14 cm in diameter) are stripped completely around their circumference, almost like barber-pole designs. Thus the preferred orientation of the strips is north to northeast, with again slight bimodality, centering on west (Figure II-12). FIGURE Figure II-12. Histogram of orientations of stripped areas on lodgepole pine trees. A total of 92 cut marks were observed on the trees, with 50 of these occurring on north facing strips, 20 on northeast facing strips, and the circumferentially-stripped trees totalled about 20 visible cuts. There does not seem to be any association of cut mark orientation with the width or length of the strips. At present, it is not known during what season the stripping takes place, nor May 18, 2003-83 precisely why it takes place, although Lane (1953:43) clearly places it in the late winter/ early spring \"starvation\" period, and Teit (1909b:781) notes that lodgepole pine cainbium was \"much relished\". As noted above, some informant testimony has been obtained indicating that cambium is sometimes given to children. Obviously, the mere presence of the trees limits the time depth of a possible in-depth study, and it is unlikely that physical evidence of trees older than about 80 years could be obtained. Furthermore, recent pine-beetle infestations of pine forests of the interior are currently destroying nearly all pine trees, so the information will not be around for much longer than say 5 to 10 years. The only age information apart from that provided by Parker, was obtained from a standing but wind-broken tree at Henry's Camp. This tree was cut, the rings counted, and an age of 11 or 12 years was obtained for the cut, the tree being about 22 years at the time of the stripping. This matched informants' testimony that the site had been occupied about 10 years ago, when the cambium had been fed to children. Informal experimentation informed us that the lodgepole cambium is quite sweet in late spring, and loses its quality through the summer. Young trees appear to have sweeter cambium than older ones, though both of these observations cannot be taken as conclusive. One can only speculate at present that the reason for north-northeast preference for cut orientation is also related to the quality of the cambium. Although the information obtained was not conclusive, preliminary investigation of the stripped trees at Eagle Lake showed that yet another aspect of traditional subsistence practices can be studied in the region. Future research should attempt to more precisely date the specimens, obtain a larger sample, and interview informants for their knowledge of the practice. This should be related to other subsistence practices, May 18, 2003-84 since on their own the trees have actually not much to offer. In 1984 I had written that any future work should take place within the next five years before the trees are completely destroyed by the current pine b eetle infestation. Since that time not only have culturally-modified trees (CMTs) studies become a mainstay of British Columbian archaeoiogy but spreading beetle-kill has destroyed many of these trees. Visits to Eagle Lake in 2000 and 2001 to locate the trees recorded above resulted in failure, although some other CMTs informally located in the 1980s were found. From the state of these I conclude that the recorded trees have died, fallen, and their bark has fallen off. Concurrently, however, studies of CMTs have become a standard archaeoiogical practice in CRM archaeology which in interior B.C. is largely driven by the forest industry. Although most of the studies remain exercises in inventory, their potential to contribute to greater anthropological knowledge is best exemplified by Prince (2001 ). In a study of CMTs in the Nechako Plateau (west of Prince George, B.C.), Prince was able to demonstrate strong relationships between cambium usage and salmon availability. His research indicates that pine cambium was more important to northern interior B.C. diet than is commonly believed. May 18, 2003-85 Chilcotin Settlement Pattern In contrast to the Plateau Pithouse Tradition, not enough archaeological investigations had taken place in Chilcotin territory--or in neighboring Athapaskan areas--to generate a good model of the full settlement pattern (Wilmeth 1978; Donahue 1973,1977; Fladmark 1974, 1976; Borden 1952). Thus we were forced to reiy heaviiy on the ethnographic descriptions for this purpose. However, there is some information about expected archaeological artifactual differences with which we will end this section. As in the PPT, the core activity for the Chilcotin, where available, was late summer salmon fishing. Since the Chilko River is clear, unlike the lower Chilcotin, the main technique used in the Eagle Lake area would be harpooning or \"gaffing\" as it is referred to today. The Henry's Crossing area today is an important sockeye fishing area, which typically continues for about a month from mid-August. In the past, an eariier Chinook run may have also been important. These fishing camps and the drying racks which are very obvious today have been described in the previous section by Burnard-Hogarth. We assume that in the recent past they were very similar. In the past, though, the preserved salmon were stored nearby. Lane (1953:45-46) states \"Dried fish were often stored in the pit caches, which were concealed near campsites where the owners intended to camp at a later time.\" Thus we would expect near-river cache pits for later use for people who intended to camp later not too far away, and for families wintering nearby, cache pits near, but not at the winter sites. These locations are analogous to both the ravine cache pits found at the MOC, and what we inferred was a possible function of the ecotone May 18, 2003-86 cache pits there. In general terms, both forms of cache pit sites would be located relatively near the Chilko River. A variety of activities could have taken place between the salmon runs and the winter base camp. Lane (1981 :406) indicates that \"some went back into the mountains for more marmot trappingn1 others went north for caribou hunting, fished for kokanee, or prepared for the fall hunt. But by November people congregated at the winter camp area. It is clear that the Chilcotin winter base camp was smaller, occupied for a shorter duration, and w.as less likely to be regularly reused in following years than the PPT winter village. According to Lane (1953) only one or two families would winter in a single location. The structure(s) could be a small pithouse or the gabled lodge reviewed above. Lane (1953:46) states \"The gabled house, which was earlier among the Chilcotin than the pit house was considered to be an excellent shelter. Well covered with brush, bark, and dirt, it was warm; and yet, the opening extending the length of the ridge pole provided arnple ventilation and allowed light to enter. The fire, which was laid in the center of the house along the space under the ridge pole, had to be kept burning constantly to provide adequate heat. This took a great deal of firewood, and, if the camp was occupied for some time, maintaining a supply of fuel became a problem.\" Lane (1953:47) also gives information on the typical location, \"House and camp sites were selected with a view to concealment from enemies, a supply of firewood, and conveniently located fisheries or hunting grounds. Usually they were near lakes, but not actually on the shores. They were set back in the woods where they would be less conspicuous.\" This quote almost exactly describes the setting of the Bear Lake site, described in detail later on. There are clear distinctions from the PPT in size of winter May 18, 2003-87 settlement and location but the house forms used included the pit house as found in the PPT, as well as the distinctive gabled lodge. Although the main source of food would be that stored from earlier seasons, along with hunting, winter whitefish procurement was probably also an important activity. One has the impression that the winter base camp was usually abandoned by the end of February, with cambium stripping and hunting being activities until the (trout, suckers) fish began running when the ice left the spawning streams. The spring-time spawning runs at some locations resulted in relatively large aggregations. Hunting, fishing, root collecting and processing, carried out by smaller groups, occupied the later spring and early summer. The alpine areas were visited in the later summer for the mountain potato. The Potato Mountains, immediately south of Eagle Lake, are the best known alpine location and was visited by relatively large aggregations. In the late summer salmon fishing locations were again occupied. In terms of settlement pattern, then, there is broad overlap with what is predicted for the PPT. The main difference discernible through archaeology is the winter base camp. The location near lakes, but not on the shore, and the size, one or two families, is different from the multiple family winter pithouse villages, located adjacent to potable water, of the PPT. Isolated pithouses, though, are found in the PPT, and are relatively common in some areas, although not at MOC. The presence of an isolated rectangular structure or two in such a location discriminates between the PPT and Athapaskan occupations. Rectangular depressions are found in the PPT, and are often thought to be the remains of mat lodges used in cold weather. As mentioned before, these are usually May 18, 2003-88 thought to be relatively recent, perhaps related to the use of the horse and the changes in transportation that the horse allowed. So while the use of rectangular structures for the winter is distinctive for the Chilcotin, it is not unknown for the PPT. Rectangular lodges appear to be an old Athapaskan trait, with many Athapaskans using them from the middle of Alaska (Clark 1996) to the Chilcotin. As Lane states (1953:144-156) rectangular lodge variants are the most common northern Athapaskan winter dwelling type. Archaeologically, they were present at the site of Chinlac, near the confluence of the Nechako and Stuart Rivers (Borden 1952) and at Punchaw Lake (Fladmark 1974, 1976). Lane (1953:46,160), Ray (1939) and Teit (1909b) all agree that the pithouse was a recent introduction to the Chilcotin; of course, they may all be wrong. Interestingly enough, Lane's (1953:275) informants did not believe that the large pithouse sites were made by the Chilcotin or that the Chilcotin ever wintered along the Chilcotin/Chilko River. Here we again have a clear separation from the PPT in winter basecamp location and type. Given the limited information about the Chilcotin settlement pattern, the only obvious difference is in the likely most common type of winter dwelling prior to 1800, and its location. Even in winter dwelling form, though, there is overlap, in that pit houses were also used and relatively isolated small pithouses would be expected in both PPT and Chilcotin settlement patterns. The rectangular gabled lodge, though, is quite different from anything described as a winter domestic structure for the PPT. May 18, 2003-89 Distinctive Northern Athapaskan Material Culture What are the material culture items which might be archaeologically useful to distinguish Athapaskan from PPT components? From previous research Ludowicz (1980) and Magne and Ludowicz and (1984) identified five possible items. Projectile point styles, primarily the \"Kavik\" point type, \"tci-thos,\" spurred scrapers, a lithic technology utilizing a relatively high number of flakes, and the rectilinear winter lodge style of dwelling. Each of these will be discussed in turn. Distinctive stemmed and triangular side-notched projectile points have been associated with complexes interpreted to be ancestral to historic Northern Athapaskans. The Kavik or Klo-kut point, a contracting stem point with a slightly convex blade, is characteristic of a recent complex from Anaktuvuk Pass (Campbell 1968), and an examination of its distribution supports this interpretation (Dumond 1978: 56). It is reported from sites in Eastern Alaska, Western Yukon (Morlan 1973; Workman 1978), and south to the centrai Interior Plateau. Borden (1952) states that contracting stemmed points are typical of late prehistoric Carrier assemblages. Donahue (1973) reports them from a late component at Ulkatcho. Similarly Wilmeth (1978:159) found these points associated with the protohistoric Chilcotin component at Anahim Lake . Chinlac, Ulkatcho and the protohistoric component at Anahim Lake are all considered to be recent Athapaskan. It appears that these points become less frequent towards the southern plateau. Sanger (1970) reports no small stemmed points at Lochnore-Nesikep, and none are reported by Matson et al. (1984) from the mouth of the Chilcotin River. At Chinlac, Borden (1952) interpreted a series of miniature and triangular side\u00C2\u00ADnotched points also as part of the late prehistoric Carrier complex. The miniature points May 18, 2003-90 are not longer than 2 cm. and the larger side-notched points generally have indented bases and slightly convex blades, not common in the PPT. Aiso, the bases are often wide and the side-notches are most often smail and shallow. Similar points are absent from the much earlier Natalkuz Lake assemblage, which Borden indicated is a preceding PPT occupation (Borden 1952: 39). The differences between Athapaskan and PPT \"Kamloops\" side-notched points are described in more detail by Magne and Matson (1982,1987, this volume). Donahue's Group 3 points (1977: 441; Fig. 7) from Tezli show close resemblance to this Athapaskan point style, as do Montgomery's small side notched points reported from Punchaw Lake (1978: 126; Fig. 22). All three of these sites are in Carrier territory. The variation among triangular side-notched points found both outside and inside Athapaskan territory is discussed in greater detail later in this volume. It should be noted that both the Kavik point and \"Athapaskan\" style side\u00C2\u00ADnotched points will turn out to be useful in distinguishing between Athapaskan and PPT components at Eagle Lake. At Dixthada, tci-thos, which are large bifacially retouched scraping tools, were added to the Kavik complex (Dumond 1978: 55). This is a relatively crude tool type, likely created for a specific function and never very abundant. As Clark (1975:69-69) points out, \"tci-thos\" can refer to cortex spall tools, as well as bifaces and tabular tools. Thus this biface implement is likely an alternative form used for a s ingle function, that can be filled by a number of artifact types. Perhaps when the material to produce cortex spalls is not easily available these large bifaces are made, or they may be much modified cortex spalls. Wilmeth (1978) includes this class in what he considers distinctive Athapaskan artifacts and records three for Anahim Lake, along with only May 18, 2003-91 three, relatively poor examples of cortex spalls. Thus this class is not very likely to be very diagnostically useful for this purpose though it is not common in PPT assemblages (Richards and Rousseau 1987) in contrast to cortex spalls which are common in some PPT assemblages, including some collections from MOC. Wilmeth (1978:105-107) adds spurred end scrapers to Athapaskan complexes of the central plateau. This is also a functional type, and as such may cross cultural boundaries or may not appear in all components of a single Athapaskan tradition. Except by Wilmeth, these end scrapers have not been used as a distinctive Northern Athapaskan artifact type, although Gunnerson (1960) uses this type to help define the Dismal River complex. Rather than use these last two items as diagnostic types, we include them as artifact classes and use the entire assemblage composition, rather than single artifact types, to characterize the assemblages and see if the Athapaskan tradition is distinct from the PPT. A technological strategy of stone tool manufacturing involving a high proportion of retouched and utilized flakes is also considered by some to be characteristic of Carrier Indians. At Chinlac, Borden (1952: 32) found \"the thrifty use of such flakes, which many other groups would discard as waste\". This technological strategy was not fully evaluated prior to 2000 despite our ability to do so in the 1980s. Our impression then was that this thriftiness is more a result of raw material availability. Further, this factor did not appear to be distinctive in Magne's (1983,1985a) analysis of Eagle Lake and PPT lithic traditions. As an example of the \"lithic economy,\" we noted that the Anahim Lake collections excavated by Wilmeth (1978) contain thousands of unused flakes, and in view of the proximity of these sites to the Obsidian May 18, 2003-92 Creek source, this is not surprising. This idea, though, was evaluated in 2000/2001 with results discussed below, despite our earlier rejection. Dwelling structure style is the final variabie of Athapaskan material culture to be seen as archaeologically distinctive. The most detailed description i s found to be in ethnographies, although our work at the Bear Lake adds to this. As reviewed in more detail above, Lane (1953: 146) states that a ground level rectangular gabled structure about 15 feet wide and 20 feet long is the typical winter house of the Chilcotin. This dwelling would be above ground, usually with one rafter support post at each end of the house, and several poles leaned against it to form a gable-like frame. The sides were covered with horizontally laid poles, bark and/ or split logs, while the ends were enclosed with vertical members (Lane 1953: 145). The use of circular and rectangular pit houses is seen as a recent and not wide spread practice (Lane 1953: 160). Archaeological investigation has revealed a variety of recent house structure types in ethnographic Athapaskan territory. These include the remains of rectangular house frames at Chinlac (Borden 1952), Anahim Lake (Wilmeth 1978), and in Taseko Valley (Tyhurst 1996) while rectangular \"platforms\" are recorded at Punchaw Lake (Fladmark 197 4, 1976). We report on the remains of two well\u00C2\u00ADpreserved historic lodges at the Bear Lake site, and a less well-defined prehistoric one, as well as another less investigated one at Lingfield Creek with a tree-ring date of A.D. 1890. A variety of circular housepit types, on the other hand, are also recorded at Tezli (Donahue 1977) and Anahim Lake (Wilmeth 1978, 1979), also associated with what are interpreted as Athapaskan components. In general it appears that while circular May 18, 2003-93 features are more abundant and range over a greater period of time than do the rectangular features, the few tested rectangular structures are more recent than most circular housepits, thus supporting an association with recent Athapaskans arrival. Further study of the circular depressions may reveal a distinction between the housepits preceding Athapaskan migration and housepits adopted by them. Lane suggests that the Chilcotin pithouses are smalier than those used by their Shuswap neighbors (Lane 1953: 175). Evidence of variation in the placement of houses, benches and supporting structures may also occur. The relative lack of investigation of Athapaskan rectangular winter lodges is probably because of their lack of archaeological visibility. The absence of deep pits and rims resulting from the typical PPT pithouse make the rectangular lodge structural evidence that much less visible. This combined with the smaller and shorter duration of occupation make these winter base camps much less visible than the remains of the usual PPT winter pithouse village. In sum, five distinctive material culture traits that can be perceived archaeologically have been suggested that may distinguish between Athapaskan and PPT components. The two we found most useful were the rectilinear lodge and the differences in projectile points. These along with the settlement pattern differences reviewed previously, make up the core of our attempts to distinguish between the two. Although no single site or artifact type can be assuredly interpreted as Athapaskan\u00C2\u00ADproduced, the presence of a number of these can make such an interpretation definite when used in the Parallel Direct Historical Approach. May 18, 2003-94 The Parallel Direct Historic Approach R.G. Matson Introduction The problems involved in ethnic identification of prehistoric archaeological material are major ones. The very assumption that there is an equivalence between an archaeological culture and an anthropological culture is dubious. Cases can be cited to show that a single archaeological culture can consist of two ethnic cultures and that a s ingle ethnic culture can be divided into two archaeological cultures (Willey and Phillips 1958:49 - 50). In many cases, however, a close correspondence exists. Situations where the archaeological cultures do not correspond tend to be more frequent when changes in environment occur, either within a single ethnic territory, or where a single environment extends across several ethnic territories. In this section, the parallel direct historic approach (Matson 1982, 1991:317-320) is put forward as a solution to this problem by elimination of environmental (and temporal) factors. First, however, the question of what is meant by ethnic identification should be discussed. Ethnic Identification Ethnic identification can be correlated with language use, but this is not a very satisfactory definition for archaeologists, ethnographers or historians. Ethnic identification includes other aspects, such as shared beliefs about the world and how one makes a living. Researchers working with living people often see ethnic identity as a self identity of a group in contrast with others. This view of ethnicity, though, is not May 18, 2003-95 very useful for non-western peoples. It is clear that many cultural groups seen by anthropologists as consisting of separate ethnicities were not \"recognizing their unity and difference from other formations (self-awareness) and expressing this in a self\u00C2\u00ADappointed name\" (Dragadze 1980:162 cited in Renfrew 1988:216). They identified with much smaller villages or bands, as these were the iargest effective political groups. Fried (1975) in fact went so far as to argue that \"tribes\" which correspond roughly to e thnic groups politically were the result of European pressures in the Post-Contact world. So an ethnic group as used here is an anthropological \"culture\" as opposed to language use or self identification. It is also distinct from the usual archaeological definition of phases. Thus the relationship between archaeological culture and ethnic identification is an open one that needs to be determined on a case by case basis. Renfrew (1988:217) points out that ethnicity is \"a matter of degree\" and that the amount of self-awareness is variable, and, in fact, may be absent, or shared among several hierarchical layers, as among the classic Greek city states. Obviously, the sort of ethnicity to be discovered with non-perishable archaeological material at the Archaic stage (as defined by Willey and Phillips 1958) will be only the more definite boundaries, not subtle differences between villages and bands. I now turn to how ethnicity can be determined under such conditions. The Direct Historic Approach The direct historic approach was first proposed with this name by Julian Steward in 1942 (Steward 1977) and refers to working back in time from a known ethnographic basis. In this approach, the most recent archaeological material is equated with the May 18, 2003-96 ethnographic material and is interpreted in light of the ethnographic descriptions of cultural activities. From this known baseline, as one works back in time, the changes in archaeological material can be inspected to see what activities change. Breaks in the sequence are equated with the arrival or development of a new ethnic group. When some material object diagnostic of a culture is no longer present in the archaeological sequence, the ethnic group is also considered to be no longer present. Dealing with items that are associated with historic groups is risky since in prehistoric times the association of objects with a particular ethnic group may not be valid. There is no 11a priori\" reason why any object or group of objects should have a popularity that coincides with the life of an ethnic group. Material objects can also diffuse from one group to another, further complicating the situation. The direct historic approach has been used by Donahue (1977) and Wilmeth (1978, 1979) in their attempts to discover when the A thapaskans arrived in central British Columbia. The timing of the arrival of the Athapaskans requires their ethnic identification. Both Wilmeth and Donahue had problems in determining this date, because changes occur in any regional sequence whether or not any change in ethnic identity takes place. Thus, additional information is needed to decide which change in the sequence is also a change in ethnicity. The Parallel Direct Historic Approach In order to deal with these problems, I proposed (Matson 1980) the use of the parallel direct historic approach: two sequences, one in \"one ethnic area\" and another \"in an adjacent ethnic area\" are compared in detail. It is important to hold May 18, 2003-97 Historical AthapaskanMigration Present Migration Absent? B ? ? C ? D E 1800CCB B A A a) Direct Historical Approach\u00EF\u00BF\u00BDAthapaskanb) Paralllel Direct Historical Approach where one Archaeological Culture (C) is found in both areas and continues to contact times in oneAthapaskanMigration A.0.0Historical AthapaskanArea A Historical PPTAreas 1800 A.D.Oc) The Athapaskan and PPT situation in the abstractAthapaskan PPTEagle LakeLulua PhaseEagle Lake Phase\u00EF\u00BF\u00BD-? ? - - .- \u00EF\u00BF\u00BD - - - - - - = -. \u00C2\u00B7 . . ..\u00E2\u0080\u00A2 : : \u00C2\u00B7Jr1h t True North t \u00C2\u00B7rue ,orth l --- ----- -- --\u00EF\u00BF\u00BD \u00EF\u00BF\u00BD - _. -- -_..._ ..... ..._--..... _ -- ..... - ..... _..... ..... \u00EF\u00BF\u00BD-..... _ ..... --.QUAD\u00EF\u00BF\u00BD- ....__ IP e It I I I I I I -0 \u00C2\u00B7 80 100ln ,, , , '\" Slope Directon .. \u00EF\u00BF\u00BD _ ......... _ .... _ ..... \u00EF\u00BF\u00BD \u00EF\u00BF\u00BD-Denee Aid;, WIiiow Mal'lh \u00EF\u00BF\u00BD .._ _.., -- ----- 41-- ..... -..... /4. rn:'#'\u00EF\u00BF\u00BDX1\u00EF\u00BF\u00BD \u00EF\u00BF\u00BD \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD : . : :\\u00C2\u00B7\u00C2\u00B7::: v.1.1.t\u00C2\u00B7IJ.1:1 ii:, True North t )/f(\u00EF\u00BF\u00BD.,\u00C2\u00B7'\u00C2\u00B7 . .. ,._\/.-.:.:\u00C2\u00B7? :_:\u00C2\u00B7.\u00C2\u00B7.\u00C2\u00B7.' ::-.: ,f :yts.fr.it: ., .. ,.. , .... \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00C2\u00B7 .. :\u00C2\u00B7 .. , ... \u00C2\u00B7,.?\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD;?: \u00EF\u00BF\u00BD\u00C2\u00B7- Z{.?tU \u00C2\u00B7r\u00C2\u00B7: \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \::::\u00C2\u00B7:._\u00C2\u00B7:\u00C2\u00B7.-..:,::-_: _ _.... < .:::,:':\"::\u00C2\u00B7 .-..:\u00C2\u00B7 : \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 ;. :\u00C2\u00B7\u00EF\u00BF\u00BD1f\u00EF\u00BF\u00BD.: ;-. ::\"\u00C2\u00B7 \u00C2\u00B7:\u00C2\u00B7:: ;-. :\u00C2\u00B7\u00C2\u00B7.:.-:,,,..,,.,., 0 l!O 100ffl True Nort, t True Norll t , ......... , 0 50 100., l111d1111I 0 50 1001a OllNu ..Figure III-6. Location of Taseko Lake Quadrats. Magne (1984:4) hypothesized that areas used today were also those used in the past and therefore divided the frame into 'P' (Primary) quadrats that evidence of recent use, and those that did not, 'S' (Secondary) quadrats. He initially selected 4 'P' and 7 'S' quadrats, but found that two quadrats in Yohetta Valley (Figure III-6) were in a Spruce and Swamp environment could not be surveyed and replaced them with two T (\"Judgementally\" selected) quadrats, thus modified his original sampling scheme. In the end 11 quadrats (Figures III-6,7) were surveyed 6 'S', 3 'P', and 2 'J'. FIGURE Figure III-7. Taseko Lake Quadrats Sketch Maps. The elevations of the quadrats ranged from a low of 1340 m (4400 ft) to 1675 m (5500 ft.) although only one has any significant area above 1525 m (5000 ft). Collection and Survey Procedures The procedures used for survey and site collection at Eagle Lake had originally been developed by Matson (1971; 1991) and Matson and Lipe (1975) during their work in the U.S. Southwest. The searching of quadrats is done by crews of 4 to 6 individuals spaced at about 10 m intervals. Stray artifacts are collected during quadrat sweeps, but May 18, 2003-119 concentrations of artifacts or features such as cultural depressions are marked with flagging tape so that they can be mapped and collected later. After a quadrat is searched any such sites are revisited and a grid system is set up for mapping and collection. Each 25 m X 25 m or 50 X 50 m area is sketched and all artifacts within it are located and keyed to the map, with all material within a 1 m radius given the same location number. Forms dealing with botanical, physiographic and cultural features are also filled out for each site and for each quadrat (see Appendix Ill for the forms used in 1979 and 1983). The forms are similar to the ones used at the Mouth of the Chilcotin with modifications for the local situation, and with the important addition of \"overview'' Oudge 1973), which Pokotylo (1978) found useful at Upper Hat Creek. Briefly, overview is a measure of how much of the site surroundings can be seen from the site. The Potato Mountain investigations followed closely the procedures used at Eagle Lake and all prehistoric sites located during both quadrat surveys were completely collected. An important part of our \"collection\" procedures was our evaluation of pit features. In many cases, the function, roasting pit, cache pit or housepit, is obvious based on size or presence of fire-cracked rock. In the size range of 2-5 m though, it is not always obvious what the function is via surface inspection. In some cases, a small trowel test can provide definite evidence, with fire-cracked rock and dark,. charcoal soil are usually abundant on roasting pits, but our standard was the use of a King tube (an \"Oakfield soil sampler\") 2.5 cm in diameter. This would usually provide good information with little disturbance, as cache pits usually had a fair depth, while roasting pits usually had abundant rocks close to the surface. Although not all depressions could be definitely classified, by the use of these procedures, most were. May 18, 2003-120 In the Taseko Lake Project very similar survey procedures were used as at Eagle Lake, except that for the initial sweep of the quadrats with smaller crews, a distance of 20 m between individuals was used (Magne 1984:9). The same mapping procedures and forms as for Eagle Lake and Potato Mountain were used, but with two exceptions, no sites were collected. Soil probes were used sparingly on cultural depressions to determine whether smaller depressions were cache pits or roasting pits as at Eagle Lake. Thus, the Taseko Lakes project located and mapped sites, using very similar procedures as used at Eagle Lake and Potato Mountain, but did not usually include collecting artifactual material. All surface prehistoric sites found in both years of the Eagle Lake quadrat survey were completely collected. The only exception to the complete collection rule was one historic site. We did plan to subsample sites if they were very large or had very dense lithics but no such sites occurred, although Quadrat 19-1 came close. In the event that a number of quadrats surveyed would have extensive duff it was planned to use a modification of the technique used by Spurling in the Peace River area (Spurling 1978; 1980) and since then by others. This field methodology was developed for surveying in a forested area (Peace River) where much of the ground surface is obscured by vegetation litter. The procedure consists of shovel excavating and screening a 1 m X 1 m pit in each 50 m X 50 m subsection of each quadrat. Since Spurling used quadrats 500 m on a side, 100 such pits are excavated in each quadrat. Although we did survey a number of quadrats with poor surface visibility we decided not to use a variant of this method. The use of it is very costly in terms of time and we thought that the advantages were not sufficient to outweigh this cost. Also, May 18, 2003-121 because of the abundance of sites with pit features at Eagle Lake, unlike Peace River, we did not think that many sites would be missed since the pit features are visible in forested conditions. Finally, it would be difficult to compare quadrats that had been searched in this manner with those that had not, since only a few would have required this method. Still, there are strong arguments for the use of this search method under dense forest cover situations. Klassen and Ridington (1998:26) carried out an interesting subsurface testing variant in a survey that overlapped with both the Eagle Lake Quadrat and Chilko River surveys. In their case they tested likely landforms on landforms of \"archaeological potential\" that did not have good surface exposure. This enabled them to discover a number of sites without the higher costs involved in the Spurling method, but one wonders about the bias involved. Chllko River Survey Design As indicated earlier, in addition to the quadrat survey at Eagle Lake, we also carried out a non-collection survey along the upper end of the Chilko River, mainly to locate sites that could contribute to resolving cultural history questions regarding the Pre-Plateau Pithouse Tradition, i.e., the period prior to 3000 B.P. Since there appears to be a concentration of sites along the rivers on the Plateau, we thought this would be the best area to survey to attempt to find a stratified non-house pit site with evidence indicating a separable early component. The absence of house pits was desired because the presence of house pits usually means that the deposits are highly disturbed (Wilmeth 1977). Mav 18, 2003-122 .I The procedure we developed was to survey both sides of the river, focussing on the immediate river banks and lower, open terraces. We used the same cultural forms as on the quadrat survey and produced equivalent sketch maps, but usually did not do any collection. Our intensity of survey was also less than that on the quadrats, as measured by the distance between searcher. On the other hand, more attention was spent on sites that looked like they might have significant depth of deposit. Those sites that looked like they might have a significant depth of deposit were later tested. The area surveyed was the Chilko River, from near its outlet to about 30 km downstream (northward). In addition to this, the area at Bidwell Creek, the 11Quiggly holes\" site, was also mapped and tested, about 25 km (15 miles) further downstream from the end of the continuous survey zone. We also extended the survey to the shores of Chilko Lake adjacent to the outlet, and actually began our survey there. In all 105 Chilko River sites were located, of which six had been previously recorded, several by the quadrat survey. The distance from the river that was surveyed was variable, and often dependent on the landform and amount of ground coverage, with broad open terraces surveyed even if they extended some hundred meters away from the river. In an idiosyncratic survey carried out in 1997 (Klassen and Ridington 1998) that included some areas that we surveyed along the Chilko River, a housepit site (with 5 housepits) was discovered that we apparently had missed (EkSa 142). On inspection (Klassen and Ridington 1998b:129-131) it turns out that we found an adjacent site (CR #9; EkSa 116) closer to the river and the housepit site (EkSa 142) actually began more than 100m from the river in a lodgepole and aspen forest. So this ''error\" appears to be more a limitation of the survey May 18, 2003-123 technique than an survey error since the site was not on an open area adjacent to the river. As we will point out later, housepit sites are located up to 3 km away from the ChHko River according to our data -- a finding Klassen and Ridington (1998:83 \"two kilometres 11) agrees with -- so a riverside survey will miss those that are not adjacent. What this means is that there are clearly more housepit sites along the Chilko River than we located; those that we may have missed, and those further from the river than we searched, and EkSa 142 is an example of the latter. We will also report on another \"near Chilko River\" housepit site located by Klassen and Ridington (1998) in another context. May 18, 2003-124 3 20 2 6 46 3 0 0 4 9 8 5 3 6 6 0 0 7 0 5 8 0 1 9 5 16 10 0 26 11 0 0 12 15 33 13 J 6 (5 Repeated)Median 3 6 Confidence 0-6 (98%) 0-26Intervals 0-3 (73%) 5-16Interquartile Range 0-6Mean 3 .42 Estimated Grassland 266.5 Total 0-2613. 421046.5 'fable IIT-1 Mouth of Chilcotin Grassland Quadrats Formed F ' . Lre Cortex fine Coacse Lithic Crack. Spa 1J Basalt Basa 1 l Chen: l)roj.6 6 0 6 1 1 1 1 5 5 0 6 1 1 1-61-5\"'ools Rock 3 + 75 + 0 261 + 40 + 6 0 0 il':i + 0 + 0 24 + 40 + 6 + 0-7S --+ 0-40 +-+0 75 --+ 3.]7 113./ 247.3 3403 Tools \u00EF\u00BF\u00BD Dcbitaae Poi nls 0 17 0 8 1 1 i.. _i 0 0 G 1 0 () () 1 /' 1l 0 Ci 11 0-80 1 J ,109. ':> 0 r\ V '::,J 11 0 () 253 33,1 JO 86 0 0 0 0 () () '::,l) -111 S(' d 0 2 6 1 'a ,_ / 30 86 n u 0 J 0 0 so G 86 0 t11 0-86S 1 . 8189:,.I] 1\036.S 0 0 JO 6 0 {' d 216 S 1 / 9 J 0 0 0 0 {) 0 1)9 13 0 () 0 f\ (j 0 90 l) (J 3 0 0 6 0-9 0 3 0 30 0-63./S 13868 ?SJ.S (f) w (f) (f) <{ 0::: Cl (f) ID a: 1.1.J ::c en !Il :::, a:: :c (f) (I) w w a: \u00EF\u00BF\u00BD XX X x xx \u00C2\u00B7x\u00C2\u00B7 \u00C2\u00B7x\u00C2\u00B7 \u00C2\u00B7x . . . . .X \u00E2\u0080\u00A2\u00E2\u0080\u00A2 x.x x.x.x. X X XX X .x. \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \"' a, \u00E2\u0080\u00A2 .. x . . . xx.x_x.x_x.x,.., ,,...\u00EF\u00BF\u00BD .><. \u00E2\u0080\u00A2 x .. x.xx:x: :xx .. x.xX X xx xx : :xx: : : .x. xx XXX .\u00E2\u0080\u00A2 :x x:x:x:\u00C2\u00B7x\u00C2\u00B7x\u00C2\u00B7x x\u00C2\u00B7x\u00C2\u00B7xx:x:x:x : xx X .x.x. x .. x.x X X .xx: : :x \u00C2\u00B7x\u00C2\u00B7x\u00C2\u00B7x x\u00C2\u00B7 \u00C2\u00B7x xx xx\u00C2\u00B7x\u00C2\u00B7x\u00C2\u00B7xx\u00C2\u00B7x\u00C2\u00B7x\u00C2\u00B7xx\u00C2\u00B7x\u00C2\u00B7x\u00C2\u00B7xX X X X X X X X x.x.x xxx X X X X X X x\u00C2\u00B7 . . X \u00E2\u0080\u00A2 .>< \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00C2\u00B7xx\u00C2\u00B7 \u00C2\u00B7x\u00C2\u00B7.x;x : :x: x:x:x:x x . . .x.x.x ... >< x. :x: :x: X >< \u00E2\u0080\u00A2\u00E2\u0080\u00A2 ><. X \u00C2\u00B7 \u00C2\u00B7 \u00C2\u00B7x .x. . x: : :xxx xxxxxx xx -N\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BDw\u00EF\u00BF\u00BD\u00EF\u00BF\u00BDmO-N-Nmv\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD::c u z w m w z 0 N 'l'i:tlil r::i E$ II\ '-'2, Qua ' \ \u00EF\u00BF\u00BD I I I ..__,_ S 400 E\u00C2\u00B7O OUADRAT \u00EF\u00BF\u00BD \u00EF\u00BF\u00BD Ero __________________ \u00EF\u00BF\u00BDE-400 0 STEEi\" SLOPE fOIIESTEO IIULLIES &011 \"'OOEIIATELT DENSE fOIIEST 2900\u00C2\u00B7-----5\u00C2\u00B7400 1\u00C2\u00B7400 [\u00C2\u00B7 0 [\u00C2\u00B7400 OUADRAT 3 S\u00C2\u00B7O [\u00C2\u00B7400 E \u00C2\u00B70 S\u00C2\u00B7O 1111.ISSL.t.lt0 8ENTLE 111.t.SSLAND 0 so ... 1\u00C2\u00B7400 [\u00C2\u00B7O E'*EN QUADRAT 5SLOI\"\u00EF\u00BF\u00BD 1\u00C2\u00B7400 \u00C2\u00A3\u00C2\u00B7400 Figure \u00EF\u00BF\u00BD Sketch Maps of Quadrats 1-6 S\u00C2\u00B7O &\u00C2\u00B7O E\u00C2\u00B7O \u00C2\u00A3\u00C2\u00B7400 r\u00EF\u00BF\u00BD-\u00EF\u00BF\u00BDT_\u00E2\u0080\u00A2:\u00E2\u0080\u00A2\u00EF\u00BF\u00BD\u00E2\u0080\u00A2\u00EF\u00BF\u00BD:c..\u00EF\u00BF\u00BD;;,:,-;,=,\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD:::::::\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD:::N::::::=1 t=====11= ..=_____ =g=[==11===== .. =lt=l=V=!_.=R:-_=-_=-_=-_=-_-_t_--lSE U0N.t.L CRE C \u00EF\u00BF\u00BD r::. we\u00E2\u0080\u00A21 tt\u00E2\u0080\u00A2to 1A1t1 \u00EF\u00BF\u00BD , .. ..,.,;,!-;_ _ .S 400 E\u00C2\u00B7O OUADRAT 2 l\u00C2\u00B7O E\u00C2\u00B7O ,,.-------, ... -;I ,,,.o.,o \u00E2\u0080\u00A2\u00E2\u0080\u00A2 ,.. ' ' I , __ ... ! _ .... \u00EF\u00BF\u00BD ,_, \ I , __,. t TEIIR.t.CES N t IIODER.t.TE SLOPE i \u00EF\u00BF\u00BDASSLAlt0 0 $ 400 E \u00C2\u00B7400 S\u00C2\u00B7O [\u00C2\u00B7400 SOIi 1\u00C2\u00B7400 $\u00C2\u00B7400 E\u00C2\u00B7O E\u00C2\u00B7400 OUADRAT 4 IIOLLIN$ 0 so ... 1\u00C2\u00B7400 [\u00C2\u00B70 OUADRAT 6 . .SCATTERED ARTIFACT! N t IIIIASSLANO 1\u00C2\u00B7400 [\u00C2\u00B7400 l.Y Table III 3 Mouth of Chilcotin Forest Quadrats \u00EF\u00BF\u00BD\u00EF\u00BF\u00BDSites \u00EF\u00BF\u00BD F-1 0 0 0 F-2 0 F-3 0 F-4 0 F-5 0 F-6 0 F-7 0 (Repeat of F-1) Median 0 Confidence (98%) 0-0Intervals (87%) 0-0Intequarcile Range Mean 0-00. 05 0 0 0 0 0 0 0 5 0 0 0-00.83 2 0 0 0 0 0 0 0 2 0 0 0-00.33 Environment Unlogged Douglas Fir Forest Transitional between Forest and Grassland. Logged off forest. Logged off forest. Logged off forest. Logged off forest \u00EF\u00BF\u00BD Unlogged Doublas Fir Forest. 1\u00C2\u00B7400 f >O OUADRAT F - I S\u00C2\u00B70 E\u00C2\u00B7O L0GUD 0f'F THICK IIElllt0WTH OUADRAT F-'3 0 IIOM. ..._.__. $\u00C2\u00B7400 E\u00C2\u00B7O L0GIIED OH OUADRAT F-5 -nr--\u00EF\u00BF\u00BD1\ 1\u00C2\u00B7400 (\u00C2\u00B7400 N I l\u00C2\u00B7O [\u00C2\u00B7400 S\u00C2\u00B7400 [\u00E2\u0080\u00A2400 l\u00C2\u00B7O E\u00C2\u00B7O SJ'OIIA0IC \"' SCATTEREC, /' Of'EN (4, AREAS 0 IIOM HEAVILY FOflESlfD VUf'f'[II lllcEWf:11 CK. ltD. 1\u00C2\u00B7400 [\u00C2\u00B70 QUA DRAT F - 2 S 0 \u00C2\u00A3\u00C2\u00B70 S\u00C2\u00B7O E 400 1\u00C2\u00B7400 [\u00C2\u00B7400 i S\u00C2\u00B70 [\u00C2\u00B7400 ,011\u00C2\u00A3.STEO, SCATT[ll[D OPEN GRASSLAND f'ATCH[S 0 S 400 E>O IOM. OUADRAT F-4 , ..' LOGGIN(. /STAGING AREA S 400 E 400 10 so rE_-o ________________ E\u00EF\u00BF\u00BD\u00C2\u00B7\u00E2\u0080\u00A2oo ICATTEIIEO TOUHG DOUGLAS ,i11 S\u00C2\u00B7400 E\u00C2\u00B7O IIIIAISLAND \Sf'ORADIC Y0UNI. DOUGLAS fllli o\u00EF\u00BF\u00BDo 1\u00C2\u00B7400 l-400 QUADRAT F-6 Figure .W Sketch Maps of Quadrats F-1 to F-6 FIGURE Figure III-8. MOC Grassland Quadrats 1-6. FIGURE Figure III-9. Sketch Maps of MOC Grassland Quadrats 7-12. FIGURE Figure III-10. MOCGrassland Quadrat Locations on Aeriai Photograph. In contrast, only one of the six forested quadrats had any recorded sites, and that one was at the edge of the forest (Table III-2 and Table III-3). Sketch maps of each of these six quadrats are shown in Figure III-1 L That three of the forested quadrats were in logged off areas (Figure Ill-12) may have increased the likelihood of empty quadrats, however. as the ground in those areas was both very disturbed and heavily covered with the trash resulting from logging. In summary, though, it is clear, the sites are concentrated in the Grassland stratum. TABLE Table Ill-3. MOC Forested Quadrat Summary. FIGURE Figure III-11. MOC Forested Quadrat Sketch Maps. May 18, 2003-126 FIGURE Figure IIl-12. MOC Forested Quadrat Locations Aerial Photographs. As a means of providing another perspective on quadrat variability and on the differences between the grassland and forested quadrats, multivariate analyses were undertaken of the quadrats environmental characteristics and of the sites' environmental and cultural characteristics (Matson et al. 1984:38-43). The environmental relationships are probably best shown by Figure 111-13, the metric multidimensional scaling (aka Principal Coordinate Analysis ( Torgerson 1958; Gauch 1982:162; Jongman et al. 1995:152; Matson and True 1974) of the presence-absence environmental information of Table III-2, using Jaccard' s complement as a measure of similarity (Matson and True 1974; Sneath and Sokai 1973). The first two dimensions account for 63 % of the trace, or squared distance from the centroid. The dimensions can be considered as the underlying factors that \"explain\" the differences between the quadrats in terms of environment. The numbered groups are those derived from a related cluster analysis (Matson et al. 1984). Note that 5 of the 6 forested quadrats are well separated from the grassland quadrats, with the six, F -2, in Group III of similar ecotone quadrats. Group II consists of two quadrats located on the steep area between the river and the upper \"Bench\". Note how all the steep quadrats are on the right hand of the figure. Group IV is on the edges of the upper bench with Q 6 including some of the steep ravines, and 11 is next to McEwan Creek. Neither had very much archaeological material present. FIGURE May 18, 2003-127 :c: (!) :i: /. - _ _ _ + 0/M I/ F-4 ----- ---$ F-3 + $F-5 I, I \ \ \ \\ \ F-6 \" (9 \"-..._ ___ _ -+ --+ + '-' ' \" \ F-1 ID \ I I / / / ,/ ----3 0 ..J + / ..:::i:.r.-, / -i- \ / 12 J + - 3 '\+ ,,,..-v * /01M/ / >\u00C2\u00ADa:: 0 / / / / / ,,,\" / / / / / V /,,,-/ I $ \u00E2\u0080\u00A2 I I ..., I ' --{ f14 ' \ 9 \ \ I / \ / \ / \ 9(1 /', / ._.,,, FLAT \\u00EF\u00BF\u00BD // $F-2 111 / \u00EF\u00BF\u00BD (9 8 / \. ---'---:-+ + STEEP Figure t& Metric Multidimensional Scaling of Quadrats by Environmental Info rmatio\u00EF\u00BF\u00BDDimensions l and 2, 63% of Total Distance. Roman Numerals refer to Clusters in Figur e 11. \u00EF\u00BF\u00BD Figure III-13. Metric MDS of MOC Quadrats by environment. Almost all of the archaeological material exclusive of cache pits is found in or adjacent to Cluster V. These include the bulk of the upper Bench and are some of the flattest quadrats. All four have house pit sites present. Quadrat 5 in this plot should be a member, although it was an isolate in the cluster analysis and shares most characteristics with the other four, including a significant house pit site. Quadrat 12 (see Figures III-9 and-10) has the most forested area of any in this cluster. Quadrat 2, of course, is along the Fraser River, and also has abundant archaeological material. It is located at the bottom of the plot, indicating its position as the lowest and driest quadrat--at least in term of its terrestrial plants. The recording of the quadrat environmental variables, then, is in good accord with the quadrat location on the landscape. The archaeological material, as well, is clearly positioned in particular environments, with house pit sites concentrated in the lower left of the plot. Wetness, as caused by elevation, and steepness, are the two main factors of the environment in this area, with the house pit sites being concentrated in flat, dry areas, but adjacent to the forest, and, as discussed before, usuaily with non\u00C2\u00ADFraser River sources of domestic water. This scaling analysis demonstrates that the botanical and physiographic information that we recorded are good measures of the \"real\" environment. This pattern, counter clockwise of low and dry, to steep and high, can also be thought of as a progression, spoiled only by the flatness (relatively speaking) of quadrats F-6 and F-4. If the entities were time sensitive archaeological assemblages, May 18, 2003-128 Table 111-4 Final MOC Synthetic Site Classes \u00EF\u00BF\u00BD \u00EF\u00BF\u00BD Site Number #HPs #CPs Cachepit Sites I Ravine Cachepit Sites 7-1 5 10-1 5 10-2 8 10-3 6 10-4 3 10-5 4 II Ecotone Cachepit Sites 1-3 1 1-4 5 1-5 1 1-6 1 4-4 1 8-1 1 9-3 1 9-4 7 12-4 8 12-5 2 F2-1 2 F2-2 3 Housepit Sites II! Housepit and Cachepit Sites 1-1 2 1-2 1 11 4-1 8 4 5-1 3 6 9-1 3 4 9-5 2 4 12-1 7 8 12-2 1 1 12-3 2 4 12-6 5 10 Riverside Sites IV 2-1 2 5 2-2 23 2-3 3 2-4 9 2-5 4 6 2-6(?)Chert Debitage Sites V 4-24-59-24-6(?)Unique Sites VI 4-3 3 cachepits plus a few artifacts (disturbed) 6-1 a few artifacts only 2-6(?) a few artifacts only such a progression would have been interpreted as a \"seriation\" (Matson and True 197 4), but in this situation the two factor interpretation appears to be superior. In any case this plot shows a strong relationship between environment and sites for the PPT at the MOC. This relationship was explored in detail by Matson et al. (1984:43-65) by c lustering and scaling sites in a similar fashion as the quadrats were scaled and comparing the results of clustering and scaling the sites on the basis of the artifactual material present there. The results of these analyses were six site types, four of which correspond closely with ethnographic expectations of the PPT which is turned to next. The six site types are tabulated in Table 111-4. These site numbers are slightly different than those reported earlier for housepit sites, as these show as separate sites housepit(s) that are more than 50 m from each other, while previously we used a cutoff of 100 m to separate sites, which is probably closer to ethnographical reality, although not easy to deal with archaeologically. Cachepits sites are divided into two classes, those that are at the ecotone between forest and grassland, the same general location as most housepits, and those located away from other sites, but along the ravines between the upper bench and the Fraser River, the Ravine Cachepit sites. A member of the latter, outside of any quadrat, was tested (\"Cache Pit\" near Quad rat 7 on Figure 111-10), and found to be a cachepit, complete with tree bark, which ethnographically is said to be used to line cache pits (Teit 1900:234). TABLE Table 111-4. Final MOC Synthetic Site Classes. The third site class is that of the Housepit sites. Note that in this set, all also had May 18, 2003-129 cachepits within 50 m of the houses, and the cachepits were often intermixed with the housepits. Although 10 sites are shown here, all on the upper bench, if the 100 m criteria is used (as used earlier), only 8 are present, but the number of housepits and cachepits per housepit site is increased. All of these 10 (or 8) sites are found in relatively level areas, close to the edge of the forest and water sources. Although in several cases the local source of water took some searching, the winter situation may have been more obvious than it was to us in the summer. In short, a very good fit with ethnographic expectations for the PPT. The fourth class found in the MOC survey was the Riverside sites, all found in a single quadrat (Q 2). For our purposes, there appear to be three separate site classes within this one quadrat, one or two winter housepit sites, riverside cachepit sites (making a total of three separate classes of cachepit sites), and riverside fishing sites. The number of housepit sites is dependent on whether one uses the 50 m or 100 m criteria. This part of the river does support a fringe of trees, including Douglas fir, along the shore, providing a local source of water, and there is a ravine which runs through the quadrat (Figures III-8 and III-10) which may provide an alternative source of water to the Fraser in the winter. Thus, adding in the one (or two) Quadrat 2 housepit sites to the rest, in total 9 (or 12) housepit sites are present within MOC quadrats, and all fit the environmental locations expected for the PPT model. Further, all housepits are within 100m of other housepits, meaning that all (or all but two, as recorded in the field) are multiple housepit sites, with most housepits being located immediately adjacent to other housepits. This points out how clustered the housepit sites can be in the PPT. May 18, 2003-130 Three sites were found in quadrat 2 that had cachepits and no housepits. Two of these, (Q2-2 and Q2-4) had 23 and 9 cachepits (respectively) and very little other information. These, then, fit the model for large cachepit sites for the storage of salmon, located immediately adjacent to fishing locations. The third site with cachepits, Q2-3, had only 3 and a great deal of lithic artifacts, indicating activities well beyond storage (Appendix Ill; Table 1). Located immediately adjacent to the river, and to a historic fishing location next to an eddy, this is a good fit for a riverside fishing site. Thus, we have divided one site class, the riverside sites, into three groups, all fitting the expectations of the PPT model. Given the small numbers this step was not formally taken by Matson et al. (1984), although these sites were discussed in a similar fashion there. The fifth site class is that of the chert debitage sites. These are sites that were dominated by pink chert debitage, in contrast with the \"basalt\" dominating all the other sites. No features were present on these, although they were located in areas with good overview, and were associated with chert atlatl points. Matson et al. (1984) argue that they are Pre-PPT on the basis of lack of features, atlatl points, and presence of chert. Sanger (1970:117) notes that the use of vitreous basalt increases through time, so, by extension, a site with very little vitreous basalt should be early. Similar chert was found in an alpine site near Lillooet, as reported by Vanags (2000). Finally, a few small MOC sites could not be squeezed in any categories, and grouped together as \"Unique\" sites, the sixth site class. The MOC settlement pattern analysis thus acts as the \"expected\" for the PIT pattern, as summarized earlier. It should be remembered, though, that this is an May 18, 2003-131 -.... --,PIT HOUSE FILL BROWN SANDY LOAM WITH SPECKS Of RED CLAY CHARCOAL ROOT \u00EF\u00BF\u00BD ROCK .... CROSS SECTION Of FLOOR PLAN SHOWN IN PROFILE ORIGINAL SURFACE CONTOURS BEFORE EXCAVATION 0 CONTOUR INTERVAL, 0.2 METRE EXCAVATED s;TERILE, HARD PACKED, LIGHT GRAVELLY, C1LAY SILT RED CLAY. PFlESUMED BURNED ROOF FILL CONCENTRATION IM[ TRE ' \ . \ \ \ j H,UIMC\u00EF\u00BF\u00BDTON\u00C2\u00A3 ,.; WITHPfTCH ' \ ' I T 0 \u00E2\u0080\u00A2 I \"' \ .. ,,-\u00EF\u00BF\u00BDrt BEAVE/f: Pfl.VIS...J.> ESSICW. \u00E2\u0080\u00A2 : r \u00E2\u0080\u00A2\u00E2\u0080\u00A2 OCORE 2 I ' 'I /-ELK ANTLE/If SUTT 3 ANTLE/If TIN\u00C2\u00A3 \ ,o ,. I ' ' . 6 '\u00E2\u0080\u00A2 \u00E2\u0080\u00A2 i I 'I ' \u00EF\u00BF\u00BD ! 'I / I HAAIAIERSTON 'I> ;-, ., /' \ ..... -\u00C2\u00B7 \u00E2\u0080\u00A21-1' ' I ) EkRo 48 . ' -o, ,. ' I FLOOR PLAN OF EXCAVATION UNITS AT CONTACT WITH \"FLOOR\" 5 0 6 0 0 30 2.0 3.0 40 \u00EF\u00BF\u00BD O EAST WALL, UNIT 4 3 O UNITS 2, 3, 4 II 6 EXCAVATED TO TOP Of STERILE IMMEOIATEL) BELOW FLOOR UNIT I EXCAVATED 40 CMS BELOW FLOOR. EXCEPT FOR ROOENt HOLES DEPOSIT BELOW FLOOR WAS STERILE C-\u00C2\u00BBIWALL, UHITS Ill 6 PRESU\"'ED RODENT BURROWS :nr-Figure J;!. EkRo 48, Floor Plan and Cross-Sectionof Excavated Housepit.CROSS\u00C2\u00B7SEC TION $HOW\"' IN PA'C,:IL( extremely densely occupied area, probably the most densely populated area within Shuswap territory (Teit 1909a:464-6), so the relatively large numbers of sites found within a small number of quadrats is not to be expected elsewhere. Mouth of Chilcotin Tes ting In addition to the survey, four minor excavations were carried out in 1974. First, a previously known and collected site (by the O.F.Y. project, Mohs 1973) EkRo 48, had a small housepit tested with five 1 x lm units (Figures III-11,-15 and 16) revealing a well\u00C2\u00ADp reserved floor, with a number of artifacts on it. Two charcoal samples were dated; 870+ I - 80 B.P. (GaK 5326) and 1450+ / -75 B.P. (GaK 5327), with the latter on the floor. Since two Kamloops side-notched points (Appendix III; Tabie 3) were found during this excavation, the first date is more apt to be valid for the last occupation of this house. The older date indicates an occupation in the Plateau Horizon (Richards and Rousseau 1987), an occupation that is confirmed by the presence of large barbed points found in MOC collections, including one from EkRo 48 (Matson et al. 1984:169), which are markers of this Horizon. FIGURE Figure III-14. Contour Map of EkRo 48. FIGURE Figure III-15. EkRo 48 Excavation Floor PLan. Two similar units were also placed in a housepit on site EkRo 31 (Quadrat 4-1), but no clear floor was present, and one unit appeared to be highly disturbed, although May 18, 2003-132 abundant archaeological material was present (Appendix III; Table 3). The presence of five Kamloops side-notched arrow points, though, indicates occupation in the last 1200 years. The combination of these two minor excavations supports the surface collection indications that the PPT occupation is concentrated in, or at least continues strongly into, the Kamloops culture. Three 1 meter units were also excavated into site EkRo 18, a site of very large and deep housepits (Figure 111-16). This is believed to be part of the \"Lillooet\" Phenomenon, explored by Stryd (1974; Stryd and Lawhead 1978) and Hayden (2000, 2001) in the Lillooet area. This complex in the Lillooet are is very large sites of very large and deep housepits dating between 900 and 1500 years ago (Hayden and Ryder 1991; Lenert 2001) . Some sites, even in the Lillooet area, have modest numbers of these very large pithouses (the Bell site in Stryd and Lawhead 1978). EkRo 18 stands out in the MOC as h aving uniquely large and deep housepits, and we wished to discover if it dated to the saine time as similar sites in the Lillooet area. We also had hopes that we may be able to get a dendrochronological cross date with material from Lillooet. Only a modest amount of material was recovered (Appendix III; Table 3), but a radiocarbon date of 1290+ I - 80 B.P. (GaK 5325) did confirm the same time frame. Wood that may be dateable through dendrochronological techniques was also obtained. In the end it was judged that EkRo 18 was the local variant of the Lillooet phenomena. FIGURE Figure 111-16. Contour Map of EkRo 18. In sum, the modest excavations supported the general inferences made from the May 18, 2003-133 N.,., __\u00EF\u00BF\u00BD - ao,\u00EF\u00BF\u00BDu\u00EF\u00BF\u00BD,A Settlement Pattern mouth of regional survey. In addition, the date from Ek.Ro 18 is in agreement with the inference that this is a representative of the Lillooet phenomenon. Thus, almost ail aspects of the PPT expected for the last 2000 years is found at MOC, with the possible exception of the use of mid-elevation root processing areas, which had yet to be archaeologically discovered at the time this survey was completed. Figure IIl-17 summarizes the observed settlement pattern. FIGURE Figure IIl-17. Observed Settlement Pattern at the Mouth of the Chilcotin. May 18, 2003-134 Eagle Lake Quad.rat Survey Statistical Analysis of Major Cultural Variables A total of 43 quadrats were sampled at Eagle Lake, of which six are repeats (sampling with replacement), so that actually 37 separate quadrats were surveyed. Of the total of 43 sampled quadrats, 22 are grassland quadrats (G-quads) and 21 are forested quadrats (Q-quads). The grassland quadrats yielded a total of 36 prehistoric sites, and the forested quadrats a total of 15 prehistoric sites (fable III- 5). A total of 175 cultural depressions, of which 38 are housepits, 26 are roasting pits, and 111 are cachepits, were discovered. About 2045 stone tools and flakes were collected and 34 culturally modified trees were recorded, the trees only being noted in the 1983 survey since we were unaware of their existence in 1979. In ali we believe we obtained an excellent sample of both lithic scatter and cultural depression sites. All 23 quadrats which had prehistoric sites present are illustrated as sketch maps in Figure III-18a to III-18f. The quadrat locations are shown in Figure III-19. The average number of sites per quadrat is higher in the grassland quadrats (1.68 sites i quad) than in the forested quad rats (1.00 sites/ quad), but the number of cultural depressions is quite a bit higher in the forested quadrats than in the grassland quadrats (5.57 versus 3.05). House pits and cachepits (1.19 and 3.7 6 / quad) are more abundant in the forested quadrats than in the grassland quadrats (0.59 and 1.82 / quad, May 18, 2003-135 QUADRAT CULTURAL FEATURES (1979) Sites Cult. Roast. Cache Lithics C.M.T. G l l 10 G 2 4 4 1 3 8 G 3 (22) 1 85 G 4 1 1 1 G 5 2 1 1 9 .l. G 6 1 19 G 7 1 7 G 8 (31) 0G 9 ( 1) 1 1 1 G 10 (8) 0G 11 (14) 2 1 1 ' '1.\u00EF\u00BF\u00BD GRASSLAND G 12 (16} 1 37 QUADRATS G 13 (18) 6 l7 3 1 13 G 14 (20) 1 2 1 1 G 15 (21) 0G 16 ( 2 6) 2 1 1 111 G 17 (27) 1 6 5 1 1.!. G 18 (32) 1 4 1 1 2 434 3 G 19 (34) 1 2 G 20 (44) 9 29 2 8 19 211 1 G 21* 0 G 22* 1 85 -\u00EF\u00BF\u00BD--------------------------------------TOTALS 37 67 13 14 40 1003 33 MEAN (n=22) 1.7 3.0 0.59 0.64 1.8 45.6 1.5 ------------------\u00EF\u00BF\u00BD-----------------------------Q 2 4 5 1 4 Q 3 4 44 ,.., 1 L, \u00EF\u00BF\u00BD 23 Q 4 0 Q 5 0 Q 6 0 Q 7 2 16 3 1 12 Q 9 0 Q 10 0 FORESTED Q 11 0 QUADRATS Q 12 1 21 1 6 14 l .l. Q 13 2 4 4 Q 15* 0 Q 17 1 4 Q 19 1 18 4 14 1122 1 Q 23* 4 5 1 4 Q 24 0 Q 25 0 Q 28 0 Q 29* 0 Q 33 0 Q 35* 2 4 4 --------------------------------- ----------------TOTALS 21 117 25 13 79 1127 1 MEANS (n=21) LO 5.6 1.2 0.62 3.8 53.7 0.048 May 18, 2003-136 g.N t \u00E2\u0080\u00A2,ag.N t .g.N ag.N t I (I/ I I I j I I I 0 !10 100m e Cultural Oeprwsion 0 Tree Lile , , \ Slope Direction .P Small Cache Pit Depr-.ona & , , , '\u00EF\u00BF\u00BDo',, 'i\u00EF\u00BF\u00BDm OT-Line,,{ Slopel:llrection WMarshSmall Cache Pi! Depr-ions 41 Cultural Depression \u00C2\u00B7:\u00C2\u00B7\u00C2\u00A5:-.. l\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD,Pi(,\u00EF\u00BF\u00BD\u00C2\u00B7\u00C2\u00B7 .. \u00C2\u00B7 .. : . . ... . ... Sp\u00EF\u00BF\u00BDtJHt,t\u00EF\u00BF\u00BD-,.\u00C2\u00A5_-\u00C2\u00B7\u00C2\u00B7 .. : ... : -: ::\u00C2\u00B7:. . \u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7.\u00C2\u00B7\u00C2\u00B7.\u00C2\u00B7\\u00EF\u00BF\u00BD:<\u00C2\u00B7\u00E2\u0080\u00A2./\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7. \u00E2\u0080\u00A2 \u00C2\u00B7 .. \u00C2\u00B7 .. \u00C2\u00B7.' \u00E2\u0080\u00A2:. .. : ... \u00C2\u00B7T\T-B,C SITE! 1 Grid B-2 @ Ulhic Scatter 1,2.... Artlact Locations 0 mag.N mag.N srre 1 GridA-3 !50 611, ...J , ' Old Shoraline .:.:.\u00EF\u00BF\u00BD : \u00C2\u00B7\u00C2\u00B7.\u00C2\u00B7:.: \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00C2\u00B7,1 . \u00C2\u00B7.\u00C2\u00B7 . \u00C2\u00B7\u00C2\u00B7.\u00C2\u00B7:\u00C2\u00B7\u00C2\u00B7 .\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7./\u00C2\u00B7::\u00EF\u00BF\u00BD:::\u00EF\u00BF\u00BD -\u00EF\u00BF\u00BD:\u00EF\u00BF\u00BD:\u00EF\u00BF\u00BD'.\u00C2\u00B7::\u00C2\u00B7;.:\u00C2\u00B7 .. . SITE 1 GridA-3 SITE1 Grid X-3 100m 81TE1 GndX-3 . . \u00C2\u00B7\u00C2\u00B7. \u00C2\u00B7 . 0 T (\J 0 NO. OF P!T FEATURES 1II'-i2. Figure \u00EF\u00BF\u00BD- Plot of the number of pit features per quadrat by distance from Chil.ko River, a: w > a: 0 ::.:: ..J :c: u :E 0 a: i.,.. (/) w a: 1-w ::E 114 I I I \u00C2\u00B7\u00EF\u00BF\u00BD (\J a II II ro t0 v,. (\J it) \u00E2\u0080\u00A2 N N I!! V w (IJ Iii Ill N 0 V C\I 0 0 0 N LI.I U) ::.:: 4 ..J LI.I -o J (!I 0 \u00EF\u00BF\u00BD \u00EF\u00BF\u00BD w r<) t :::: 0 a: LL \u00EF\u00BF\u00BD 1/) I a: rI-LI.I :::E co 0 0 IO V ,0 0 0 1J!'\" _;i\u00EF\u00BF\u00BD NO. OF PIT FEATURES Figure\u00EF\u00BF\u00BD Plot of number of pit features per quadrat by distance from Eagle Lake. All quadrats greater than 2000 metres from the Chilko River. 116 / !' The general relationship between features and distance from lakes and rivers now seems to be well accepted, although the grounds for this are not always clear (Helmer 1977a: Klassen and Ridington 1998:18). In our case, this was to be expected from the ethnographic accounts because of the important fish present in some rivers and some lakes. Given that sites with pit features include winter habitation sites and that habitation sites will have non-habitation pit features clustered near them, this pattern is expected by both PPT (habitation sites expected close to the Chilko River) and Athapaskan (habitation sites expected adjacent to the lake, but not close to the shore) settlement patterns. Water is also stated to be used in roasting roots (T eit 1900:237; 1909a:517,780) so that roasting pits would be expected to be located close to water s ources, as well, as is demonstrated later (Klassen 1997). FIGURE Figure III-22. Plot of the number of pit features per quadrat by distance from Chilko River. FIGURE Figure 111-23. Plot of number of pit features per quadrat by distance from Eagle Lake. All quadrats are further than 2000 meters from the Chilko River. Small lithic scatter sites are much more common on the grasslands on the north shore of the lake, showing another concentration of activity. This concentration is not simply the result of surface exposure for two reasons. First, the concentration is found on the north, but not the south side. Second, these sites are not found in forested May 18, 2003-143 quadrats, even though many forested quadrats did have open areas within them. The high number of sites at either end of the lake (Figure 111-21) may be indicative of the importance of water outflow locations, perhaps for fish trapping purposes (Kokanee, mountain whitefish, suckers) while the tendency for housepit sites to be more common in forested areas may be typical of sheltered, winter habitation location preferences indicated by both postulated settlement patterns. Site Classification and Comparison One of the most basic questions is how much archaeological material is there in the Eagle Lake area? Because of the regional sampling approach used, we can answer this question with some confidence as seen in Table III-7. Here, in addition to the interquartile ranges of the seven classes of archaeological material found in the quadrats surveyed are listed, confidence intervals for the predicted values in the survey area are also given in terms of expected material per quadrat. Note that a wide range of material was found with a surprisingly large amount of pit features but, as mentioned before, disappointingly small amounts of lithic material. Table 111-8 lists the prehistoric sites found during the quadrat survey with a brief summary of what was found there, including the variables used in Table 111-7. These two tables are used both for summary comparisons with other surveys, but also in further analyses. May 18, 2003-144 SITE SIZE # # lf # # # # # # 1979 Des, RPTS Gl-1 36000 0 0 0 0 0 0 10 (1983) G2-1 81 1 1 0 0 0 0 0 0 0 (1983) G2-2 7500 0 0 0 0 0 1 0 500* 0 (1983) J_ G2-3 5625 0 0 0 0 0 0 25* 15* 0 (1983) G2-4 3750 3 0 0 3 0 4 0 0 0 (1983) G3-1 150 0 0 0 0 2 83 0 0 0 Q22-1 G4-l 30 , 0 1 0 0 0 100* 0 0 (1983) .l. G5-l 48 0 0 0 0 l 8 0 0 0 (1983) GS-2 3 \" 0 0 1 0 0 0 0 0 {1983) J_ G6-1 140000 0 0 0 0 0 0 0 0 19 (1983) G7-1 1000 0 0 0 0 0 7 0 0 0 (1983) G9-1 10 1 0 1 0 0 0 100* 0 0 Ql-1 Gll-1 81 1 1 0 0 0 0 10* 0 0 Q14-1 Gll-2 160 0 0 0 0 2 11 0 0 0 Q14-2 G12-1 3750 0 0 0 0 5 32 0 0 0 Q16 1G13-1 1500 10 1 0 9 0 0 0 0 0 Ql8-1 G13-2 9 1 0 1 0 0 0 100* 0 0 Q18-2 G13-3 48 3 0 0 3 0 0 0 0 0 Ql8-3 G13-4 81 1 1 0 0 0 0 0 0 0 Q18-4 G13-5 15 1 0 0 1 0 0 0 0 Q18-5 G13-6 81 1 1 0 0 0 0 0 0 0 Ql8-6 G14-1 42 2 0 1 1 0 0 100* 0 0 Q20-1 Gl6-1 5 1 0 0 1 1 0 0 0 0 Q2 6-] G16-2 9375 0 0 0 0 2 108 0 0 0 Q26-2 G17-1 800 6 5 1 0 2 0 100* 100* 0 Q27-l Gl8-l 5000 4 1 1 2 37 407 1000* 30* 3 Q32-1 G19-1 2000 0 0 0 ,, 0 2 0 0 0 Q34 1 \) G20-1 20000 11 1 7 3 0 0 1000* 0 0 Q44** G20-2 2250 0 0 0 0 3 119 0 0 0 Q44 G20-3 480 l 0 0 1 15 73 0 0 0 Q44 G20-4 200 1 0 0 l 0 1 0 0 0 Q44 G20-5 2750 4 l 0 3 0 0 0 0 0 Q44 G20-6 750 4 0 0 4 0 0 0 0 1 Q44 G20-7 32 2 0 0 2 0 0 0 0 0 Q44 G20-8 800 5 0 0 5 0 0 0 0 0 Q44 G20-9 28 1 0 1 0 0 0 100\" 0 0 Q44 Q2-1 104 2 0 1 1 0 0 100* 0 0 Q2 l Q2-2 9 1 0 0 l 0 0 0 0 0 Q2-2 Q2-3 9 1 0 0 l 0 0 0 0 0 Q2-3 Q2-4 9 1 0 0 l 0 0 0 0 0 Q2-4 J_ Q3-l 250000 34 16 0 18 0 0 800* 0 0 Q3 l Q3-2 9 1 0 0 1 0 0 0 0 0 Q3-2 Q3-3 1500 3 2 0 l 0 0 0 0 0 Q3-3 Q3-4 5250 6 3 0 3 0 0 0 0 0 Q3-4 Q7-1 250000 14 2 1 11 0 0 100* 0 0 Q7-1 Q7-2 200 2 1 0 l 0 0 0 0 0 Q7-2 Ql2-l 2580 21 1 6 14 0 1 1000* 0 0 Q12-l Ql3-1 625 3 0 0 3 0 0 0 0 0 Ql3-1 Ql3-2 30 1 0 0 l 0 0 0 0 0 Ql3-2 Q17-1 1 0 0 0 0 0 4 0 0 0 Ql 7-1 Q19-1 40000 18 0 4 14 57 1065 400* c:;* 1 Q19 1 May 18, 2003-145 * = APPROXIMATE** Quadrat 44 was selected in 1979 but not surveyed until 1983.Table III-8. Eagle Lake Quadrat Site Summaries (1979 and 1983 surveys). One of the obvious comparisons to make is between these tables and one derived from Pokotylo's dissertation (1978) on the 1976 survey at Upper Hat Creek, also using 400 m quadrats and similar survey procedures. We have mentioned previously that the Eagle Lake area in many ways was similar and we expected, at least in part, similar material. A summary of the Upper Hat Creek survey is given in Table Ill-9. Note that no housepits or cachepits were identified in 1976 at Upper Hat Creek. The best comparison of material would be between the Grassland quadrats from Eagle Lake which excluded quadrats near the Chilko River (Table IIl-10), an important salmon resource not comparable with anything at Upper Hat Creek. We have grouped together Pokotylo' s forested and grassland quadrats, as many Eagle Lake Grassland quadrats were at least partially forested and this combination only occurred at Upper Hat Creek in the forested sampling stratum. TABLE Table III-9. Upper Hat Creek Quadrat Survey Summary TABLE Table III-10. Upper Hat Creek Quadrat Surveys compared with Eagle Lake Quadrats. May 18, 2003-146 n of n Table III 9 Uppe\u00EF\u00BF\u00BD Ha\u00EF\u00BF\u00BD Creek 1976 Survey Forest and Grassland Quadrats Combined (n=44) of Formed n of Proj r1 of \u00EF\u00BF\u00BD .\u00C2\u00A3ill Tools Debitaae Pits Poim:.s L.S.Median 0 0 Third Quartile 2 0 Mean 1.82 0.75 ----\u00EF\u00BF\u00BD--\u00EF\u00BF\u00BD-\u00EF\u00BF\u00BD---\u00EF\u00BF\u00BD \u00EF\u00BF\u00BD-Total 80 33 0 0 11 446 17.6 2940 ----\u00EF\u00BF\u00BD--\u00EF\u00BF\u00BD-----\u00EF\u00BF\u00BD 773 129355 0 0 0.296 0 0 \" n ..L \" 7 --\u00EF\u00BF\u00BD-\u00EF\u00BF\u00BD--13 85 0 l L52 67 Table III 10 Eagle Lake \"Open Quadrats !i Quadrats with some open areas, not adjacent t:.o Chilko River \u00EF\u00BF\u00BD Sites L.S. RJ:..,_ 1 1 1 8 0 14 2 1 16 1 1 20 1 21 0 22 1 26 2 1 ..L 27 1 30* 2 1 31 0 32 1 34 1 1 ---------- \u00EF\u00BF\u00BD-\u00EF\u00BF\u00BD--\u00EF\u00BF\u00BD-Median 1 0 0 Third Quartile 2 ' 0 .L Mean 1.08 0.38 0.08 Total 13 c:: \" .J .L 30* repeat of 26. ( 1979 Sample) Vonnod Proj. r- of r\u00E2\u0080\u00A2 cf n or C.P. H.P. 'i 100] s \u00EF\u00BF\u00BD Pc i nls C. i'. lL..l:...,_ p. p. 1 l 1 2 9 l 1 4 37 1 1 l l l 2 87 1 3 138 l 1 6 c:: 1 .L 1 3 138 1 21 203 4 2 1 2 ----- ---------0 0 1 6 0 1 1 3 130 0 0.23 0.23 2.8 46 0.46 3 3 .., ,, 600 6 .) I 0 ,, \J 0 \"\u00EF\u00BF\u00BD,, L .5 , ..) 0 1 0.54 7 0 0 r, ') \"l u . L,..) 3 A cursory look at this table shows much higher mean density of iithics at Upper Hat Creek. It is in the debitage category that the difference is most striking with a mean value of 46 items per quadrat at Eagle Lake and nearly 3000 for Hat Creek. On the other hand, the values for the median and upper quartile number of lithic scatter sites per quadrat are the same and when tested using the Wilcoxon two sample test (lots of ties) or Chi square are not significantly different at the .10 level. This appears to be the case even though the mean number of lithic scatter sites is 0.50 at Eagle Lake and 1.52 at Hat Creek. On inspection of the quadrat information, differences are seen in that at Hat Creek, a relatively small number of quadrats has a great deal of material, and a large number of lithic scatter sites, while no Eagle Lake quadrat in similar environmental situation had more than 2 lithic scatter sites and none were of any great size. Two types of cultural depressions occurred at different rates in the two areas, with mean rates of housepits of 0.60 per quad rat at Eagle Lake and 0.0 at Hat Creek and a cache pit mean of 1.8 with 0.0 at Hat Creek. Since it is apparent that cachepits were not recognized at Hat Creek it is unclear whether this latter difference means that they were absent there, or generally not identified. Given its distance from significant salmon resources and relatively high elevation, prehistoric house pits would not be expected to occur in any numbers at Upper Hat Creek, and were not identified there. Roasting pits, on the other hand, had a mean occurrence of 0.75 at Hat Creek and only slightly less frequent at 0.6 in comparable quadrats at Eagle Lake. Given their similar elevations (circa 1050 m (3350 ft) to 1350 m (4500 ft) at Upper Hat Creek, 1129m (3700 ft) to 1280 ( 4200 ft) for Grassland Quadrats at Eagle Lake) and the presence of the important Balsarnroot sunflower in both areas (although more abundant at Upper Hat Creek) it is May 18, 2003-147 not surprising that similar numbers (and we will show below, similar sizes) of roasting pits occur in both areas. The two factors that probably account for most of the differences observed between the two areas are the lack of older, Pre-PPT lithic material at Eagle Lake which are important at Hat Creek and the presence of habitation sites at E agle Lake, (habitation sites are essentially absent at Upper Hat Creek, at least those with the h ousepits of the PPT). In 1980 Matson (Matson et al. 1980:44-48) judged that there were greater differences between Hat Creek and Eagle Lake than were expected previous to field work and argued that more quadrats away from the lake in open areas were needed to resolve this issue leading to the additional quadrats surveyed in that environment in 1983. Although the numbers involved in the comparisons used did not change greatly, with the additional grassland quadrats, the differences no longer appear so large and are now accountable. Although only about one third of the 1979 Eagie Lake quad rats can be easiiy compared with those from Hat Creek (and about one half after the additional 1983 survey) about two thirds of the 1979 and 1983 surveys can be compared with the work at the Mouth of the Chilcotin. Both survey areas are near major salmon streams and include open areas. The Mouth of the Chilcotin (Table III-11) can be compared with most of the Eagle Lake project quadrats by eliminating from the latter only those in located in thick forest and at high elevations (Table III-12). Although the median number of sites per quadrat in Table III-12 is the same, the third quartile and means show more sites at the Mouth of the Chilcotin. However, TABLE May 18, 2003-148 Table III-:..1 Mouth of Chilcotin Grassland Quadrats Summary 2 6 1 3 2 3 0 0 0 0 0 4 6 3 2 l 5 ' 0 0 l J_ 6 1 1 0 0 0 7 ' 0 1 0 .L 8 l 0 0 0 1 9 5 1 ') 2 L 10 5 0 5 1 11 0 0 0 0 0 1 \u00EF\u00BF\u00BD - L 6 0 0 2 4 13 1 0 0 (5 Repeated) ----------------------------Median 1 0 l 1 ------- ----------------Interquartile Range 1-6 0-1 0-3 0-2-------\u00EF\u00BF\u00BD------------------Mean 3 .17 0.5 l .46 1 \u00EF\u00BF\u00BD23 ----------------------------6 0 9 3 0 0 0 ':, 0 0 1':, J J 0 6 ---\u00C2\u00B7-3 ! I 46 0 8 6 0 ::., 1 C lu ,--:b (} 33 6 0-? 6 1 3 A-0 - i ./_ IS 0 2 6 J 40 6 n u 0 I 7 ':, 0 0 I:() Ii 3. 1 Coclcx 17 D 8 J 1 () f} /) 1 0 () () 1 r;l 1() (J IS P,oj. :)cbit. Po:nl:s 0 0 6 0 0 ? i ':,7 1 / 3 0 [) D (j 0 () / () 1 B ,, \" 3 0 0-6Table III 12 Eag2..e Lake Q'Jadrat.s comparable with Mouth of Chilcotin, Open and Near River 1 l 0 1 0 2 4 0 1 3 0 3 10 0 0 3 7 7 2 0 0 0 2 8 0 0 0 0 0 12 1 0 0 0 1 13 ,., 0 0 2 0 L, 14 2 1 0 0 1 16 1 1 0 0 0 17 1 1 0 0 0 18 6 0 1 ,., 3 L, 19 2 ,., 0 0 0 L, 20 1 0 0 1 0 21 0 0 0 0 0 22 1 1 0 0 0 23* 4 0 1 3 0 25 0 0 0 0 0 26 2 1 0 1 0 \u00EF\u00BF\u00BD7 1 0 0 0 1 L , J_ 3 0* ,., 1 0 l 0L, 31 0 0 0 0 032 1 0 0 0 !34 1 1 0 0 0J_ _l_ 35* 2 0 0 2 0--------------------------------Median 1 0 0 0 0 Third Quartile 2 0 0 1 0 Mean L96 0.38 0.17 0.75 0.67 -------------------------------Total 47 9 4 18 1 r J. 0 (1979 0 0 13 3 0 1 0 1 0 0 6 0 0 0 0 0 0 0 s 0 0 1 0 0 0 0 1\u00EF\u00BF\u00BD2S 30 Sample) i6 12 0 20 4 0 0 0 1 () 1 \u00E2\u0080\u00A2 !'I 0 0 1 0 0 () (' .! / () \u00C2\u00B710 0 3.8 91 1 l 0 1 0 0 0 0 0 0 3 0 0 0 0 () 0 1 () 0 0 0 0. t161 0 Formed Too s 0 0 0 0 0 0 2 4 0 0 '']1 f J 0 2 I' ti (J 3 ?, 3 (\ )1 () 0 0 r1. \u00EF\u00BF\u00BD -110 Spa 11 'l\"o --- --- -ls ---Dcbil. 1 0 0 0 0 1 0 9 3 \"/ I) n\"1393 (\u00E2\u0080\u00A2 \" (J 8 I() () l?B 6 1?8 () ?03 ,., / [) --0 11 83. /.199:i. * Repeat Quadrat. Q23 repeat of Q2, Q30 repeat of Q26, Q 3\u00EF\u00BF\u00BD repeat of Ql].i'roj. Poi nls 0 (' ,) 0 0 0 0 0 0 (J ?7 () () 0 (l 0 () ,1 0 0 () () 1 .?9 3 1 Table IIl-lL Mouth of the Chilcotin Summary. TABLE Table IIl-12. Eagle Lake Quadrats near Chilko River and in open environs. this difference is not significant according to the Wilcoxon two sample test. Moreover, the relative abundances of the various kinds of sites and pit features are similar between the two areas, with cachepits the most common, housepits next, and with lithic scatter sites, the least common. Remembering that roasting pits were not recognized during survey at the Mouth of the Chilcotin, their abundance cannot be compared, although later site inspection confirms that they are not common there. However, both cachepits and house pits were at least twice as common at the Mouth of the Chiicotin, as at Eagle Lake. This is in accord with the Teit's (1909a:450,464) report of the MOC being particularly densely settled. Furthermore, the number of both housepits and cachepits per site is much larger at the Mouth of the Chilcotin. Since that survey was located close to the river, and the sites were clustered around sources of water away from the Fraser River and close to the forest, this resulted in a more concentrated pattern. Much of the Eagle Lake survey was located much further from the major salmon resource, which unlike the Fraser is good domestic water and forested, unlike the Fraser at the Mouth of the Chilcotin. Thus, the lower number of these expected PPT sites and occurrence in smaller clusters May 18, 2003-149 at Eagle Lake is also in accord with the PPT settlement pattern expectations. A better c omparison might be the resuits of the Chilko River survey, in which, as reported below, numerous house pit and cachepit sites were located. They occurred in larger clusters along the Chilko River than found in the Eagle Lake quadrat survey. Another difference between Eagle Lake and MOC is the presence of the Pre-PPI ''chert debitage sites\" found at the Mouth of the Chilcotin. No clear Pre-PPT site class was found during the Eagle Lake Quad rat survey--or the Chilko River survey for that matter. The presence of Pre-PPT assemblages at both the Mouth of the Chilcotin and Upper Hat Creek indicates that the Pre-PPT occupation in the Eagle Lake region is less visible than at the other two areas. In terms of conventional site types, according to the tabulation in Table 111-6, the most common site type is a cache pit site, defined as a small depression (or depressions) in the absence of larger pit features or substantial lithics. Eighteen of the 46 sites tabulated in 1979 were cachepit sites. Eighteen housepit sites were found, eight lithic sites and only four roasting pit sites, aithough roasting pits were also located at four other sites. The 1983 survey added 3 more roasting pit sites and 9 more roasting pits. The site with the most dense lithic material (and 14 cache pits and 4 roasting pitfeatures), Q 19-1 (Figure III-18f, 19), is located on a low bank on a bend of the Chilko River. This is a good fit for a riverside fishing site, being similar to the Quadrat 2 sites at the Mouth of the Chilcotin and not at all to any from Upper Hat Creek. Put in another way, the six PPT site types expected to be present at Eagle Lake region, in rough order of importance, are: winter house pit sites, late summer riverside fishing sites, cachepit sites (in at least two different settings), roasting pit sites, and May 18, 2003-150 alpine sites. The last is evaluated as part the Potato Mountain survey summary. As reviewed above, the house pit sites, riverside fishing sites, and roasting pit sites are all present in the expected locations and with the expected attributes in the quadrat survey. The house pit sites, riverside fishing sites and adjacent -to-river cachepits are also very present in the Chilko River survey, as discussed below. In particular, clusters of house pits were found either close (less than 1.5 km of the Chilko River) or adjacent to Eagle Lake. The latter are also within 3 km of the Chilko River. The two largest clusters were found in quadrats Q 3 and G 17, (Figure III-19 and III-18b, 18d) both within 1 km of the river, with a smaller cluster found in Q 7(Figure III-18c), a bit further away from the river. All three of these sites are immediately adjacent to forested areas. On quadrat G 17 (Q 32) (Figure III-18d), a cluster of five housepits were found immediately adjacent to the eastern shore of Eagle Lake, site EkSa 32. Another site, the Shields site, EkSa 13 (Figure III-19), independently recorded in 1979, also consisted of a tight cluster of (6) housepits next to the eastern shore of Eagle Lake. Both of these house pit sites are about 3 km from the Chilko River. Most of these sites include cachepits and others have known cachepit sites within several hundred meters. In terms of being located relatively adjacent to a major salmon stream, with immediate access to both wood and potable water (assuming that Eagle Lake in the past was good water) these sites fit the expectations of the PPT. The tight clustering of housepits often touching each other, also correspond to what is expected for the PPT. Isolated housepits were also found, some near the Chilko River, and some near Eagle Lake. These, of course, could be expected with both the PPT or Athapaskan May 18, 2003-151 settlement patterns. The vast majority of the housepits, though, in the quadrat survey (28/38) were found in the clusters reviewed above. The Chilko River survey shows a similar pattern. A partial check on the location of several site types is found in the survey by Klassen and Ridington (1998), briefly introduced previously. In this wide ranging, but dispersed, study, four survey units were either located with in, or adjacent to, the Eagle Lake survey area. Two of their survey units (LL2, LL3) were on the south side of Eagle Lake and are reviewed in more detail later when we discuss root-roasting pits, since that is the only material that was found in (one of) these units. The other two units (PS4 and SS4) included housepits and so are the subject here. The sampling units used were 200m by 1250m, for a total area of 25 hectares, about 50% larger than the 16 hectare Eagle Lake units. These were all located in what are assumed to be archaeologically rich areas, with the short end oriented to a rich resource area, in three of these cases, either Eagle Lake or Chilko River. The fourth unit, SS4, was orientated across Lingfield creek, near the Chilko river (Figure Ill-19) with the closest part of the sampling unit being about 300m from Chilko river, and the furthest, about 1400m. Thus, the cardinal direction of the orientation of the long side of these units varies. Only unit SS4 is outside the Eagle Lake sampling frame and the furthest portions of that lie within 2 km of it (Figure IIl-19). The procedures used for survey included ground survey techniques not too different from Eagle Lake ones, plus the addition of the blind testing procedure discussed earlier, but no subsurface investigations of cultural depressions, which means the functional identification of many are unclear. May 18, 2003-152 Unit S54 included three house pit sites, one with 6 house pits and 14 smaller depressions (EkSa 145), one with 2 house pits and two smaller depressions (Ek:Sa 146), and one (Ek:Sa 147) with a single depression, a house pit. The other unit (PS4), aligned with Chilko river, included a previously located site (EkSa 116 aka CR 9), and four other sites with cultural depressions-- EkSa 140 with 2 cache or roasting pits, EkSa 141, 2 house pits and 6 cache or roasting pits, EkSa 142, 5 house pits, and 20 cache or roasting pits, and Ek.Sa 143, a cachepit (Klassen and Ridington 1998). Unit PS4 was also discussed earlier under the topic of the Chilko River survey methodology. Although it is very difficult to compare in any quantitative fashion the findings of this survey with the Eagle Lake quadrat survey given the difference in unit size, sampling and field procedures--let alone a sample size of three--the impression is of the same pattern that we found with the quadrat survey. In particular, house pit sites are located close to Chilko river, with larger sites consisting of a number (6 maximum in these four units) of house pits along with a number of smaller cultural depressions, down to sites with single, isolated house pits. Also close to the river are sites that have only smaller depressions, either cachepits or roasting pits. (As mentioned before, the sample units on the south side of Eagle Lake (LL2, LL3) with roasting pits is discussed, separately, later.) Turning to other site types, Quadrat Q 19-1 (Figure Ill-18e, Table III-8) fits all the expectations one would have a riverside fishing site. Given that a sample of one is limiting, we also collected a riverside fishing site (still being used in the 1980s) located during the Chilko River survey, EkSa 33 (aka the Brittany Creek site, CR 92), as described in the section on the Chilko River survey. The roasting pit sites are of the May 18, 2003-153 size, as reviewed later in this volume, and located in the environment where one would expect them according to the PPT settlement pattern. Cachepits in both the PPT and Athapaskan settlement patterns would be located, among other places, adjacent to the major salmon locations. Quadrats Q2, Q3, (Figure III-18b) and G13(Q18) (Figure III-18d) have cachepit sites and are located close to theChilko River (Figure IIl-19. Of course, these may have had other functions as well. And quadrat Q 19 also had numerous (14) cache pits immediately adjacent to the river. Cache pits are also located in other areas, but usually in low numbers and close to Eagle Lake. An interesting exception occurs at the west end of the lake where quadrats Q 2, 13, and G 20 (Figures lll-18b, 18c, and 18b) have large numbers of cachepits (a total of 51) well away from the Chilko River. This end of the lake (Figure 111-19) , however, iswhere the short creek that flows between Fishtrap and Eagle Lakes occurs and is also the spawning location of kokanee, suckers, and mountain whitefish. Thus a concentration of cache pits here is in accord with the importance of this resource location. The different types of cultural depressions tend to co-occur with each other, and to occur in numbers in the Eagle Lake quadrat survey. The median number of house pits per house pit is 1, but the mean is 2.5, with 5 of the 15 sites having 2 or more house pits present (all figures from Table 111-8). Ten of these 15 sites also have cachepits present, 6 have roasting pits and only 4 (all single house pit sites) did not have other types of cultural depressions present. Roasting pits also tended to occur with other types of cultural depressions, and sometimes occurred in substantial clusters with a median of 1 and a mean of 2.2 pits per site. Three of the 12 sites had more than a single May 18, 2003-154 example present (4, 6, and 7 pits), and only four (of 12) sites consisted of isolated roasting pits with no other cultural depressions present. A total of 29 quadrat sites had cachepits presents, with a median occurrence of 2 and a mean of 3.8. Fifteen of these sites had two or more cachepits present and 13 were on sites with other types of cultural depressions present. Cachepit sites without other types of cultural depressions present tended to have fewer cachepits with a median of 1 and a mean of 1. 9 cachepits per site. For all three types, the majority of the depressions are found on sites with more than one pit of that type present and for all three types, the majority of each type are found on sites with other types of depressions present, showing a quite clustered distribution. In summary, the majority of the surfacial archaeology of the Eagle Lake area is a good fit with the expectations of the PPT settlement pattern. This is true whether one looks at the various cultural attributes in isolation, or looks at the associations of cultural depressions with each other, as in the previous paragraph. This fit is in accord with the area being long occupied by this tradition and only relatively recently being replaced by the Athapaskan pattern. Given the overlap between the Athapaskan pattern and the PPT, though, much of the material cited above as fitting the PPT settlement pattern can also be cited as fitting the Athapaskan pattern. Even though further differences might be later discerned between these two traditions, given the relative few details known for Athapaskan patterns in central B.C., both the riverside fishing sites and cache pit sites fit with what we know today about the Athapaskan pattern. Similarly, the roasting pit sites appear to be good members of both settlement patterns. The distinctive lodgepole May 18, 2003-155 pine cambium stripping is abundantly confirmed for the recent historic past by the 1983 survey, but we have no evidence of its prehistoric existence, although it has recently been confirmed for late prehistoric times further north (Prince 2001 ). The Klassen and Ridington (1998) also located two areas with bark stripped lodgepole pine trees in their sample units PS4 and S54. Isolated, small house pits, remnants of winter base camps, also occur in both settlement patterns. In the PPT, though, they would be expected relatively close to salmon streams, while in the Athapaskan pattern, they would be iocated in wooded areas, back from the shore of Eagle Lake. In the quadrat survey, isolated housepits were found in both situations, although they are located between Eagle Lake and the Chilko River, and could be argued to fit both patterns. The most distinctive difference between the two ethnographic settlement patterns is the isolated rectangular lodge, back from the shore of the lake which is expected only in the Athapaskan pattern. Quadrat G 18 (Q 32), (Figure III-19 and III-18a) located in such a place, produced just such remains, the Bear Lake site (EkSa 36), complete with a shallow rectangular cultural depression, roasting pit, two cachepits, and a relatively extensive lithic scatter, as well as stripped lodgepole pines adjacent. Needless to say, this site figures large in the remainder of this report. Further comment is in order in terms of the age of most of the surfacial archaeology. Above we have commented that most of the archaeology fits well with the settlement type and location of the PPT indicating that the Athapaskan occupation is relatively recent. We also commented that unlike the MOC and Upper Hat Creek surveys, no Pre-PPT site classes were identified. Beginning with the latter first, no May 18, 2003-156 microblades were found and only two possible atlatl points were collected out of 30 points collected during the 1979 survey, indicating that most of the lithic material is quite recent. We will report later that microblades were found during excavation at a site that surfacially is PPT and which returned dates indicating that the majority of occupation was in the last 2000 years. This conclusion that almost all Eagle Lake material is not older than the beginning of the PPT is verified by the radiocarbon dates reported under results of excavations and the testing program associated with the Chilko River survey. The second inference that the PPT did not abandon this area too far in the past is supported not only by radiocarbon dates reported further in this volume, but also by the presence of numerous Kamloops side notched points, associated with the Karnloops culture, and dated to between A.D. 800 and 1800 (Sanger 1970: 42-44, group 10; Stryd 1973; Richards and Rousseau 1987). These projectile points associated with the clustered pithouse sites also indicate a PPT occupation within this period. Further, a few multiple side-notched points were also recovered and Richards and Rousseau (1987:43) argue that these are found only between A.D. 1450 and 1800, and these are generally accepted as dating to the later part of the Kamloops culture. Of course, although this style is associated with the PPT and abundant examples were present at the MOC, it may also be true that Athapaskans made them at Eagle Lake. Although these lines of evidence are not of the same strength as radiocarbon dates and associated assemblages they do point to the likelihood of a relative recent replacement of the PPT. May 18, 2003-157 Multidimensional Scaling of Environmental Data from Eagle Lake and Taseko Lakes As a means of providing another perspective on quadrat variability and on the differences between the grassland and forested quadrats, multivariate analyses were undertaken of the quadrats environmental characteristics and of the sites' environmental and cultural characteristics parallel to the analyses carried out earlier for MOC. The quadrats and sites from the Taseko Lakes survey (Magne 1984) are included here since many similarities are expected, but also because knowing the higher altitude and different overali environment of that area can aid in interpreting some of the more problematical locations at Eagle Lake. The first analysis is of the environmental features of the Eagle Lake and the Taseko Lakes quadrats. In this analysis the botanical and physiographic forms that were filled out for each quadrat in the field were converted to presence-absence format using the method shown in Appendix III; Table 3. Although 41 actual variables were used, the presence-absence method means that a total of 73 variables are being read by the computer. The data shown in Appendix III; Table 4 were used to compute a Jaccard's complement distance matrix (Sneath and Sokal 1973) between each pair of the 48 Taseko and Eagle Lake quadrats. No repeat quadrats are used since they would be duplicates of the \"real\" quadrats and no new information would be obtained. The resulting matrix was factored and scaled as per metric multidimensional scaling (Matson and True 197 4; Torgerson 1958) or Principal Coordinate Analysis (Gauch 1982:162; Jongman et al. 1995:152), and yielded nine dimensions, of which the first two May 18, 2003-158 \u00E2\u0080\u00A2 0 G a TASE KO OUADRATS EAGLE LAKE FORESTED OUADRATS quodrots with sites quodrots without sites Eagle Lake grassland Eagle Lake forested T Taseko Lakes JU:-\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD TS4 \u00E2\u0080\u00A2 0 OIi 134 010 0 II EAGLE LAKE GRASSLANDS G8 0 \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD 07 4bQ25 0 017 Eagle Lake a. Toseko Lakes quodrots N = 48 Environmental Doto T scole dimensions I a 11 20.6\"1. + 13.9\"1. iroce Figure\u00EF\u00BF\u00BD Multidimensional scaling of environmental data for Eagle Lake and Taseko Lakes quadrats. are plotted in Figure III-24. These two dimensions combined account for only 34.5% of Trace, or overall variability in the distance matrix, yet appear to separate the major environmental trends quite well. FIGURE Figure III-24. Multidimensional scaling of Quadrats by environments. Figure III-24 indicates that the Eagle Lake grassland quadrats are indeed different from the forested ones, and that outlines can be drawn around the members of the two groups without overlap. Thus we can conclude that our original stratification of the region was appropriate, and that our field forms are useful monitors of the differences. Secondly, the Taseko Lakes quadrats are quite discrete, with some overlap occurring with the Eagle Lake forested quadrats, but not with the grassland ones. This pattern was expected, and it could also be hypothesized that the Taseko Lakes region settlement pattern should be different too, but bear a close resemblance to the pattern in forested environments at Eagle Lake. Observe that Figure III-24 includes symbols for the presence or absence of sites in each quadrat, a variable which was not included in the data that produced the distance matrix. Here it is apparent that, although the separation is not perfect, quadrats with sites tend to occur in the upper right corner of the diagram, those w ithout, in the lower left. In general, then, sites occurred either in open areas, or in forested quadrats along the forest edge. May 18, 2003-159 The wettest, swampiest quadrats of both regions, Q33 at Eagle Lake and TP3, TS6 and TS3 at Taseko Lakes, occur at the extreme left, the driest (G7,G2,G3, etc.) at upper right, and recently burned over areas at bottom center (Q10, Q6, Q4, etc). Again, \"swampiness\" or \"degree of burn\" were not variables directiy input but clearly account for much of the environmental variability in this data set. To summarize, the open grassland environment is towards the upper right, the wettest, most forested area is toward the extreme left, and recently burned areas, at bottom center. These appear to be the three most important factors accounting for the environmental variability among these quadrats. This results are similar to, but far from identical with, those reported earlier for the MOC analysis (Figure (IlI-13) with the main trend of \"elevation\" being the first axis in both cases, albeit at the top of the plot for Eagle Lake/Taseko and at the bottom, for the MOC analysis. This difference a \"reflection\" is strictly arbitrary and carries no significance. In both situations, sites are concentrated in the lower and drier quadrats. For the MOC, the second dimension was \"steepness\". The highest, wettest, and steepest quadrat at Eagle Lake is Q 33 which is at the extreme left of Figure III-24, showing that this trend occurs in both sets of data. It is far from perfect, as the highest and steepest Taseko quadrat, TJ2, is near the center of the plot The multidimensional scaling analysis of environmental data for the Eagle Lake/Taseko sites (Appendix III; Table 5) is shown in Figure IIl-25. Here the first two dimensions shown account for a total of 32.6% of Trace, and again outlines can be drawn without overlap around the groups of \"grassland\" and \"forested\" sites. May 18, 2003-160 1 38 ii ASPEN/PINEGRASS COMMUNITY .02-2 G RASSLAND SITES 02\u00C2\u00B71 \u00E2\u0080\u00A2 FORESTED SITES HIGH EXPOSURE 0 * Gl8\u00E2\u0080\u00A21 G2\u00C2\u00B71o O Gl9\u00C2\u00B7t0 0 O Gl2\u00C2\u00B71 G7\u00C2\u00B71 G2\u00C2\u00B74o Ea gle Lake Groulond Eagle Lake Forested Toseko Lakes G6\u00C2\u00B7t0 * Pl-I * Pl\u00C2\u00B72 S1\u00C2\u00B73 61\u00C2\u00B74 * e 013-1 02-3 \u00E2\u0080\u00A2 Gl4\u00C2\u00B71 012-1 * 0 S1\u00C2\u00B72 S!H Gl3-4O .07\u00C2\u00B71 03\u00C2\u00B73\u00C2\u00B7 0.,,., \ 03-4 \u00E2\u0080\u00A2Pl-3 * \u00E2\u0080\u00A203-2 * S5\u00C2\u00B7'1 * Pl\u00C2\u00B7'I, LOW EXPOSURE LODGEPOLE PINE COMMUNITY Gl3-!I 0 Eagle Lake 8 Toseko Lakes Situ N \u00E2\u0080\u00A2 66 Environmental Doi o T scale dimensions I 8 II 2u\u00E2\u0080\u00A24 + IL5 \u00E2\u0080\u00A24 trace \u00C2\u00B7 J[.ZS-1 \u00C2\u00B7a\u00C2\u00B7 \u00C2\u00B7 1 1 \u00C2\u00B7 f \u00C2\u00B7 1 a Figure.,..,.. Mu t1 1mens1ona sea 1ng o env1ronmenta ata for Eagle Lake and Taseko sites. FIGURE Figure 111-25. Multidimensional scaling of Quadrat sites by environment. Note here that these outlines largely disregard whether the sites occur in quadrats labelled as grassland or forested, since the site records take into account only the immediate site area. These data are much more localized, being limited to the archaeological sites themselves and more \"precise\" than the quadrat data as they do not include areas without sites, unlike the quadrat anaiysis. The two major axes appear to sort the sites on the bases of degree of exposure and occurrence in lodgepole pine forest, trembling aspen-pinegrass forest, or in open grassland. The Taseko Lakes sites strongly tend to group on the left hand side of the diagram, as forested, low exposure sites, but with a few being relatively highly exposed sites (P2-1, Pl-1, Pl-2, Jl-1). As with the quadrat environmental scaling analysis, the patterns revealed are general yet the overall differences are quite discrete. To sum the environmental analyses, it is apparent that forest and grassland strata are separate environments, and sites occur in both, but are most frequent in grassland quadrats. In terms of immediate site environments, sites are frequently forested, but tend to have mid to high degrees of exposure. The Taseko Lakes region sites are usually forested and environmentally overlap with the Eagle Lake forest stratum, and several Taseko Lake sites occur in areas with quite low exposure. May 18, 2003-161 An important point is that the quadrat environmental data sorts those containing sites quite well from those without sites. Therefore in a hypothetical situation within the Chilcotin where identical environmental data were available, without prior knowledge of site distribution, we would feel confident in predicting which areas should contain sites. Scaling of Cultural Eagie Lake and Taseko Lakes Data To provide a view of overall interrelationships of the sites with respect to the cultural features recorded during the quadrat survey, a multidimensional scaling analysis was undertaken with the raw data shown in Tables III-8 and III-13. These tables include total site size and frequencies of lithic tools and debitage, although these are not broken down into types, and also include estimates of firecracked rock abundance (essentially 100 for each roasting pit) and frequencies of historic items. The final, combined table is shown in Appendix III; Table 6. These variables proved useful in the Taseko Lakes project (Magne 1984), in providing meaningful groupings of sites. Given the extreme range of values for the variables, they were first ranked from Oto 66 (no repeats), essentially a form of standardization, then a City Block (Sneath and Sokal 1973) distance metric was computed between each pair of sites. These values were then double centred and metrically scaled as before (Torgerson 1958; Jongman et al. 1995). The first two dimensions of the scaling analysis are shown in Figure IIl-26. FIGURE May 18, 2003-162 + + + .,e+ 142 -- ---HOUSEPIT ,-\u00EF\u00BF\u00BDzo., SITES 1 \u00E2\u0080\u00A2 + : : I I : OS\u00C2\u00B7!e / I I ' I I / I / I I ! 012\u00E2\u0080\u00A2! I \. ' ' I ------ Q17-10 ,/ ____ ... + \u00E2\u0080\u00A2It + + LARGE SITES WITH PITS t'l, LITHlCS 0 \u00E2\u0080\u00A2 M ---------;!9\u00C2\u00B71o\u00EF\u00BF\u00BD 011-1 I ., , ---------\u00EF\u00BF\u00BD Eaol\u00E2\u0080\u00A2 Lake Qro111ond quodrot Eoolt Lake fore,ttd q\u00E2\u0080\u00A2odrot T 011ko Loku quad rot -n:r-CbFigure Taseko \u00EF\u00BF\u00BD. Multidimensional Lakes sites surficial LARGEII SOME LITHICS Ill SMALL PIT + Eovl\u00E2\u0080\u00A2 Loki 8 ToHko Loh& Situ N\u00E2\u0080\u00A266 Ranked Cultural D010. 10 't'OriOblH T ,colt dimtAliOf'II I 8 U <18.9% + 29.!1% irOCI scaling of Eagle cultural data. Lake and Figure III-26. Multidimensional scaling of Eagle Lake and Taseko L akes sites by cultural attributes. Together these dimensions account for 78.4% of trace of the distance matrix, which is a good solution. On this basis of this analysis, eight types of sites are apparent in the quadrat data. Three sites can be classed as quite large sites with pit features and abundant lithics. These included P2-1 at Fishem Lake, G18-1 (lithic scatter and surface features at the Bear Lake site), and Q19-l, adjacent to the Chilko River. Each one of these is in fact a different kind of site, yet all were intensively occupied. P2-1 is a large, likely Lillooet phenomenon village site, as will be pointed out later (see also Magne 1984), G18-1 is an Athapaskan lodge site with historic and prehistoric components, and Q19-1 is a riverside fishing site. A second site type is of house pit sites, here including S5-1, G20-1, G17-1, Q3-1, Q7-2, and Q12-l. These contain between one and 16 house pits and very few or no lithics. This class includes both the clustered house pit sites expected in the PPT tradition, and some of the isolated pithouse sites found in both cultural traditions. The third class of \"larger pit sites\" is comprised of 15 sites containing both cachepits and roasting pits, while the fourth, \"small pit sites\" class, is comprised o f 22 sites with one or two pits. Both of these two classes include some members of the isolated house pit class. A fifth type of site is exposed lithic scatter sites, which contain lithics but no pit features and totals 12 sites. When this class was first found (Magne and Matson May 18, 2003-163 1984:142-143) we did not see it as very significant, more as a grab-bag of possibly functionally diverse lithic scatter sites. Further analysis in (2000-2001) demonstrated that this group overlaps with a group of sites noted in survey, as lithic reduction sites (see also Magne 1983, 1985a:246). With the exception of the two sites in G20 (Q44), all the Eagle Lake sites are located on the north side of the lake in exposed area and are dominated by lithic extraction and reduction of the cobbles and boulders exposed in the glacial deposits. Some of the flakes removed were truly large, being too heavy to remove. Further reconnaissance in 2000 located yet more of this site class in the same environment. It may well be that these sites do have a variety of functions, but they definitely include that of lithic extraction and usually have a high degree of oversight. These sites are concentrated in the high exposure end of the previous environmental analysis. The two sites on G 20 (Q44), between Fish Trap and Eagle lakes, are in a similar environment, but without so much overview. Two sites from the Taseko survey also are members of this group. The sixth site type consists of the two stripped tree sites (Gl-1 and G6-1) which are fairly discrete on the plot, although the trees, themselves, are well spread out over their quadrats. This class' location close to the lithic scatter sites on the plot is likely to be a function of their being part of the 1983 grassland stratum extension survey which was focussed on the north shore and which is also where the lithic scatter sites are concentrated. The historic Chilcotin sites (Pl-3 and G2-3) form the seventh site type and are relatively closely associated with the lithic scatter sites, stripped tree sites, and small cachepit sites with lithics. This probably occurs because of their small size and few May 18, 2003-164 features. Finally, three sites can be considered small cachepit sites with some lithics (G2-4, G20-4, and S5-5) make up the eighth site class. It is interesting that sites with roasting pits and cache pits turn out to be so indistinguishable and that sites with isolated pit houses turn up in both groups. More data on cache and roasting pits site may permit subdivision of these features, possibly on the basis of pit size, age, and evidence of re-use. Both pit classes are present on all four sides of Eagle Lake, showing no real concentration, except being concentrated close to the lake. Overall we can say that large sites are very distinctive from other sites, especially exposed lithic scatters, and that pit sites, other than the clustered house pit villages, are quite frequent and difficult to differentiate using this limited set of variables. The \"exposed lithic scatter\" sites were likely created by individuals or very small groups, likely in iate spring-fall seasons, perhaps are related to the stripped tree sites and small cachepit sites with lithics, since all of these three kinds of sites occur in grasslands and open areas. The small pit sites were probably formed by task group or nuclear family groups, while the larger pit sites could have been formed by repeated visits of such groups or multi-family cooperative groups. The multiple housepit sites are relatively discrete, and represent single and multiple family winter habitations, while the large sites with pits and lithics can be interpreted as multiple occupations at which a variety of tasks were accomplished. In general, then, the scaling of the sites by these cultural variables was not too informative. The more traditional approach focussing on pit feature type, numbers, and location relative to the lake or river was generally more useful. Note that location May 18, 2003-165 relative to lake or river was not part of the scaiing analysis. The size of site and number of debitage and artifacts is as dependent on ground cover as function, as sites in the forest with thick duff or grass are not going to have surface indications of their lithic assemblage, while those in more exposed, unforested areas, may reveal large numbers of debitage, tools, and therefore, a large site area. Nevertheless, the exposed lithic scatter sites, not a site class expected by any, did appear in this analysis (if only belatedly recognized by the analysts) and is clearly an important component of the archaeological record in this area. The conversion of the raw numbers to ranks, too, may not have contributed to the clarity of the result. The major features, though, of both \"traditional\" and this analysis are similar, with large house pit sites at one end, a variety of cache pit and roasting pits in the middle with lithic scatter sites at the other extreme, including site QI 9-1, the riverside fishing site . Taseko Lake Project Sites Although we have incorporated both the environmental information from the Taseko Lakes Quadrats (Figure III-6, Figure III-7) and some site information in the previous analyses, we have not tabulated the information from the sites there in equivalent detail. Remembering that collection did not occur as a rule in that project, the information available on each site does vary from that available for many Eagle Lake, Potato Mountain, and MOC sites. The basic information from the 11 quadrats surveyed, including site summaries is shown in Table III-13 (Magne 1984:Table 1 ). The location of the Quadrats are seen in Figure III-6 and the quadrats are mapped, including site locations in Figure III-7. May 18, 2003-166 Quadrat N of No. Sit\u00EF\u00BF\u00BD Sl 4 -1-2-3-4S2 0 S3 0 S4 1 ss 5 1 -.L -2-3-4-5S6 0 Pl 3** -1-2-4P2 1 P3 0 Jl 1 J2. Totals 15 Quadrat Mean L36 Quad Median 1 * estimated values;with 1 historic of No. of0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 2 1 1 0 0 2 0 0 0 0 l 0 0 0 0 0 0 0 0 0 0 50* 50* 0 0 \" 0 V .Q_ .Q_ 53 C, -_,,, 4.82 L72 0 0 ** An additionalartifact at 7m 2 \u00E2\u0080\u00A2 No. of 11 2 2 1 6 0 0 0 '7 I 2 4 0 1 0 0 2 0 1 L 1 10* 0 9 _O 39 ,-, c; r .5 \u00EF\u00BF\u00BD ..) '.) 0 historic Table III-13. Taseko Lake Quadrat and Site Summary. No. of No. of '!'ot.al \u00EF\u00BF\u00BD i:'..t:\u00EF\u00BF\u00BD 0 0 3760 0 0 45 0 0 700 0 0 14 0 0 3000 0 0 0 0 0 0 6 0 6000 ') 1 1790 ,l, 0 0 500 0 0 800 0 1 1000 0 0 9 ,.., 0 200 L, 0 0 0 25 0 2413 25 0 2400 0 0 4 0 0 9 1000* 10* 200000 0 0 0 0 0 10000 .Q_ .Q_ 1033 11 223970 94 LO 20361 0 0 1790 site (Pl-3) was present in Pl From Table III-13 it can be seen that the archaeoiogical materiai is largely concentrated in two quadrats, S5 and P2, both located on the shores of Fishem Lake (Figures III-6, 7, ). These are the only quadrats that had either house pits or cachepits, and P2, located at the outlet of Fishem Lake has a truly large site (EhRv 2 aka Site 1), albeit one that was badly disturbed. Because there was so much material on this quadrat, it dominates the quadrat totals. In such a case, the medians are apt to be better May 18, 2003-167 estimates of quadrat averages. The two other house pit sites in S5 have 1 and 2 house pits present, indicating, that with the P2 exception, such house pit sites as exist in this area are likely to be very small. No house pits were found in any other quadrats and only one was noted in the area, outside of the quadrats, near Tuzcha Lake (Magne 1984:71). Thus, for most concerns, it looks like the survey area is at the very limit for winter house pit sites that we associate with the PPT. In this regard, it is interesting that the site at P2 is so large. Magne (1984:71) indicates that the size and number of house pits present makes it likely that this site is a member of the \"Lillooet\" phenomenon which is found during the Plateau and very early Kamloops culture (Hayden 1997, 2000, 2001) and we will later report on the radiocarbon dates that support such an interpretation. During the review of MOC results this phenomenon was discussed in respect to Ek.Ro 18. The presence of EhRv 2 in P2 may indicate that Fishern Lake had a large sockeye salmon run in the past. The second important site class present is the roasting pit site. Five of the quadrats had roasting pits present, an amazing total of 11 of the 15 prehistoric sites had roasting pits present. Balsarnroot sunflower is present on many of the quadrats that have roasting pits present, indicating that this was likely the resource exploited. The mean value for number of roasting pits per quadrat in the Taseko Lakes project is 3.6, or 2.9 if one deletes quadrat P2, compared with 0.6 for both the grassland and forested quadrats at Eagle Lake and 0.3 for the 1976 quadrat survey at Upper Hat Creek. The Potato Mountain survey, however, produced even larger numbers of small roasting pits in the high elevation \"parkland\" stratum, with a median value of 5 roasting pits May 18, 2003-168 (and 17 cachepits) per quadrat. Magne (1984:71) suggests that the larger roasting pit sites in Taseko Lakes were used for many other functions as well, while the smalier ones were used for only root processing. The final site dass is the very limited one of 3 small-to-very-small lithic scatters in Quadrats S4, Pl, and S5. The focus on areas dose to water is supported by the observation that the two quadrats that did not include water CT2, S3) did not have any archaeological material. Thus the Eagle Lake observation about sites and water bodies is supported by the Taseko Lakes Project. Although the Taseko Lakes survey was limited in scope, it has provided additional information about the environment around lakes, close to the forest edge including a surprising large number of roasting pit sites. It gives us a more detailed look at the archaeology of the environment at the edge of thickly treed spruce forest, an environment at the upper edge of the intensely occupied zone. Mav 18, 2003-169 .., Chilko River Survey Results Introduction The Chilko River survey was primarily intended to discover stratified sites with regional culture history significance with a minor goal of extending our knowledge of sites in the area. Some 105 sites were recorded from the outlet of Chilko Lake to 30 km downstream, on both banks of the river. What was sought was a stratified site that was not a housepit site, and so would not be disturbed prehistorically. In this relatively large number of sites no such reasonably good culture history site was found. This was disappointing, but not too important in the end for the major goals of the project. A number of sites from the Chilko River Survey were tested and three charcoal samples were radiocarbon dated. Microblades were located at none of the sites in the Quadrat or Chilko River Survey and only two probable atlatl points found. One would h ave to estimate that we have little evidence of material older than 3000 B.P. The Chilko River Survey is also valuable in terms of giving us a full picture of the nature of site distributions along the Chilko River, the environment with the highest density of sites in the area and verifies the general picture obtained from the quadrats. Because the riverside area was densely occupied prehistorically and because of the less detailed recording criteria and the lack of collecting, 103 separate sites (two sites were later joined with others after further analysis) were found in a relatively smail time. As is seen in the accompanying table (Table 111-14) most sites had pit features with over 340 cachepits reported and an average of about one and a half house pits per site. The survey also included over a dozen lithic scatter sites which were a lmost invariably May 18, 2003-170 found on low open terraces immediately adjacent to the river and often with historic fish camp remains simiiar to Quadrat 19 site L These are almost surely prehistoric salmon fishing camps. The sites very neatly fit into the three categories listed, house pit, cachepit or riverside fishing sites. The ratio of cachepit to house pit sites appears to be about the same as found on the Quadrat survey, slightly more cachepit sites than house pit sites (18 to 16 in the 1979 quadrat survey, 50 to 37 in the Chilko River survey). Lithic scatters were proportionally more common on the quadrat survey. The number of house pits per house pit site is more than twice as many on the river survey as on the Quadrat survey (means 4.0 to 1. 9) as might be expected given the importance of salmon and the PPT pattern of multiple pithouses close to salmon streams which is the expected Pre-Athapaskan pattern in this area. It should be noted that the 37 house pit sites recorded in this 30 km stretch is a minimal estimate, given our concentration on open areas close to the river. In fact, Klassen and Ridington (1998) recorded three other house pit sites in the same stretch, albeit further from the river than we surveyed. Today 40 house pit sites are known in that 30 km and many more undoubtedly exist. We thus have an minimum estimate of 1.33 house pit sites per km of river and an minimum estimate of 5.3 house pits per km of river (4.0 x 1.33). The mean number of house pits per house pit site along the Chilko River (4.0) is very similar to that found in the MOC quadrat survey. Depending on how one treats the repeatedly sampled quadrat (either 41 or 44 MOC housepits), and whether one uses the 50m or 100 m of no surface visible site criteria for definition of separate site ( 9 or 12 May 18, 2003-171 Total Number of Sites \u00EF\u00BF\u00BD No. House pit 37 Cachepit 50 Lithic scatter 13 Single Artifact .1 103 Number of House pits per House pit site: Mean 4.0 Median 2.0 Number of House pits per All Types site: Mean 1.44 Median 0.0 Interquartile 0-1 Number of Cachepits per Cachepit site: Mean 5.0 Median 2.0 Number of Cachepits per All Types site: Mean 3.4 Median 2 Interquartile 1-3 Site Size, in Square meters: Mean 5023 Median 300 Interquartile 15-1700 Table III-14, Summary of Chilko River Survey MOC house pits sites) the mean number of house pits at MOC ranges from 3.4 to 4.9, a distribution which includes the Chilko River estimate. As far as we can tell, the size of May 18, 2003-172 house pit sites along the Chilko River and MOC are the same. In contrast to the larger number of house pits per house pit site along the Chilko River than in the Quadrat survey, the mean number of cachepits per cachepit site is 5.0 along the river compared to the nearly identical 4.79 for the cache pit sites discovered during the 1979 Quadrat survey (and 3.8 for the combined 1979 and 1983 survey results). So, although the house pits occur in larger concentrations along the river, the cachepits occur in the same size groups. If we compare these numbers with those found at MOC, we see more similarities then differences. A total of 99 cachepits on 21 cachepit sites were found in the MOC quadrat survey for a mean of 4.8, not significantly different from the Chilko River survey or the 1979 Eagle Lake Quadrat survey. At the MOC, however, we divided the cachepit sites into three locational variants, Ravine Cachepits sites (mean of 5.2 cachepits), Ecotone cachepit sites (mean of 2.8) and Riverside Cachepit sites (mean of 11.7). One might argue that it is the last class that should be compared with the Chilko River cachepit sites, but since only three such sites are found at the MOC (with 23, 3 and 9 cachepits), the apparent difference in size in not significant. So the reliable numbers are very similar between the three areas. Some sites did not fit in with this three part classification. One was tested, Chilko River site 64, and will be discussed below. Another was a recently abandoned sweat lodge which not only fit informant descriptions (see section by L. Burnard-Hogarth) but also solved the mystery of small piles of river cobbles found in Quadrat Ql 9. In general the survey gave us an increased sample of some of the most common sites found in the quadrat survey and verified the pattern found near Chilko River using the more May 18, 2003-173 intensive quadrat technique. Specifically, we previously iisted six site types expected for the PPT, three of which we expected to be reiatively near the Chilko River, winter pithouse villages, riverside fishing sites, and near-river cachepit sites. These three site types include almost all of the sites found along the Chilko River. The relatively large numbers of house pits found close together near the river are expected only in the PPT settlement pattern, confirming its presence along the Chilko River. The numbers of both house pit and cachepit sites and the number of the pits present are close to that seen in the MOC survey, much closer than between the respective quadrat surveys. In both surveys the riverside fishing site are less common than the other two site types. These observations indicate that the area immediately adjacent to the Chilko River is the most appropriate to compare with, and the most similar to, the MOC Grassland Quad rats and indicate the strong presence of the PPT along the upper Chilko River. The other cache pit and riverside site types are associated with both the PPT and the expected Athapaskan settlement patterns. The large numbers of those two and the house pit site classes confirm the pattern based on the smaller numbers found during the quadrat survey, particularly the very smail numbers of riverside fishing sites, as only two quadrats, Q 12 and Q 19 actually had riverside areas present. This site class is discussed in more detail below under EkSa 5, the Canoe Crossing site. The extent of the Chilko River survey on either bank depended on the topography. The survey area in some cases included wide lower terraces extending as much as 500 meters and in other steep, almost canyon like areas, only extending to 50 meters. The drawbacks of this approach were discussed earlier in the methodology May 18, 2003-174 section when we reviewed Klassen and Ridington' s (1998) overlapping survey. In spite of the relatively large number of sites we discovered none of these that we investigatedfurther turned out to have clear potential to contribute to our knowiedge of the Pre\u00C2\u00ADPPT. Two places with multiple buried soil horizons were found but no archaeological material was found at either location. Of these sites, the ones that appeared to be best suited for culture history work were tested. A total of six sites were tested and a very brief description of each follows below. With the benefit of both hindsight and 20 years more experience, the choice of sites for testing no longer appears optimal. In all cases 1 m x 1 m units were used in excavation. Further details of the testing are available in Matson et al. 1980. Tested Riverside Sites Some fourteen riverside lithic scatter sites were located during the Quadrat and the Chilko River surveys. They shared a number of characteristics (that is most of these sites, had most of these characteristics). These included (by definition) extensive lithic scatters along the Chilko River that are extensive in size and quantity relative to other sites found in the study area. These lithic scatters were located along low level, open, grass covered terraces. Although the absolute height of the terraces varied, these sites were always located on the lowest terrace available that is not currently being seasonally flooded. Very frequently, including EkSa 5, they were used in the 1980s by the Chilcotin as late summer fishing camps. They also very frequently had obvious fish resources adjacent; at EkSa 5, a spawning ground for Chinook salmon existed immediately in front of the site, more typical would be some sort of constriction in the May 18, 2003-175 river. At three or four of these sites, house pits were found, but never in the lithic scatter, usually in timber and usually further from the river than the lithic scatters. Along with these characteristics all four tested had very shallow cultural deposits. All these sites appear to be prehistoric salmon fishing sites. In view of their shallowness of deposits, their frequent use today as aboriginal fishing sites and as camping sites for others, this group of sites appears to have a troubled future. EkSa 5; The C anoe Crossing Site, CR 2 This was the first site to be tested. Located at Canoe Crossing, a wide slow moving (relatively speaking that is) part of the river just four kilometres downstream from Chilko Lake, this site includes two adjacent concentrations of housepits, that included a total of 70 cultural depressions as well as a large lithic scatter located on a grass covered terrace next to the river (Figures IIl-19 and IIl-27). It was on this terrace that the four 1 x 1 m test units were placed. FIGURE Figure III-27. Map of EkSa 5, the Canoe Crossing site. Very little material was recovered from these units, which were excavated in 5 cm, levels to a maximum depth of 19 cm. Such artifactual material as was recovered (42 items) was found in the top most sandy soil layer, and none in the gravel layers underneath. The sandy layer was usually less than 10 cm deep. In view of the size of the Canoe Crossing site, including many house pits in May 18, 2003-176 another portion of the site and of the size of the lithic scatter, which extends more than 200 by 60 meters, the shallowness of the deposit is disappointing. Unfortunately, shallow deposits appears to be the usual case for this type of site. A total of 28 house pits and 42 cachepits were identified at this site. A total of 10 house pit depressions were over 10m or more in diameter, with three (E, Fl, and I) having the largest diameters of 14 m indicating that this is likely a Lillooet site, although this would need to be confirmed by dating. It is not, however, as good as case the next site discussed. In any event, it is one of the largest PPT sites along the Chiiko River. ElRw 4 (Quiggly Holes; Bidwell Creek Site) This site is much like that at Canoe Crossing, being the other very large site known in this area prior to our investigations. This site was actually to the north of our continuous survey area, and had a total of 169 pit features in the area we mapped (Figures 111-28 and IIl-29). Lithic scatters existed in the terraces between the housepits and the river and on a large slump deposit between the river and terraces. A total of three 1 by 1 metre units were excavated, one on the lowest terrace and two on the \"slump\" next to the river. The material recovered was relatively evenly distributed among the three units, but with only a single excavation unit (No. 3) having cultural material to a depth of 20 cm. Several good looking radiocarbon samples were also collected from this unit, as well as a few pieces of bone. One of the radiocarbon samples was assayed (SFU 16) and gave a date of 280+ /- 80 B.P. FIGURE May 18, 2003-177 CONTOUR JNTERVAL 200 FT PAftT Of 9\u00EF\u00BF\u00BD 0/13 W 0 I KM Jll-\"2._0Figure\u00EF\u00BF\u00BD iocation of ElRw 1\u00C2\u00B7 (From N.T.S. map 92 0/13W\u00EF\u00BF\u00BD NW Corner). 62 /4 0 0 00 0) I-\"' lt L \u00EF\u00BF\u00BD\u00C2\u00B7'-V. 'f)(J) w a:: 1-w ::E g 0 \u00EF\u00BF\u00BDo Figure III-28. Location of ElRw 9, \"Quiggly Holes'' FIGURE Figure 111-29. Map of ElRw 9. Quiggly Holes. This site then is similar in many ways to EkSa 5, consisting both of a large series of house pits including some of very large size (17m in diameter according to Klassen and Ridington 1998, three of 15m in diameter according to our 1979 records), and a riverside lithic scatter site. The actual number of house pits present is unclear, as there is a large number of 4-Sm in diameter depressions that do not appear to be roo_t roasting p its, but without testing it is unclear what function they have. Both Quiggly Holes and Canoe Crossing were used in the 1980s as fishing sites and have been historically. While these groups of large house pits ought to be Pre\u00C2\u00ADAthapaskan and thus dates from them ought to pre-date the Chilcotin, the lithic scatter site locations, if not used today by Chilcotin, were until recently and one might expect were also used previously by non-Athapaskans. A date from the lithic scatter area can only be associated with one of the other cultural traditions by artifact associations, which are not present in sufficient numbers by the test excavations done. The site of EIRw 4 is located at the downstream end of what the river rafters call the \"white water mile\" and a nice but small calm pool exists there. This pool is what is used today by the Chilcotin and poles and dry rock holding pens are found along the pool as well as fish drying camps located up on the terraces. In addition to the Quiggly Holes site itself, there are several other sites in the area, some of which were recorded by Klassen and Ridington (1998). The Quiggly Holes site, itself, a very large site, with May 18, 2003-178 I I 124\u00C2\u00B0o\u00EF\u00BF\u00BD\u00C2\u00B7wf 5 CONTOUR INTERVAL 200 FT ur-3-a 0 I KM Figure --6,. Locations of CR I 64,73 and 92. (From N.T.S. map 92N/16E, SE Corner). Figure 3-1, 4-1 adjoins lower left corner.65 \u00EF\u00BF\u00BD\u00E2\u0080\u00A2 C \u00E2\u0080\u00A2 0 0 contour inter vol= 20 cm datum is orbitrory 20m ---\"\"'-- Chilko River .,: u .. Q. E 0 \u00E2\u0080\u00A2 .. \"O C: 0 \u00E2\u0080\u00A2 0 \u00E2\u0080\u00A2 Q. 0 .. C, \"O 0 .. E .. .. 1983 survey orid extension to 1983 orid / / / / I I .... 0 //\"' ,o o' ,,,,\"\"' /\u00EF\u00BF\u00BD coche of gaff poiu Figure 1* The Brittany Creek Site.Jif-3.f historic fish comp I I I 63 / I i I I ,--I\u00EF\u00BF\u00BD I I I I /\u00EF\u00BF\u00BD C, Cl 0 -.. the presence of enormous house pits, at least 15 meters in diameter, its good condition and dramatic setting make it ideal for an archaeological park. Its size would make any comprehensive archaeological investigation a task of overwhelming magnitude. It is both the largest and most impressive site in the area, and as such does have extra importance. Chilko River No. 92, EkSa 33, the Brittany Creek Site. This site, near the mouth of Brittany Creek, is another riverside lithic scatter site. In 1979 about half of it was surface collected as well as two units excavated in natural layers to a maximum depth of 10 cm where sterile was found (Figures IIl-30, IIl-31 and Table 111-15). In 1985 the rest of the site was collected (Alexander and Matson 1987). The collection and excavation were done to obtain comparable data with Quadrat 19 of the Quadrat survey, particularly for the lithic analysis as well as for possible culture history work. FIGURE Figure III-30. Location of CR 64 (EkSa 34), CR 73 (EkSa 35) and CR 92 (EkSa 33). FIGURE Figure IIl-31. Map of Chilko River 92 (EkSa 33), the Brittany Creek site. The material recovered (Table III-15) was not numerous, but a single charcoal May 18, 2003-179 sample was obtained as well as several small bone fragments. A date of 860\u00C2\u00B1 80BP (SFU 14) was obtained from this sample. As one expects this site was located on a low grassyterrace, and the cultural material was all in the surface sandy material which lay on a sterile gravel layer. Although not located at an obvious fishing location, historic fish camps indicate that it has been used as one recently. This site appears to be typical of riverside lithic scatter sites with abundant surface material but little depth. Of interest is the presence of a stemmed Kavik point found in the surface coliection (See Ormerod, this volume). A multiple side notched p oint was also found on the surface, indicating presence of both projectile point traditions. The radiocarbon date indicates the presence of a Pre-Chilcotin occupation, as suggested for ElRw 4. It also confirms the probable mixed nature of the site and lack of substantial time depth. Quadrat 19, Site 1 (EkSa 27) The archetype of the riverside lithic sites was also tested. This was the only site located on the quadrat survey which appeared to have any Pre-PPT culture history p otential. In addition it was desired to try to obtain some dateable material from this site. Three units (Figure IIl-18f) each 1 by 1 metre were excavated to sterile which was 10 cm with one exception, an ash feature which went to 25 cm. As seen in Table 111-15, not much material was recovered, but 3 charcoal samples were recovered, as well as a few bone fragments. The best charcoal sample was sent off but when combusted turned out to be too small to date. TABLE May 18, 2003-180 ':'able III-15 Chilko River Artifact Summary (Excavations and Surface Collections) Canoe Quigg}y Artifact Class Crossing Ho]es EkSa 5 E]Rw '1Bifaces Small Points. Bi facially Retl Flakes Pieces Equ. Uni facial Ret, Flakes, Narrow Angle Unifacial Ret I Flakes, Stp. Angle Heavy Duty Cutting/Scrap, Cortex Spall Tools Utilized Flakes Misc/ Tools, etc Basalt Debitage Obsidian Debitage Bipolar Debitage Total Tools Total 0 1 0 0 0 0 2 0 0 0 0 0 8 0 35 1073 7 36 0 3 0 7 ! 42 JO / 9 CR 97 EkSa 33 P,xcav. s_ Coll. 0 1 I ' I1) 0 0 ':,3 0 1 9 0 ,, \) 1 2 0 0 1 9 168/* 93 () l 6 *:,::, 1 / 11 l::xca \"' \"'9-1[) (j u () u (} () () l? () 1-:ksa 3 11 i\u00C2\u00B7:ksa 3 ':: Ci\u00EF\u00BF\u00BD 611 Cf\u00EF\u00BF\u00BD '/3 0 () 0 0 '\ \, () \} 1 0 (l () 1 1 0 0 0 0 0 1 6 39 77 () ?O 80 Miscellaneous Tools, Etc,, includes 1 di-ill in Ji]f:?wt;, piece ol grouncl slaL(' 1n Q 79 1, 2 pieces of decorated slate, 5 quart/iLc flakes, 3 \"waisLed\" c;Lor,c\u00C2\u00B7s, ? inci1cd bones frag., 2 polished pebbles, and 2 ce]L [1-agmcnl_s in CH /3, sc\u00E2\u0080\u00A2,:, '!<.1bic 'ii 13 for J',kSc, 33 1 * Figures from Table VI-11 Flakes less than ':,mm in greaLcst: dimension dclclcd for Lhisset only/A J L 0 -11.-x\u00EF\u00BF\u00BDASPEN TREE JUNIPER TREE LODGEPOLE PINE EXCAVATION UNIT FENCE DOWN \u00EF\u00BF\u00BDLOPE 70 0 10 METRES -m--\u00EF\u00BF\u00BD 5'2-. Figure \u00EF\u00BF\u00BD Map of CR IJ 64. Ek.\u00EF\u00BF\u00BD 3 'fTable III-15. Chilko River Excavation and Collection Artifact Summary. In general none of the riverside lithic scatter sites showed the culture history potential that we had hoped for, although other useful information was obtained. In addition to the riverside lithic scatter sites two other sites were also tested, each one in some way unique. Chilko River 64 (EkSa 34) This site located north of Marsh Lake (Figure III-30) was immediately adjacent to the river on a small alluvial terrace in front of a basalt lava flow. Site 64 was partially cut by the road, but was noteworthy in that large mammal bones were present. Two 1 x 1 metre units were excavated (Figure III-32) yielding 44 pieces of debitage and two tools. Large mammal bone were present, however, to the extent of about 230 grns. These appear to be either elk or deer (most likely) bones, all well fragmented. In addition to the flakes and bone, abundant fire fractured rock was also recovered. In excavation unit 1 a concentration of fire fractured rock and bone was discovered in the south west corner. F IGURE Figure III-32. Map of Chilko River 64 (EkSa 34). In general the obvious interpretation of this site as a marrow processing site May 18, 2003-181 ] seems to be supported. Certainly this site is very different from any of the others excavated and a shallow deposit (maximum 17cm). Again the cultura l deposit was in relatively fine alluvium on top of sterile gravels. (Although in this case more fines were found mixed with the gravel than at other sites). Chilko River 73 (EkSa 35) While surveying the Chilko River, a house pit truncated by the river was found near Brittany Creek (Figure III-30). In the exposure several burned pieces of wood were noted. In view of the possibilities of dendrochronological dating and the dating of at least that house pit using radiocarbon we excavated four 1 by I metre units. This site was excavated using natural layers (Figure III-33). The projectile points included one stemmed point (possible Kavik or Klokut point) and a triangular side point with an indented base. Both of these styles are found in Athapaskan sites but not in recent Salish sites (Magne and Matson, 1982, 1987, this volume). Besides lithic artifacts, abundant fish bone and some large bone was recovered. The main purpose of the excavation was amply fulfilled as no less than 17 samples of charcoal were recovered for possible radiocarbon dates as well as two possible dendrodate samples. In none of the four units was the cultural deposit very deep, with the units in the centre of the house pit hitting sterile (gravel again) within 10 cm, but with up to 25 cm of deposit seen near the rims. FIGURE Figure III-33. Map of Chilko River 73 (EkSa 35). May 18, 2003-182 68 C4 C ____ .__ ______ __, 10 m \u00EF\u00BF\u00BD collected in Im square units \u00E2\u0080\u00A2 1984 datum!985 datumJ:rr:-\u00EF\u00BF\u00BD 3t\ Figure\u00EF\u00BF\u00BD. The Fishtrap Lake Site. cultural depressionQ trees \u00EF\u00BF\u00BD slope One of the radiocarbon samples has been processed, giving a date of 360+-80 BP (SFU 15), well within-the expected range of the Athapaskan migration, and likely within the range of living tree chronologies. Both of the large \"dendrodate\" samples were not datable. Although the sample from this site is very limited, the information we obtained in 1979 is in an agreement with what would be expected for a Chilcotin occupation. For this reason the site was re-excavated in 1985 (Alexander and Matson 1987), with two additional 2 one by one meter units but only three more lithics and several additional radiocarbon samples were added to those recovered in 1979. Fishtrap Lake site (EkSb 27, T 84-27) The lithic scatter collected at the Fishtrap Lake site (EkSb 37 or 84-27) is located on a small knoll on the north shore of Fishtrap Lake (Figure lll-34) near Quadrat G 20 (Q44) (Figure III-18\u00C2\u00A3). FIGURE Figure III-34. Map of The Fishtrap Lake Site (84-27; EkSb 37). In 1985 we decided to collect this site which had been located and mapped during 1984 ethnoarchaeological investigations (Alexander et al. 1985). Down slope and to the southeast of the scatter are numerous other historic and prehistoric features which may be part of the same site. They include the remains of three cabins, an outhouse, a storage cellar, two corrals, three drying racks, numerous other historic structures and artifacts, and a second, smaller lithic scatter. Two roasting pits were May 18, 2003-183 located in the area of the lithic scatter (Figure IIl-34). Except for a few historic artifacts there is no evidence of historic activities in the area of the large lithic scatter. There are three important resource locations near this site. To the south is an area typified by numerous small lakes and rugged terrain that has large, dense patches of balsamroot sunflower and was also used in the historic period for the hunting of deer and elk. Northwest of the site, in what is now a dry part of Eagle Lake, is a large historic weir which was used to catch spawning sucker (Catostomus sp.). Finally, to the east is another weir on the stream draining Fishtrap Lake which was used to catch spawning kokanee and the mouth of this stream is also the location that the mountain whitefish (Prosopium williamsoni) spawned. A single 50 cm X 50 cm unit was excavated that revealed a 5 cm cultural layer underlain by sterile aeolian silt. Only 49 small flakes were found in the excavation. No charcoal for radiocarbon dating was recovered. A total of 2075 catalogued items were recovered from the site, 29 of which were historic artifacts. The majority of the iithic collection (72.1 % ) were less than 1 cm in maximum dimension, much smaller than most other surface collections and only 28 were retouched tools. Of the 28 tools, bifaces (18) are more common than unifaces (10) including four Kamloop side-notched points or fragments. Summary of Chilko River Testing Programme Some generalizations can be made about these sites. In none was there any indication of a reasonably deep deposit as sterile was located in all units within 25 or 30 c m with 10 cm being the more usual situation. In none was there any indication of May 18, 2003-184 antiquity in that neither microblades nor atlatl points were recovered although 14 small points or fragments were, The only evidence of \"stratification'' was seen at EiRw 4 where groups of different size housepits suggested progressive occupation of the area. In addition to ElRw 4 and EkSa 5, EjSa 11 (aka CR# 1) actually at the outlet of Chilko Lake is a likely member of the Lillooet Phenomenon. In 1979 we found 18 housepits there with the largest diameter (rim crest) of 14 m, with likely others destroyed by development of a road and an air strip. Ridington and Klassen (Ridington and Klassen 1998; Klassen 2002) report on mitigations efforts after the site was further damaged by the Department of Fisheries and Ocean, who have buildings on the site, and who apparently destroyed surface evidence, at least, of five cultural depressions and damaged others (1998:ii). This site is also a likely Lillooet member. Although the original mitigation effort did not include radiocarbon dating, Ridington and Klassen (1998; Appendix C:237,238) report on two projectile point that are pre-Kamloops, Plateau Horizon, the time of most of the Lillooet phenomenon, supporting such an interpretation. Klassen (2002) reports on later testing and two dates of approximately 2000 BP that are discussed in the dating section. In short none of the sites tested during the Chilko River survey was suitable for further work for the purposes of discovering more about the period prior to the PPT in this area. This does not seem to be a problem with respect to the main goals of the project as old appearing material was absent on this quadrat survey as well. It looks as if the vast majority of the material falls within the last 2000 years, the period of interest. The Pre-PPT culture history problems are important, but the Eagle Lake project will not m ake much of a contribution toward them. May 18, 2003-185 Potato Mounlain :'a1\"k1and QL;adr\u00C2\u00B7aL Summary Site \u00EF\u00BF\u00BD Pl 1 1 2 3 P2 1 3 2 2 3 2 4 12 5 1 6 3 7 5 8 3 9 2 P3 1 (Historic 2 7 3 0 4 P4 1 2 .l 2 0 P5 1 0 2 3 3 .L P6 i 30 2 1 P7 (Sub Alpine Forest and so P8 1 0 2 1 ..L 3 2 4 7 P9 1 (Historic 2 3 4 5 6 7 8 PlO 1 Median 3 Interquartile Range 2-8 Mean 3.89 Median 96% C.I. 2-81 1 0 0 0 0 0 5 site) not surveyed) Campsite} () 0 0 1 2 0 2 2 0 0 3 :1 1otal l\u00EF\u00BF\u00BDoasLjng rl'olal ____ P_i \u00EF\u00BF\u00BDt. s \u00EF\u00BF\u00BD i Ls \u00C2\u00B7133 ' J I !0 (Lit.hie Scactcr\u00C2\u00B7, n::://) 3 ,1 () () ? 06 'l \, 18 0 )9 31 / ()6 { !) l us l Ii L h i c Sc a LL c r , n 3 7 \u00EF\u00BF\u00BD) 0 (l)lus !1i_Lhic ScaLLc-r-, n=?.110) 3 7 8 / ,1 ., ':, 1 3 / 10 / ? 3 1 ., 1 . /1 1i ) 31 I\ 3 ) \u00EF\u00BF\u00BD\ ILilhics (' ./ 0-? I 6 11. / 0-?. I Potato Mountain Quadrat Survev ., The archaeological settlement patterns are very sharp and clear in the Potato Mountains. The Alpine Tundra proper is bare of prehistoric archaeological remains, while the \"Parkland\" stratum contains large numbers of sites with roasting and cache pit features. No prehistoric sites of any sort and only one historic site and one historic campfire were found in the Alpine stratum (fable 111-16). There were a few suspicious depressions in Quadrat A4 (Figure IIl-4) which might be cultural, but were judged to be natural features in the field. Quadrat A4 also contained the only Alpine site of any consequence, a historic deer hunting camp. The presence of this campsite may be explained by the scattered clumps of krummholz alpine fir in this quadrat which gave it less of a \"tundra\" appearance than many other Alpine quadrats. TABLE Table 111-16. Potato Mountain Quadrat Summary. The total lack of prehistoric remains in the Alpine Tundra stratum is somewhat surprising. One would expect at least some use of this area for hunting and mountain potato gathering, even if all the resource processing and camping occurred in the Parkland. The absence of any material may be the result of a small sample (9 or 10 quadrats depending on one's definition) in a scarcely used area, or the result of environmental conditions that made the usual evidence of past activities invisible to our survey. In contrast to the Alpine, every Parkland quadrat had at least one prehistoric site May 18, 2003-186 0 124\u00C2\u00B0 1\u00EF\u00BF\u00BD\u00C2\u00B7 w km contour i\u00EF\u00BF\u00BDlervol = 1000 ft (\"\"\"\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7 c, \u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2 alpine zone :iooo\u00EF\u00BF\u00BD Figure W. Location of Sites Where Large Lithic Scatters were Collected. 44 (fables Ill-16). The median number of sites per quadrat was 3, and the mean was n early 4. The median number of cache pits was 17 and the mean 37 per quadrat. At most sites cache pits outnumbered roasting pits, the latter of which occurred at a mean rate of 11 per quadrat and a median of 5. In total 35 sites and more than 330 pit features were located in the Parkland zone. As we will show at the end of this chapter, both the cachepits and roasting pits differ significantly from those found elsewhere in the Chilcotin. The mean figures are greatly skewed by the presence of P6, a quadrat on the western edge of Echo Lake which had 31 roasting pits and 206 cache pits recorded. Actually this site (EjSb 12 or Echo Ridge site) (Figures III- 4, -36) was not fully contained inside the quadrat and included many more pit features which were recorded by Alexander et al. (i985) the previous year. Based on species abundance, Mountain potatoes (Claytonia lanceolata) must have been the dominant plant resource cooked and stored in these pits. One quadrat (Pl 0) with a single site fell into the only area that had balsamroot. In this case, balsamroot may have been the key resource. Two large lithic scatters (Mountain Pond and Mountain Fan) (Figure III-36) were found in the Parkland during the quadrat survey. Four additional sites produced a few lithics. FIGURE Figure III-35. Location of Potato Mountain and Eagle Lake Lithic Scatters. May 18, 2003-187 85 0 2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 I creek----Tfr-3 ,_.-Figure#. Location of the Sites Test Excavated at the South End of the Potato Mountain Range. 3 o z\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD \"-------'-----' contour interval : \u00EF\u00BF\u00BD00 fl EjSb site ]I-3'1-...-- - trail 46 \ area surveyed Figure ra. Judgemental Survey Area at the North End of thePotato Mountain Range. FIGURE Figure IIl-36. Location of Tested Potato Mountain Sites. Only two historic sites were recorded, although isolated historic artifacts were not infrequent. Outside the surveyed quadrats, the judgemental reconnaissances located 5 additional lithic scatters at the south end of the range and 5 at the north end (Figure 111-37).FIGURE Figure 111-37. North End of Potato Mountain Sites. Physiographic rather than cultural factors appear to be responsible for the fact that most lithic scatters are concentrated in one small area around Middle Mountain. Tnis area has more ground exposure than any of the other Parkland areas and, therefore, lithics have a greater visibility. However, this area did have exceptional overview, both of adjacent Alpine areas and open areas in the lower Parkland, and may have been a staging and campsite area for hunting activities. Potato Mountain Testing Program The Mountain Fan site (EjSb 39), on a small remnant of a fan at the base of Middle Mountain (Figure 111-36), not only had the densest lithic scatter, but also had a number of culture components, and was the only deposit of depth discovered. Some parts of the site (Figure 111-38) have no vegetation or soil over the alluvial gravels while, May 18, 2003-188 wp ) C ::> \"\"il wp \u00E2\u0080\u00A2 N 4 C :, \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 Figure , .... ___ .\u00EF\u00BF\u00BD-) E0wp 2 0 wp of e \u00E2\u0080\u00A21 \u00EF\u00BF\u00BD \u00C2\u00B7, 12 '\" \ ...... A\ 13 ,. wp, 0\ ,. 10 \\u00E2\u0080\u00A2 I '\ ',1$ ,. ., EU 3 ',, 9 \u00E2\u0080\u00A2 I 116 \\u00E2\u0080\u00A2 EU2 owp \u00E2\u0080\u00A2 0:, I wp .. 0 'O .. \ -, , , ... (\ \ \ bore 9round ., / ' 0 \ \ \ 11 \ I -0 ,. \ \ \ 19 \ & \ ,_.,. .. \ \ \ \ \ 119 I \ \ \u00E2\u0080\u00A2 20 \ _(\"\ , , ' I I '\ ,, \ \ I I I I I Site. 0 0I \u00E2\u0080\u00A2 wp of k \u00E2\u0080\u00A2:, \u00E2\u0080\u00A2 E U 75 cultural depreuion artifact location whitebork pine alpine fir juniper kinnikinik boundary between ve9etotion and bore 9round excovotion unit on other areas of the site, grasses, kinnikinnik and smail shrubs cover aeolian deposits of more than 40 cm in depth. Lingfield Creek is currently ca. 5 m below the site. FIGURE Figure III-38. The Mountain Fan Site (P8-3, EjSb 39). Five units were excavated at the site. Two 50 cm X 50 cm test units were initially dug at the site to test the depth of the cultural deposits and check the artifact density. These tests revealed three cultural horizons but a generally low artifact density. Since it is the only known stratified site in the study area and since each cultural horizon contained large quantities of charcoal, two additional 1 m X 1 m unit s were excavated where the soil deposition was greatest. Radiocarbon samples from the two lowest layers produced dates of 960 +/- 80 B.P. (WSU 3374) and 2220 + /- 80 B.P. (WSU 3375), stratigraphically in order. An additional 50 cm X 50 cm unit was also placed in the center of one of the small depressions on the site which confirmed its identification as a roasting pit. This roasting pit has an accumulation of 27 cm of firecracked rock and charcoal in the center indicating a relatively deep initial pit and the charcoal produced a date of 1680 + / - 90 B.P. (WSU 3380). On the surface it appeared to be a shallow, moderately sized pit with an inner rim diameter of 100 cm and a rim crest to rim crest diameter of 240 cm. Evidence indicates that the pit was lined on the bottom with a rock pavement and the cooking fire constructed in the pit rather than outside. May 18, 2003-189 78 I I/ j O ht \"o\\O'fl0 I \"t-t \ OL OK J I 0 0 H G IF OD \u00EF\u00BF\u00BD E 6 0 oc \u00E2\u0080\u00A2 B \"' /0 Figure 21>. The Mountain Pond Site. A total of 426 items were catalogued from the Mountain Fan site: 397 flakes and 27 retouched lithic artifacts; 221 flakes and 20 of the tools from the surface. No points were found at the site, and most (23) of the retouched artifacts were unifacial. The relatively high frequency (6.3%) of retouched artifacts and non-basalt artifacts (6/27) and lack of evidence of primary lithic reduction (only 2 flakes had any cortex) indicate that curated tools were common. The Mountain Pond Site This site (EjSb 54 or P8-1 or 84-14, Part 6) is on a mid-slope bench on the west side of Middle Mountain overlooking Lingfield Creek, the Mountain Fan site, and a large, open area to the west (Figure 111-36). It is in a Parkland environment with an open area of grassy meadow fringed with stunted alpine fir on the west. A small, shallow melt-water pond is located 23 m to the east. The site consists of a lithic scatter and 18 cultural depressions representing both cache and roasting pits (Figure 111-39). FIGURE Figure III-39. The Mountain Pond Site (P8-1, EjSb 54). It is in the area traditionally used by the Redstone band during the historic period and, in fact, a few tin cans were collected from the site indicating a recent occupation. Two 50 cm X 50 cm units were excavated at the site to test the depth of cultural deposits and artifact densities. These revealed a low artifact density and a thin (ca. 3 cm) cultural deposit over sterile aeolian silts. May 18, 2003-190 Only 317 lithics were recovered from the site. There is little evidence of primary reduction at the site. Only four retouched artifacts and no points were recovered. Middle Mountain This site (EjSb 52 or 84-14, Part 4), located at an elevation of 1920 m (6320 ft) is near the top of Middle Mountain at the south end of the Potato Mountain Range (Figure III-36). A large lithic scatter and 12 cache pits were found at the site. The site is in aparkland (alpine-subalpine ecotone) environment consisting of an open area bordered on the north and south by small clusters of stunted alpine fir. In some places the surfaceis bare of vegetation with weathered bedrock exposed, while in other spots grasses or kinnik.innik are established. There is a small melt-water pond ca. 30 m to the northeast and a good overview of alpine and parkland areas to the south. Ethnoarchaeological evidence indicates that this area was traditionally used by the Redstone band as well. They camped on Middle Mountain during the late summer while they collected and processed mountain potatoes, hunted and processed deer, and participated in an Indian-only rodeo near Lingfield Lake. An historic camp is present at the south end of the Middle Mountain site with a hearth, possible drying rack, tin cans, and glass. In addition to being mapped and surface collected (Figure 111-40) five units were excavated on this approximately 27 m X 42 m site. Two adjoining 75 cm X 75 cm units bisected a cache pit, and two adjoining 50 cm X 50 cm units were placed where surface lithic concentrations were high. An additional 1 m X 1 m unit was excavated in a spot with a heavy grass cover. All units revealed ca. 10 cm of cultural deposit underlain by May 18, 2003-191 0 L 5m contour interval= 20 cm. cultural depression \"\"D\".r-YOA2 Figure :.z. The Middle Mountain Site\u00C2\u00AE 72 Al 01 0 0 .., ' b e d h k m n 0 p 2 0 . 0 em ,n q r rrr-4 IFigure\u00EF\u00BF\u00BD- Diagnostic artifacts from the Potato Mountain Range, the Fishtrap Lake Site, and the Brittany Creek Site. Kamloops points: a: EjSb 3; b: EjSb 14; c-f: EkSb 37 (Fishtrap Lake); g-k: EkSa 33 (Brittany Creek, CR 92). Early Ramloops points: 1-m: EjSb 52 (Middle Mountain). Plateau points: n-o: EjSb 52 (Middle Mountain). Shuswap point: p: EjSb 53. Early point: g: EjSb 24. Drills: r: EkSb 37 (Fishtrap Lake); s: EkSa 33 (Brittany Creek, CR 92); t: EjSb 3. bedrock or glacial till. No charcoal for radiocarbon dating was recovered. A single cache pit (Feature A) was excavated at the site with a 75 cm X 150 cm unit which bisected the pit. No artifacts, firecracked rock, or useful charcoal samples were recovered. Excavations confirmed surfacial impressions of a small, shallow pit. The inner rim diameter is only 70 cm and the depth from the rim crest to the bottom of the pit is only 14cm. FIGURE Figure 111-40. The Middle Mountain Site (EjSb 52). A total of 1013 items were catalogued from the site including 154 historic artifacts (primarily glass fragments). There is little evidence of primary reduction at the site. Forty-six (5.4%) are retouched artifacts, with a predominance of unifaces (35). The relatively high frequencies of retouched artifacts and non-basait artifacts indicates high use of curated tools. Four points or point bases were found, all corner- or side-notched points (Figure Ill-41) indicating Kamloops horizon (1200 - 1000 B.P.) and Plateau horizon (2400-1200 B.P.) occupations. FIGURE Figure IIl-4L Potato Mountain Artifacts. Other Tested Roasting Pits P2-3 (EjSb 26) is a 'B' site, a member of a class of sites within the forest fringe May 18, 2003-192 that typically had only one or two roasting pits in contrast to sites in more open areas which typically had more roasting pits and included cache pits. Two of the 'B' sites were tested in 1985. P2-3 site is located in the forest ca. 50 m above the open slope to the east of Lingfield Lake (Figure 111-36) on a small bench above the high, steep bank of a small unnamed stream. The site consists of two large roasting pits. Roasting pit A was tested with a 75 cm X 100 cm unit which extended from the center of the pit to the rim crest. The excavation was done in natural layers and all material screened with 1 / 4 inch mesh. Although no artifacts were recovered, numerous radiocarbon samples were collected, one of which produced a date of 450 + / - 70 B.P. (WSU 3376). Prior to excavation the pit appeared to be a large (140 cm inner rim) pit but relatively shallow in comparison (27 cm) to its width. Excavation confirmed this impression and revealed a high concentration of firecracked rock (ca. 140 lbs or 57 kg) in the unit. Reuse of the pit after initial abandonment is indicated in the rim profile by the presence of a sterile, light-coloured rim underlain by a thick layer of soil with substantial fire-cracked rock and charcoal. P2-9 (EjSb 33) is also a type 'B' site located ca. 65 m further up slope from P2-3 (EjSb 26). It also is next to the same small stream and consists of only two large roasting pits. Feature A was excavated in the same manner as Feature A at P2-3. No artifacts were found but one of many radiocarbon samples from the pit was dated to 615 + I - 80 B.P. (WSU 3373). At the ground surface the pit was very large (245 cm inner rim diameter) for Potato Mountain and relatively shallow (31 c m) for its size. A high concentration of firecracked rock (ca. 85 lbs or 39 kg) was found in the unit. A May 18, 2003-193 sw corner 20 40 60 nw corner excavation organic debris yel\u00EF\u00BF\u00BDowish-brown silt dork brown sill fire crocked rock Jlf-\u00EF\u00BF\u00BDc_ Figure\u00EF\u00BF\u00BD- West Wall Profile From Roasting Pit A at P2-9 {EjSb 33) \u00E2\u0080\u00A2 87 profile, similar to that found at the previous site (Figure III-42) indicates that this pit was also reused. FIGURE Figure 111-42. Profile of Roasting Pit A at P2-9 (EjSb 33). Echo Ridge (EjSb 12 or P6-1) is a very large site with 342 cultural depressions recorded within the quadrat and many more observed outside, a definite type 'A' site. These depressions are on top of a Parkland Ridge which runs parallel to the west shore of Echo Lake (Figure IIl-36). Three roasting pits and one cache pit were tested at this site. All four pits were excavated in the same way as with the previous two sites. No artifacts were found. Radiocarbon samples were removed from two of the roasting pits providing dates of 1710 +/- 90 B.P. (WSU 3372) and 1910 +/- 50 B.P. (WSU 3381). However, there was not enough charcoal in the other two pits for an adequate sample. Nevertheless, the third roasting pit was also dated with a charcoal sample from a pile of firecracked rock and charcoal adjacent to and, presumably, cleaned from the roasting pit. This date of 100 + /-60 B.P. (WSU 3378) suggests a recent use for the pit which is in keeping with the shallow soil deposition in the depression. However, it is possible that the charcoal may date a natural forest fire since it was not deeply buried. The shape and content of the three excavated roasting pits varied considerably. One pit had very little firecracked rock ( ca. 30 lbs or 14 kg) and only small flecks of May 18, 2003-194 charcoal but was the one with a large concentration of firecracked rock and charcoal beside the pit. The second pit had moderate quantities of firecracked rock (50 lbs or 23 kg) and charcoal (enough to date) but the rock was generaliy small in size. The third pit contained large quantities of firecracked rock (143 lbs or 69 kg) and charcoal with many very large rocks. The cache pit at the site contained no charcoal or firecracked rock and only had 13 cm of deposits at the center of the pit. This evidence suggests a relatively recent use o f the pit. No firecracked rock was found nearby to indicate its possible use as aroasting pit. Upon excavation, this cache pit was found to have uneroded walls and nearly vertical sides with a rounded bottom, in marked contrast with the more hemispheric shape of most roasting pits. Radiocarbon dates from six of the roasting pits tested provided dates ranging from 100+ /- 60 B.P. (WSU 3378) to 1910 + /- 50 B.P. (WSU 3381). The high density of sites and pit features agrees with the Chilcotin accounts of the importance of this area. If we extrapolate from our sample to our sampling frame we arrive at a point estimate of about 400 sites, with about 770 roasting pits and 2400 cache pits for the south end of the Potato Mountain alpine zones. Moreover, excavation indicates substantial reuse of the roasting pits, suggesting that this pit estimate is a low estimate for number of root roasting episodes. These large numbers of sites and features, together with the radiocarbon dates indicating that the excavated roasting pits were only in use during the last 2,000 years, point towards a substantial number of people using the area at any one time. May 18, 2003-195 Summary of Regional Survey Studies How can four different survey projects and related testing be briefly summarized? Actually, this is not difficult to do. The Mouth of the Chilcotin project provides the example of the PPT settlement pattern, showing a close agreement with the ethnographic model based on Teit (1900, 1909a). The Eagie Lake and Taseko Lakes quadrat surveys, and the Chilko River surveys show that the majority of the material found fits well within this model. The PPT, then, is strongly present with numerous remains of the winter pithouse viilages of tightly clustered housepits, located near salmon streams and adjacent to good water and wood, both at Eagle Lake and in the Taseko Lakes region. Other PPT site types, such as riverside fishing sites, and cache pit sites are also present in the appropriate locations. Some of these same sites, and other sites, also fit the less well-defined A thapaskan settlement pattern. We will discuss how to untangle this issue in the next chapter. In all cases, though, almost all of the material found fits the pattern expected for the last 2000 or so years, indicating that the surfacial archaeology is dominated by recent material, an inference confirmed bv the numerous recent radiocarbon dates .; reviewed above. When we go to the Alpine areas, we move to material that is not present at MOC, and is not well described in the ethnographic accounts. We do, however, have dating evidence indicating that it, too, also extends back to, but is limited to, the last 2000 years or so. It also appears to be both PPT and Athapaskan, and includes relatively smallroasting pits and cachepits. Also, the mid-elevation roasting pit phenomena was not May 18, 2003-196 identified at the time the MOC was carried out but is present both at Eagle Lake and the Taseko Lakes. These, too, are not well identified in the ethnographic literature, but are well described for the PPT from Upper Hat Creek (Pokotylo 1978; Pokotylo and Froese 1983; Vanags 2000). It appears that the sizes of the roasting pits are highly correlated with the material processed (Peacock 1998) with the smaller ones used for mountain potato and the larger ones for balsamroot sunflower. Matson and Alexander (1990) show this dearly in their comparison which is next summarized with additional information from Peacock (1998). The size of pit features on the Interior Plateau varies considerably (circa 1 to 15m) as shown in Figure III-43. Most of this variation results from different functions. Within roasting pits the amount and kind of roots used influence size, with larger amounts and tougher roots resulting in larger pits. Peacock (1998:153-161) has carried out laboratory investigations which explain why the ethnographic accounts (Teit 1900:236) indicate that the Balsamroot sunflower was cooked for so long. Basically, the root must be cooked for a long time to convert the undigestable inuiin carbohydrate fraction into digestible fructose. In order to hold the heat for up to two days, a large oven must be used. The soil matrix apparently had little influence on pit size, as compare the mean diameters of Hat Creek ( 435 cm) and Potato Mountain (193 cm) roasting pits on similar soils but different roots. Both large and small roasting pits appear to date within the last 2400 years (Pokotylo and Froese 1983; Peacock 1998; this volume). The sole exception to this pattern reported by Peacock (1998:224-225) is an atypical pit, without fire-cracked rock at the Parker Site (EdRi 27) with a radiocarbon date of 3130+ /-100 (no lab# given). The May 18, 2003-197 40 30 %20 10 0.0 4or30\u00EF\u00BF\u00BD %20 10 0.0 20 30[%1\u00EF\u00BF\u00BD 0.0 Potato Mountain n= I07 1.0 2.0 3.0 4.0 5.0 6.0 7.0m Eagle Lake n=l9 1.0 2.0 3.0 4.0 5.0 6.0 7.0m Hat Creek n= 55 1.0 2.0 3.0 4.0 5.0 6.0 7.0m Roasting Pit Diameters Probabilistic Surveys next oldest Canadian in the 27 dates tabulated for the Plateau is 2360+ / -150 (Beta 57 434) from EdRb 44 in Scheidarn Flats near Kamloops (Peacock 1998: 197) . Clearly the root roasting activity, defined as root roasting pits away from housepit sites, is concentrated in the last 2400 years, if not limited to that time span. FIGURE Figure III-43. Roasting Pit Diameters, Probabilistic Surveys. We earlier commented on the observation that in the Eagle Lake quadrat survey roasting pits appeared to be distributed on all sides of the lake, but concentrated close to the shore. In addition to our quadrat survey, Klassen and Ridington (1998:20-25,47-50,115-124) report on two survey units (LL2, LL3) on the south shore of Eagle Lake, each 200m by 1250m and oriented with the narrow end next to the lake and the long direction extending in a general SE direction (but not the same exact orientation). One of these units did not have a verifiable archaeological site (LL3) according to Eagle Lake standards, but the survey of the other (LL2, basically between our Q4 and the lake [Figure III-19]) located four sites, all within 450 m of the lake, with 1, 2, 1, and 3 cultural depressions. Although the procedures used did not allow the determination of function, the sizes, (2.85-3.4m in diameter) and inferred soil matrix indicate that they are roasting pits. Note that to the west of these our Q13 on the edge of Eagle Lake had 4 roasting pits present. These findings support our findings about roasting pits being located along all sides of Eagle Lake and concentrated near the lake. May 18, 2003-198 50 40 30 %20 10 0 30t 20 % I\u00EF\u00BF\u00BD 30t % 20 10 . 0 Potato Mountain n =3270.0 LO 2.0 3.0 4.0m Eagle Lake n= 97 0.0 LO 2.0 3.0 4.0m Mouth of Chilcotin n=l50 0.0 1.0 2:0 3.0 4.0m Cache Pit Diameters Probabilistic Surveys The dominant root at Hat Creek and Eagle Lake was the large (and very tough) Balsarnroot sunflower, and the roasting pit sizes are large in comparison to those found on Potato Mountain, where Mountain Potato was the main resource. It is interesting to note that the roasting pits near the only area on Potato Mountain that had Balsamroot, were larger than those elsewhere (214 compared to 192 cm). Cache pit size (Figure III-44) also appears to vary significantly between areas. Pit features identified as cache pit remnants on Potato Mountain were significantly smaller than those found at Eagle Lake and the Mouth of the Chilcotin. The 112 cm average diameter on Potato Mountain probably reflects the relatively short time spent on the mountain and high transportation costs involved moving material from there to the winter habitations in lower elevations. The average diameters of 231 cm at Eagle Lake and 217 cm at the Mouth of the Chilcotin probably reflect both the greater importance of salmon and Balsamroot as long term stored food and the shorter distances betweenresource location and winter habitations. Lower transportation costs make storing large amounts for winter habitation use economically feasible. FIGURE Figure III-44. Cache Pit diameters, Probabilistic Surveys. Pokotylo and Froese (1983) have suggested that roasting pit s ize may also be related to economic unit size. This is plausible, and certainly the small sizes seen on Potato Mountain point to small groups, but demonstrating that this relationship is the May 18, 2003-199 case in a particular situation is much more difficult. Clearly there are some very large roasting pits at Hat Creek that are suggestive of larger sociai units. The numerous roasting pits (150) recorded in various non-probabilistic surveys at Scheidam Flats near Kamloops have a mean rim crest diameter of 3.51m (Peacock 1998:193) similar to the size seen at Eagle Lake and a full metre less than those found at Upper Hat Creek, even though the main resource roasted in all three areas is assumed to be Balsamroot. Scheidam Flats, though, is at a much lower elevation than Upper Hat Creek with the valley floor at 600 m which may have some bearing on the differences in size. These figures, though, indicate a high degree of similarity in the larger roasting pit sizes between Eagle Lake and Hat Creek, again supporting the PPT nature of much of the archaeological phenomena at Eagle Lake. We do have, though, an extended model of the PPT because of these archaeological discoveries. It is clear that the Eagle Lake area is in the northwestern edge of intensive root processing activity in interior British Columbia, particularly of the important Balsamroot resource. This resource extraction, though, does extend slightly northwest of Eagle Lake, with probably one of densest concentration of large root roasting pits yet known being located west of, and on the southern slopes of, Tatla Hill. This is approximately 25 km northwest of Eagle Lake (Klassen 1997), with elevation ranging from 1040m (3400ft) and 1340m (4400 ft). Although hindered by a lack of subsurface investigation by any means of cultural features, Klassen (1997:77-84) presents a detailed and convincing argument that the bulk of the 80 cultural depressions found in eight sites in his investigation were used for roasting Balsamroot sunflower, among other functions. Balsamroot is present in this area, and these features have a mean diameter May 18, 2003-200 of 3.23m similar to other samples from Eagle Lake and Scheidam Flats, although some cache pits may also be included in this sample (because of the lack of subsurface investigations at Tatla Hill). One of the sites (ElSd 4) found in this survey had fourteen or fifteen cultural depressions present and another (ElSd 5) had 32-35, although this last one was defined in such a way that it extended over 1100 m. All of these sites are located near water, or places that in the past would be expected to h ave surface water in the spring. This impressive concentration of inferred roasting pits supports the presence of this typically PPT activity in the Eagle Lake environs, and their location near sources of water supports this aspect of the Eagie Lake findings as well. Another point of similarity between the Eagle Lake investigations and the PPT is the apparent presence of the Lillooet phenomenon during the late Plateau, early Kamloops Horizons, at the Mouth of the Chilcotin, Eagle Lake, and Taseko Lakes regions, again something only associated with the PPT. We reported above on three possibilities for the Chilko River, ElRw 4 (Quiggly Holes), EkSa 5 (Canoe Crossing) and EjSa 11 (aka CR 1), with the first being the best case and the other two very close to Eagle Lake. EkRo 18 at MOC must be considered to be a good member of this phenomenon because of the radiocarbon date recovered and the size of its housepits (including a lack of small ones) and EhRv 2 at Taseko Lakes is also another likely member. In conclusion, the settlement pattern investigations combined with the surface collections and testing programs that provided important dating information, shows a very strong presence of the PPT in the areas investigated. Interestingly enough, almost all the information indicates this occupation in the last 2000 years, with the earlier May 18, 2003-201 portion of the PPT not being very visible. This patterning, though, overlaps with the proposed Athapaskan settlement pattern, and we next focus on discerning between the two and dating the initiation of the Athapaskan tradition in this area. May 18, 2003-202 DESCRIPTIVE ARCHAEOLOGY Chapter IV ARCHAEOLOGICAL SITE EXCAVATIONS IN THE EAGLE LAKE REGION Three sites were excavated to provide additional materials for the identification and dating of the Athapaskan arrival in the Eagle Lake area. These three sites have already been briefly introduced during the survey section, two representing typical PPT locations and surface remains, and the third, the more hypothetical Chilcotin Athapaskan pattern. The latter is the Bear Lake site (EkSa 36), found during survey of Quadrat 32(G18), and the typical PPT pithouse villages are the Shields site (EkSa 13), and the Boyd site (EkSa 32). In addition to the Bear Lake site, the Boyd site was also found in quadrat survey in 1979, in Quadrat 27 (Gl7). The Shields site was recorded by C. Germann and D. Friesen during a Cariboo impact assessment survey for the Heritage Conservation Branch (Germann 1979) although we did not know that when we discovered it in 1979, near our field camp, but not in a quadrat. In addition to these three significant excavations, test excavations were also conducted in 1979 at EkSa 5 (Canoe Crossing site, lithic scatter), Quadrat 19, site 1, EkSa 27 (riverside lithic scatter), Chilko River site 92, EkSa 33 (riverside lithic scatter), Chilko River site 73, EkSa 35 (isolated small, single component housepit) and at ElRw 4 (Quiggly Holes site, lower lithic scatter area tested). These test excavations have been summarized in the previous section on regional survey and are described more fully in Matson et al. 1980. All three major site investigations used 1 m X 1 m units. Scale maps of site May 18, 2003-203 topography were drawn for each site, units were excavated using natural and cultural layers, often in combination, and all tools found in situ were recorded in three dimensions. All soil was screened through 1 / 8 inch (3 mm) hardware cloth, except some 25 two litre soil matrix samples which were gathered for flotation analysis. The samples were floated and sorted into light and heavy factions in 1983 and 1984 but were not further analyzed until 2000, as described in the appendix by Radomski. Individual forms were completed to describe soil layers, cultural features, wood and charcoal samples, and flotation samples. Samples of the forms are found in Appendix UL Bear Lake Site Excavations (EkSa 36) The Bear Lake site was identified as a likely Athapaskan site during survey. Characteristics previously discussed as typical of Athapaskan winter base camps in terms of features and locations were one or two shallow rectangular depressions indicating the former presence of rectangular winter lodges, and a location near, but not on the shore of a lake. Cache pits may also be present. The Bear Lake site had all those characteristics, with one obvious rectangular depression of the expected size, and another one recognized during excavation, two cachepits, one roasting pit, and a location about 800 metres from the shore of Eagle Lake. The most distinctive Athapaskan artifact identified in Chapter II was the stemmed Kavik point rather than the Kamloops side-notched point found in recent PPT assemblages and we did find a Kavik point (as well as a small blue glass bead) from the surface of the site in 1979. Our 1983 excavations occurred in five separate areas of this site (Figures IV -1 and -2): FIGURE May 18, 2003-204 148 ; I / (/ )( \"I I ,,{ I / CP\"! 0 LODGEPOLE PINE CP CM\u00C2\u00B7 PUSH \u00EF\u00BF\u00BD .t.SPEH '! ROCK \ Figure a=\". General layout of Bear Lake site excavations. ) \ i I U') g \ \u00EF\u00BF\u00BD I \ \ I / ( \ \ \ \ \ \FEATURE A \ l/ I \ \ V!3 \ Figure IV-1. General iayout of Bear Lake site (Ek.Sa 36) excavations. FIGURE Figure IV-2. Plan of Bear Lake site excavations in lithic scatter, Feature A, Feature \u00EF\u00BF\u00BD and Feature B areas. 1. Lithic scatter: Only one area of the site had much lithic material on the surface in1979. Twelve units were laid out in a trench in the lithic scatter area for two purposes: to provide preliminary training for the field school students before the more critical and sensitive features were excavated, and to examine the nature of the cultural deposits in the area. Prior to excavation, the entire lithic scatter area was again surface collected in 1983. A hearth (Feature E, Figure IV -3) was located during this excavation that did provide an important date. 2. Feature A: Four units were excavated in what appeared to be a small featurecontaining a large amount of fire-cracked rock (FCR) and a high density of basalt and obsidian flakes immediately north of the Lithic Scatter (Figure IV-4). 3. Feature D: Two units were excavated into this roasting pit feature (Figure IV-5, -6).Radiocarbon analysis of a charcoal sample from this rock-lined feature, which was found at the bottom of the pit, yielded a date of 295 + / - 75 BP (WSU 2902), the only radiocarbon date we obtained for the prehistoric component in the 1980s. 4. Feature H: Two units were excavated in this enigmatic feature. It consists of a faintMay 18, 2003-205 rectangular outline of rocks which was actually not noticed until some trampling around the perimeter of the main Feature B had taken place (Figure IV -7). Dendrochronological analysis of lodgepole pine beams in this shallow feature crossdate with those from Feature B, and Feature H also was likely built in AD 1877. We now believe that this is a second rectangular winter lodge (Figure IV-8) although we only became fully convinced of this after a field visit in 2000. 5. Feature B: This is the rectangular feature recognized at the this site in 1979 and 42units were excavated in and about this feature. We started with a 1 m wide trench through the north-south axis, just east of center, then enlarged the trench by 1 m to the west, extended a few units to the south, then expanded eastwards through the corner of the feature, and then westwards to determine the edge of this structure. The excavations in this feature proved to be most exciting, puzzling and rewarding. Dendrochronological analyses reported below demonstrate that this lodge structure was probably built in AD 1877, and radiocarbon sampies all derived ages of less than 150 BP as reported in the dating section below. As also discussed more fully below, this part of the site contains a buried prehistoric lithic component, including the remains of a prehistoric lodge, as well. Excavations in the lithic scatter and Feature A areas (Figures IV-3, -4) yielded a variety of stone tools and debitage, with Feature A containing a total of 105 kg (225 lbs.) of FCR in four units. Cultural deposits in these areas did not exceed 15 cm below surface. A probable hearth, Feature E, though, was present in Unit 10, within the lithic scatter (Figure IV-3). A charcoal sample was obtained that was dated to 415+ / - 115 BP (BGS 2263) in 2001. May 18, 2003-206 a. Excava ons Lake site. scatter area In Feature D, a total of 270 kg (595 ibs.) of FCR were removed from two 1 m X 1 m units (Figures IV-5, -6). No artifacts, faunal or floral remains were obtained from this feature, except a smail amount of charcoal found at the very bottom of the unit closest to center of the roasting pit which was dated to 295 + / - 75 BP (WSU 2902) as reported in the dating section below. A single boulder weighing approximately 40 kg (90 lbs) was found in the bottom center of the feature. Judging from the positioning of the rocks, the feature was used at least three times. A pavement of unbroken rocks, each about 20 cm diameter, separates the lower and middle parts of the feature, and an extended lip near the top 30 cm of the feature indicates the last major modification. The attributes discovered are characteristic of root-roasting pits (Pokotylo 1978; Pokotylo and Froese 1983; Peacock 1998) and, although the evidence of multiple use is only occasionally found, a roasting pit located near a long duration site would be expected to be used many times, and some of these uses would be expected to leave evidence of their occurrence. Excavations in Feature H (Figure IV-7) yielded very few materials, The two units FIGURE Figure IV-3. Excavations in the lithic scatter area of the Bear Lake site, facing north. FIGURE Figure IV-4. Excavations in Feature A at the Bear Lake site, facing south. May 18, 2003-207 n ew ava FIGURE Figure IV-5. Plan view of excavations at Feature D at Bear Lake, facing west. FIGURE Figure IV -6. Profile of excavations at Feature D from northwest. FIGURE Figure IV-7. Excavations in Feature H at Bear Lake. did produce a small bead, metal cut wood and some wire, some charred wood suitable for dendrochronology, charred pine boughs, and a few faunal remains. Only one kg of FCR were contained in the two units excavated there. Only one flake of any sort was recovered from these two units. The function of this feature was not well determined in the field, although we now think it was another rectangular lodge. While the dendrochronological dates indicate it was built at the same time as Feature B, the dimensions of the rectangular depression were smaller and shallower than Feature B and the structure was likely less substantial. Dimensions, as measured in 2000 (dotted lines in Figure IV -2), were 4.5m by 6m, compared to approximate 7m by 6m for Feature B. Like Feature B, it is oriented due magnetic north and contains both charred pinebough flooring and charred beams of approximately the same diameter (ca. 4 -8 cm). Some thick wire or a headless nail was found under two adjoining poles in Feature H, but no such association was observed in Feature B. If Feature H was not a lodge it may have served as a storage and curing facility for both fish and mammal meat and hides, May 18, 2003-208 Bear La excavat s a ure Bl remains f -rr-1 Figure\u00EF\u00BF\u00BD. Excavation of two post holes at the north end of the lodge. The uppermost contained an in situ post. 1:l:'-10 Figure 3:t. Profile of the western half of the Feature G hearth in the Bear Lake lodge. The \"boat\" shape of the hearth is clearly visible. and possibly for roots as well. Feature B proved to be the lodge structure that we hypothesized on the basis of the orignal survey. The initial trench (Figure IV -8) through the north-south axis cut lengthwise through the eastern half of a \"boat-shaped\" hearth (Feature G), and exposed an in situ support post (Figure IV -9), as well as showed evidence of floor and roof fill deposits. A post mold was located at the south end of the trench. The south end of this trench also yielded a considerable quantity of chipped stone material. Figure IV-10 shows the excavation at the stage of exposing the central hearth. Continued excavation on both sides of the initial trench provided evidence of an interior cachepit containing articulated salmon vertebrae (Figure IV-11) in the southeast corner of the lodge, and two more post molds were exposed, for a total of four main support posts, two at each end. One of the southern molds contained fragments of rotten wood. The west side of the lodge contained charred pine boughs on its floor, and a bench about 20 cm high was revealed (Figure IV-12). Throughout the FIGURE Figure IV-8. Bear Lake excavations at the stage of trenching through the Feature B lodge remains. FIGURE Figure IV-9. Two post holes at the north end of the lodge, facing east. The uppermost contained an in situ post. FIGURE Figure IV-10. Profile of the western half of the Feature G hearth in the Bear Lake lodge. May 18, 2003-209 \u00EF\u00BF\u00BD-\\ Figure l':. View of the Feature J cachepit found in the southeast corner of the Bear Lake lodge. -rr - \ '2-Figure l:!,. View of excavations along the west wall of the Bear Lake lodge, clearly showing the shallow bench along the wall. The \"boat\" shape of the hearth is clearly visible. FIGURE Figure IV-11. View of the Feature J cachepit found in the southeast corner of the Bear Lake lodge. FIGURE Figure IV-12. View of excavations along the west wail of the Bear Lake lodge, showing the shallow bench along the wall. excavations, abundant charred pole fragments were uncovered. These were preserved for dendrochronological analysis through the application of a gas/wax mixture intended to solidify the samples. The procedure simply involves dissolving petroleum base wax in gasoline and using a paint brush to soak the charcoal or wood sample thoroughly in situ. In sunny dry weather a sample may take as many as a dozen applications over the course of two days. When it is sufficiently impregnated with wax to permit removal, the sample is provided with wooden splints and wrapped in soft tissue, cheese cloth, cotton batting or the like. Faunal remains within Feature B were varied and abundant, especially in the Feature G, I, and J locations. These remains include abundant fur-bearer, deer and salmon remains, as discussed in detail in the Faunal Analysis section by Roberts and Magne. During our initial excavations of Feature B, we thought we were excavating a May 18, 2003-210 prehistoric lodge, given the lithics and lack of abundant historic items. However, as excavation continued in Feature B, it was noted that aithough FCR and lithics were abundant in the south end of our excavations, almost none of these items were found in good context in the north half of the feature. Further, while never abundant, we continued to add to our modest historic artifact assemblage and these were found in good context (Figure IV-17). We thus came to the conclusion that this lodge was historic. The dendrodates confirmed the exact date, somewhat later than we expected but not surprising given the isolated location of Eagle Lake. It is this historical material that serves as the basis for the historic Athapaskan occupation of the area, which we call the Lulua Phase, after the well-known family of that name which has lived around Eagle Lake. What about the lithics and FCR (Figure IV -17) that were abundant in the south end of the excavation? These would be expected to be the results of an earlier, prehistoric excavation. We noted the presence of hundreds of stone flakes near the south end of the lodge area, in strata likely below those of the later, historic occupation. This prehistoric occupation is marked by a hearth feature (Feature I), which is situated outside the boundary of the historic lodge on its south east side (Figure IV -15). It is also \"Boat Shaped 11, although smaller than the one (Feature G) within the historic lodge. At the time of excavation, no charcoal samples large enough for radiocarbon dating were recovered, although flotation samples were taken. In an attempt to test this idea of an earlier prehistoric component, we were able to confirm that level of occurrence of the flakes \"inside\" the inferred prehistoric lodge had an identical elevation to the surface of a new unit placed to the east which yielded a large stemmed biface within 5 cm (Plate May 18, 2003-211 '.@ I tr& ' , t \t .\u00EF\u00BF\u00BD ,. - .-le t ' '\\u00EF\u00BF\u00BDo \"' \u00E2\u0080\u00A2 ' \u00C2\u00B7 --\u00EF\u00BF\u00BD' \u00E2\u0080\u00A2 \u00C2\u00A9,Post in situ -' Charred Wood Q Hearth \u00C2\u00AE, Fire Cracked Rock @ Cache Pit o Rock ..o-- Salmon Bones \u00EF\u00BF\u00BD Charred Pine Boughs \u00C2\u00AE Post Mould Mag.N 0 1 metre \u00EF\u00BF\u00BD \u00C2\u00AE Feature \u00E2\u0080\u00A2 0dP\u00C2\u00AE 0 \u00EF\u00BF\u00BDI -\u00E2\u0080\u00A2IYo\u00EF\u00BF\u00BD v\u00EF\u00BF\u00BD /} (l'.l) ' 41;, ' Iii/I\u00E2\u0080\u00A2 0 JAnvil *Stone \ \ - ...\u00E2\u0080\u00A2 A!.-- I!/ - JC, - A-C l 0 20 8 oJ 40 D \u00EF\u00BF\u00BD EITJ EU3O + + + EU24 EU22 + .. A (IOYR4/3) 83 (1.!5YR!5/\") 8 (IOYR6/3) B3 soil change to dork (IOYR2/I) root/ tree surface \u00E2\u0080\u00A2\u00E2\u0080\u00A211\u00E2\u0080\u00A2totion ,I \u00EF\u00BF\u00BD I \.,I\" \u00EF\u00BF\u00BD-\ ' 165 EU28 EU26 + + EU21 + .. + .. D 84 mottled < D 81 (1.!5YR!5/4) \u00EF\u00BF\u00BD rock \u00E2\u0080\u00A2 02 (IOYR!l/3) Q 8!5 \u00E2\u0080\u00A2 charcoal \u00EF\u00BF\u00BD duff Figure :3:S. bissecting North-south profile of the Feature G hearth. EU2!5 EU24 8 1 L o+ featurt G hearth -20 10 EU23 EU2O i 0 + 20 + -40 0 60 ]_J __ ,\u00EF\u00BF\u00BD Scale in centimttru with 2 times \u00E2\u0080\u00A2trlicol exo99erotion magnetic Bear Lake lodge, 166 01 EU30 EU44 L EU!l4 EU!l9 EU55 J D A sandy silt (I0YR-4/-4) \u00EF\u00BF\u00BD root 0 60 Q B2 silly eloy (7.5 YR4/-4) @] lireeraek\"d rock 0 D B21/B3 silty sand (I0YR4/-4) \u00EF\u00BF\u00BD rock D Bl 1ond (IOYR3/2) m 1ur foe\u00E2\u0080\u00A2 ve9etatioft Scale In Cl!ntim\u00C2\u00ABtru \u00E2\u0080\u00A2Ith 2 tlmtt vertical B sand (I0YR5/-4) IS3 d11lf exogqerotlon []fill] hearth feature I, firecracked rock, 20 bone (I0YR6/8) ..,-U--,_/c-)-::L-- '-' Figure B.. East-west profile of Bear Lake lodge, showing stratigraphy of prehistoric layers and the Feature I hearth. 167 1 . I .. :\u00EF\u00BF\u00BD:,ti#.\u00EF\u00BF\u00BD:. D A (IOYR4/3 0 Ii sh bone layer Q B2 (IOYR5/3) \u00EF\u00BF\u00BD C (lOYR6/3) (SJ B3 (7.5YR5/4) [Zl bark D ,andy fill with charcoal !leeks 00 wood (IOYR5/4) @] roH tinged 90il patch \u00EF\u00BF\u00BD surfoct vttoetation a litter (7.5 YR5/4) [ID charcoal stained sand 0 20 cm (!OYR2/i) \ 6 North wall profile of the Feature J cachepit. 4d). It is this prehistoric component, which we identified as Athapaskan as explained later, that is the basis for the definition of the prehistoric Eagle Lake Phase. FIGURE Figure IV-13. Plan of Feature B excavations (Bear Lake lodge ). FIGURE Figure IV-14. North-south profile of Feature B, bisecting the Feature G hearth. FIGURE Figure IV-15. East-west profile of Feature B, showing stratigraphy of prehistoric layers and the Feature I hearth. FIGURE Figure IV-16. North wall profile of the Feature J cachepit. At the end of the excavations the Feature B area revealed a complicated history of at least two structures and at least two different occupations. Figure IV-13 shows the floor plan of the excavations in Feature B, and Figure IV-14 shows the north-south profile of the historic lodge along the west wall of our preliminary trench. The inferred boundaries of the Feature B historic lodge are shown in Figure IV-18. Figure IV-15 shows the profile of units cutting through the east-west axis of the southern portion of the feature, where the separate strata of the earlier occupation are evident. Layer Bf in Figure IV-15 is associated with Feature I which we infer is the deposit associated with May 18, 2003-212 168 EkSa 36, Feature B. a - distribution of lithic tool\u00E2\u0080\u00A2 and debitage (all r\u00C2\u00ADaaterials, b - di\u00E2\u0080\u00A2tribution of historic artifacts. b \u00EF\u00BF\u00BD -\ :r_ 1 ! Figure iii\". Horizontal distribution of lithic and historic materials in Feature B. the prehistoric lodge. Note that this layer is truncated by Layer B2f /B3 which is associated with the historic lodge. Figure IV-15 also indicates that the depression associated with the prehistoric lodge was probably very shallow with the western edge shown in the middle of Unit 30. As this is only two metres from the center of Feature I this indicates an internal width of the prehistoric lodge of about four metres. Since the surface of the ground was disturbed by the historic lodge and the north-west portion of the prehistoric structure destroyed by the historic lodge, neither the depth of the prehistoric depression, nor its areal extent are certain. The extent to the east is not clear in the cross-section, nor was it clear in the field. FIGURE Figure IV-17. Distribution of lithic and historic materials in Feature B. In the distributions of historic material and of lithic materials shown in Figure IV-17, it can be seen that the lithic materials are clearly associated with the southeasternportion of these excavations, while the historic materials are clearly associated with the main feature itself, and are centered around the central hearth, Feature G. The north half of Feature I is nearly contained in Unit 54, and it must end somewhere around the middle of the unit to the north. It is likely that Feature I extends a full metre south of Unit 54, thus having a total length of 2.5 m. Note that few lithic materials are found in Unit 58 (four lithics in total, two flakes and two tool fragments, but one of those tool fragments is from the large stemmed biface and was located right on the centre of the west side of this unit ), likely indicating that this unit is outside the wall of the May 18, 2003-213 prehistoric lodge. Layer Bf was clearly present across Unit 55, but not clearly present in Unit 58, so the eastern edge of Unit 55 is an approximation of the east edge of the prehistoric lodge. Using the centre line of Feature I as the centre of the prehistoric lodge, this indicates that the prehistoric lodge is about 5 metres wide, slightly wider than suggested in the previous paragraph. The edge of the distribution of lithics to the west is less definite, but this smearing is understandable as resulting from the historic lodge excavation through parts of the prehistoric Eagle Lake Phase lodge, spreading the lithic material around. Note that in Figure IV -13, that the FCR is also concentrated around Feature I, and does not really extend to Unit 39. In Unit 58, the concentration of FCR in the northeast corner is a separate feature, clearly disconnected from the lodge remains centred around Feature I. The length of the prehistoric lodge is also unclear, but we have evidence of it for about five metres. As the historic Lula Phase lodge has the rafter support posts about 1 metre beyond the boat shaped hearth, we can use this figure to estimate the length of the prehistoric lodge. This gives us a length of about 4.5 metres between rafter support posts. If we allow slightly less than 1 m between rafter support pole in end of the lodge at the floor level, in accord with Feature B and the photograph of the Puntzi lake lodge, this results in a length of about 6 metres. We thus end up with a lodge size of between 4 and 5 metres in width and a length of 6 metres (Figure IV-18), very close to the ethnographic description of 15 by 20 feet -- and to the surface dimensions of Feature H (Figure IV -2). The orientation of the hearth, Feature I, is slightly east of magnetic east, probably slightly east of the orientation of the two historic rectangular features. FIGURE May 18, 2003-214 Figure IV-18. Boundaries of Feature B and Feature I Lodges-Figure IV-16 shows the stratigraphy of the Feature J cachepit. We are not certain whether this was a cache pit inside the historic lodge, or whether it post-dates the historic lodge. Thus there may be three separate occupations present in this area of the site. Thus our interpretation of events is that a lodge structure was built in \"prehistoric\" Eagle Lake Phase times at the site, including \"boat shaped\" hearth Feature I at the centre. Later (circa AD 1877), the Feature B structure was built, destroying part of the original occupation and scattering the remains about the south end of the historic lodge. In this interpretation the lithic scatter areas are associated with the earlier, prehistoric occupation. If we presume that roasting pits were not used in historic times, Feature D is part of the prehistoric occupation and the date associated with Feature D dates the prehistoric occupation. Similarly, the lithic scatter should also have been produced at the same time as hearth Feature I and the roasting pit. The inferred sequence of occupation at the Bear Lake site as of 1984 was as follows. There are at least two major components here. The first occupation, the Eagle Lake Phase, occurred in late prehistoric times by the time of at least 295 + / - 75 BP (WSU 2902), as indicated by the carbon sample from the roasting pit, Feature D. We would expect that the prehistoric lithic component at Feature B, the lithic scatter, the FCR Feature A area, and the roasting pit were all the products of this first occupation. Then in AD 1877, and only possibly slightly earlier, the historic lodge structure and the May 18, 2003-215 almost adjoining Feature H \"probable lodge\" were constructed, during the Lula Phase. Feature J, the cachepit, may have been constructed yet later, or at this 1877 construction date. We cannot ascertain fully the function of the lithic scatter or FCR Feature A. The FCR Feature A does not appear to have been formed in situ, but rather is perhaps the result of some disposal or \"house-cleaning\" activity. The soil about the feature is not discoloured through burning, it contains no charcoal, and only very few fragmentary faunal remains. The presence of obsidian and basalt materials amongst the FCR lead us to conclude that it was formed during the prehistoric occupation. The lithic scatter area yielded the probable hearth feature (E) at the south end of the area excavated in Units 9-12, but this was not well defined appearing only as a slight concentration of charcoal and fairly abundant FCR quite near the surface. It was likely an outside work area located adjacent to the inferred prehistoric lodge. It was located down wind of the prehistoric iodge according to the prevailing winds observed during the summer of 1983. Although we planned to carry out further excavations to more fully define the prehistoric lodge, we were unable to do so. We were able to obtain additional funding to produce the current volume and redate this prehistoric occupation. If our interpretation is correct, the charcoal sample of Feature E in the Lithic Scatter, should date approximately the same as the date from the Roasting Pit, Feature D. We had been somewhat dubious about the Feature E sample, as the deposits, as indicated above, were very shallow. A sample of 1.85 g of charcoal from Feature E in Unit 9 was dated in 2001 at 415+ / - 115 BP (BGS 2263). This date is approximately the same as the earlier May 18, 2003-216 one from the roasting pit (295 + / - 75 BP; WSU 2902), i.e., the two dates overlap at one sigma, supporting our inference that both should date the prehistoric Eagle Lake Phase occupation. Although we did not have any charcoal samples from Feature I, the inferred \"prehistoric\" boat-shaped hearth, we did have a flotation sample. This turned out to have lots of very small basalt flakes present, confirming its \"Prehistoric11 nature--at least much earlier than the 1877 occupation of Features B and H. A sample of 0.03 g from this flotation sample (No. 18) was sent off to BetaAnalytic for AMS dating, and resulted in a date of 210 + /- 40 BP (Beta 148106), confirming the prehistoric nature of this deposit, and in rough agreement with the other two dates, overlapping at one sigma with the date from Feature D, the roasting pit. If we assume that all three radiocarbon dates are of the same event, as we argued in 1984, we can average these three dates, using the procedure described by Berry (1980). If we do this, we obtain an averaged date of 245 + /- 34 B.P. This in turn converts to a calibrated date of AD 1645-1660 with no intercepts at one sigma other than this narrow period (Figure V-4). We discuss this averaged, calibrated date for the Eagle Lake Phase further below, in the section on dating. A few comments on the historic lodges are in order. Although the ethnographic accounts indicate the use of spruce boughs for bedding, the recovered material appears to be lodgepole pine boughs. Since spruce is present about 1 km away, this suggests that the preference for spruce was not great. As we have paired rafter support posts at both end of the lodges, this indicates that rather than like the structure seen at Puntzi Lake with a single ridge pole (Figure 11-2), this lodge was of the \"double lean to\" design May 18, 2003-217 with two ridge poles. This lodge style is also found among B.C. Athapaskans as Duff (1981) reports for the Tsetsaut, and suggested as an alternative form by Lane (1981) for the Chikotin. It may be, though, that the structure was rebuilt, one set of the posts was not used, etc. Rectangular lodges with two ridge poles are fairly widely distributed among northern Athapaskans. In the extreme northwest edge of the Athapaskan distribution, the Koyukuk in contact times used a double ridge pole structure ( Clark 1996; Clark and Clark 1974). These houses, though, were excavated well into the ground and included entrance tunnels (Clark 1996: Figure 4.3), so the structural similarities are limited. The size, about 5m x 5m, though is not far off that reported by Lane (1953) or Feature H at the Bear Lake site. Clark (1982) also reports on the remains of rectangular structures around Great Bear Lake, at the northeastern limits of Athapaskan occupation. These remains, most dating in the last 100 years, include structures seemingly identical with the rectangular lodge described by Lane, plus variants, but without significant sub\u00C2\u00ADsurface excavation. The Great Bear Lake lodges tend to be slightly wider than long and may include both single and double ridge pole variants, as well as forms with low vertical side walls (Clark 1982:117). Perhaps the most similar reported structure to Feature B is found in southwestern Yukon, in Charnpagne-Aishihik territory among contact structures (Greer 1985, 1986). The Sandpiper Creek House GcVa 10) in particular (Greer 1985, 1986) is a single ridgepole rectangular lodge similar to that described by Lane and Feature B with a clear entranceway at one end. The largest difference is in size, as this structure is only about 3.5 by 4.5 m. Greer (1985, 1986) reports on a number of other small rectangular May 18, 2003-218 structures made of ax-cut logs, but none are as clearly similar as the Sandpiper Creek House, and like it, are relatively recent in time. In sum rectangular lodges sharing features in common with those described by Lane (1953,1981) are common among northern Athapaskans The orientation of Feature Bis almost perfectly magnetic north, as shown in Figures IV-14 and 19. The probable historic lodge in Feature H has the same orientation--or at least the rectangular depression has this orientation. The length of the historic lodge (Feature B) appears to be about 6 metres with the rafter support posts 4 metres apart. The width is not as certain, but appears to be about 7 metres as shown in Figure IV-18. Thus the structure is siightly wider than long, and has a floor area of about 40 square metres. According to Lane (1953:145) ethnographic size estimates vary, but centred around 20 ft long and 15 ft wide. Note that this lodge has about that length but is clearly much wider. This may be an error in perception, as the \"wall\" and the ''roof' are formed by the same slanting member so that the edges close to the sides of the structure would have very low ceilings. Thus a structure of this shape may appear to be longer than wide, as one can walk from end to end, but the sides could not be approached in the same manner, and Clark (1982) reports on measurements of similar structures that tend to be wider than long and overlap in size with Feature B and H. On the other hand, if Feature B was indeed, a lodge with two parallel ridgepoles, one would expect it to be wider than a single ridgepole lodge, thus explaining the wider-than-long shape. Feature H does have the dimensions reported by Lane, although we do not have information on the structural details. Finally, Lane (1953:46) has indicated the thermal inefficiency of this structure, May 18, 2003-219 relative to the pithouse, and indicates that the fire had to be kept burning constantly. Thus one burned up a lot of wood, and this limited reuse of any area. Prior to metal implements--and even afterward---the easy way to keep a fire going in such a structure would be to drag in a small downed tree, and keep one end burning. Osgood (1932) also reports this activity in rectangular winter Athapaskan lodges, though, of a slightly different structure. Keeping a fire going was definitely necessary in such a structure during the winter. This sort of fire would result in an hearth area not round, but linear, like the \"boat-shaped\" hearths that we located in both the lodges associated with Feature B and Feature I. This ought to be another, heretofore unrecognized, useful archaeological distinction between Athapaskan and PPT occupations. It is interesting to note that Campbell (1968) in his report on the Kavik type site apparently located a small boat shaped hearth, perhaps the center of a similar, but smaller, structure. May 18, 2003-220 Excavations at Plateau Pithouse Tradition Sites. The Bear Lake Site excavations were undertaken to confirm the inferred Athapaskan present there, as well as to obtain dates and a good prehistoric Chilcoiin assemblage. Identification of an ethnic group implies a \"compared to what\" and that is the PPT in our case. We needed to excavate inferred PPT sites to confirm that there was a PPI occupation in Eagle Lake and to discover what this occupation dated to. The latest PPT combined with the earliest Chilcotin date should give us the minimum and maximum dates for the Chilcotin arrival in this area. For these reasons we excavated the Boyd and Shields sites. Both of these sites contain five or more circular housepits immediately adjacent to each other and to the Eagle Lake eastern shore. Both are immediately behind the presumed ice-raised ridge found along that shore. Our excavations began at the Boyd site (EkSa 32), primarily as that was found during the quadrat survey and it did not have the amount of tree growth found on the Shields site. Boyd Site Excavations (EkSa 32, Q 27-1) Excavations were begun on July 25,1983 at the Boyd Site (EkSa 32). As shown in Figure IV-19, five housepits are present in the shelter of the shore ridge, three about six metres in diameter and two closer to five metres. Only four units (plus a smaller text) were excavated at the Boyd site (Figure IV-19) because we stopped excavation after we found it had produced four large stemmed or corner notched projectile points, and three microblades. We considered the microblades to be likely indicators of a pre-PPT occupation, likely to be mixed, as so often happens in housepit sites, with later May 18, 2003-221 179 E A S T !A A A. A A A A A A_ ;.!:.-.,-!. --;,--::\u00C2\u00B7 .. \" \" ,,--. \" ,.W A L L L S O U T H WA L L -----;;--...-\u00C2\u00BD--'-;-_..,\u00EF\u00BF\u00BD.!_\u00EF\u00BF\u00BD!._,....:-,.;..-,\u00EF\u00BF\u00BD. A \u00EF\u00BF\u00BD \u00EF\u00BF\u00BD - -\u00EF\u00BF\u00BD - '\u00EF\u00BF\u00BD - :f \u00EF\u00BF\u00BD -\u00EF\u00BF\u00BD A-;. \"-.,--' -,\u00EF\u00BF\u00BD - '\u00EF\u00BF\u00BD - A - -\"'- _._ - :----!_.-,;:-----\" \" ,. ,, \" D A (I0YR2/I) 00 firecrocked rock [Sl 82 (I0YR5/4) lS3 root mattrial Ll 83 (I0YR<\/2) CJ 1,1ny 1oil (I0YR5/l) D 83b (I0YR3/l) @ aurtoce ve1,1e1alion Ll C (I0YR5/2) 0 boaoll flake root 20 cm \u00EF\u00BF\u00BD---v\u00C2\u00B7O Figure\u00EF\u00BF\u00BD Profile of east and west walls, Unit site. of Boyd occupations. Further, we thought the stemmed and corner notched points to be typical traits of pre-Kam.loops PPT, and thus, not likely give us an indication of the last PPT occupation in the area. As we had found numerous Kamloops side-notched points during our survey, we expected that there were housepit sites present occupied during the Kamloops horizon. In fact, this line of argument is incorrect, at least in respect to corner-notched and stemmed points, as we found they are abundant in Kamloops horizon sites, as discussed in the following artifact description chapter. FIGURE Figure IV-19. Contour map of the Boyd site (EkSa 32, Q27-1). FIGURE Figure IV-20. Profile of east and south walls, Unit 1 of Boyd site, Housepit L Housepits 1 and 2 were chosen for excavation after a king tube was used to probe the features for cultural layers and more obvious layers were present in these two structures. Strata here were difficult to define but appeared to include a floor (B3), fill (Bl), and roof fall (B2) layers (Figure IV-20). The near-surface layers of Housepit 2 contained a fire-cracked rock feature (A) that was made and used well after the main occupation of the housepit. This feature was concentrated in Unit 3 and had Layer B2 beneath it. Excavation continued to the floors of the housepits in all units. The floor was better defined in Housepit 1 than Housepit 2, but in both cases the contact with sterile May 18, 2003-222 gravei was sharp, and the floors were flat with only slight rises seen in the north ends of Units 2 and 4. A good hearth was partially exposed in the southwest corner of Unit 3, that included numerous faunal remains, similar to the hearths at the Bear Lake site. Except for this hearth being off centre, the attributes of these structures are all consistent with what is expected in PPT winter structures. Curiously enough, Test Unit A (Figure IV-19) produced a microblade and a higher concentration of lithics than found in the housepits. Is this an earlier microblade component that the pit structures were excavated through? A radiocarbon date of 505 + / - 70 BP (WSU 2905) , with a most probable calibrated date of AD 1410, was obtained from Unit 2 Layer B2 (46.4-49cm below surface of the southeast corner) House 1 at the Boyd site. This sample was from a rather large burned, but badly decayed piece of wood, which is interpreted as part of burned roof fall that ended up close to the floor. It should date the last construction of the house, as there was no obvious floor higher in the stratigraphy. This date appeared to be quite young to us in 1984 in comparison to the assemblage recovered, but further research confirmed that a Kamloops assemblage is present. This date also turned out to be the most recent PPT date obtained in the Eagle Lake area. Shields Site Excavations (EkSa 13) Excavations at the Shields site were initiated on Aug. 3, 1983 after the Boyd site failed to produce items we considered diagnostic of the Kamloops Horizon. This site is May 18, 2003-223 \u00EF\u00BF\u00BD I \ 173 EU 9 EU 10 r I I I I I I I I I I I I I I I I I I D A dark 1illhond \u00EF\u00BF\u00BD 83b compact ,ondy sill (IOYR6/3) Ll 82 1ond/9ro\u00E2\u0080\u00A2el (IOYR3/3) fSSJ 83b gro,tl lent (!OYR3/I) D 82b 1ond/9ronl (IOYR2/I) D C sand/grovel (IOYR3/3) Ll 82c 1ond/9ro1tl ( iOYR4/3) \u00EF\u00BF\u00BD root D 83 grovel (IOYR3/1) \u00EF\u00BF\u00BD firecrocked rock 830 10ndy till (IOYR 3/1) II) aurfoc\u00E2\u0080\u00A2 vegttollon \u00C2\u00A7 _--'2- .--l,Figure\u00EF\u00BF\u00BD. west wall profile of units 9 andat the Shields site.20cm 1 0 in House located on the protected slope of a high point on the beach ridge. This was likely nearly an island during times of high water, as a bay extends behind the site. We mapped six housepits, ranging in size from 8 metres (HP 1) to about 5 metres (Figure IV-21 ). Three of the six housepits at this site were excavated by a total of 10 units, with only two units in Housepit 2. Strata bearing cultural materials extended to about 1 n1eter below surface and a profile o f two units in Housepit 1 is shown in Figure IV -22. The stratigraphy at the Shields site was more complicated than at the Boyd site with more evidence of reuse, as shown in the stratigraphy profiles and photographs. Strata were difficult to define but layers indicating a floor (B3b), post-occupation fill (B3a), and roof fall materials (B2-2c) were defined. The Shields sited id produce a multi\u00C2\u00ADside notched Kamloops point, so that a Kamloops Horizon occupation was assured, although only one of the three radiocarbon dates produced supports this occupation. The Shields site contained an abundance of stone tools and debitage, some highly decorated antler artifacts, many fauna! remains, but very little charcoal. FIGURE Figure IV-21 . Contour map of the Shields site (EkSa 13). FIGURE Figure IV-22. West wall profile of units 9 and 10 in Housepit 1 at the Shields site. FIGURE Figure IV-23. Excavations of Housepit 5, Shields site, facing east. May 18, 2003-224 We FIGURE Figure IV-24. West wall profile of Unit 3, Housepit 2, Shields site. FIGURE Figure IV-25. West wall profile of Unit 6, Housepit 2, Shields site. Housepit 2 contained deposits less deep than those of the other two houses ( 40 cm vs. 90 cm), and also exhibited a concentration of FCR and black soil staining near its center. Housepit 2 also does not have the well-defined rims of Houses 2 and 5, and overall we think that Housepit 2 may have been a facility of some sort other than a habitation structure, although it does contain a sizeable stone tool assemblage. Although the features encountered during excavation were not as clear as at the Boyd site, with the exception of Housepit 2, all the attributes are consistent with the Plateau Pithouse Tradition. A charcoal sample from Housepit 1 (Unit 10) yielded a radiocarbon date of< 150 BP (WSU 2906), pre-bomb, or between AD 1800 and AD 1950 even though it was 68-72 cm below the unit datum (this is discussed further in the section on dating). This is not a feasible age for this occupation. Another sample from Housepit 2, Unit 5, layer Bla, produced a date of 750 + /- 90 B.P. (WSU 3379). A final sample from Housepit 1, Unit 2, 41-45cm below surface, Layer B3, produced a date of 1650 + /- 85 B.P. (BGS 2264).Although this date appears to be relatively old for the depth below surface, it is in the rim and inferred to be equivalent to a floor layer. May 18, 2003-225 ChapterV ARTIFACTS, DATING, AND FAUNAL REMAINS Artifact Descriptions Patricia Ormerod This section of the report provides descriptions, definitions and photographs of the stone, bone and antler, and historic artifact classes recovered in the excavations and collections. Artifact types and acronyms follow the classification used in the original Eagle Lake Archaeological Project report (Matson et al. 1980). The acronyms are used in tables in this report and in the artifact database. This classification serves as the basis of the intersite comparisons basic to the research reported here. All objects were re-examined, reclassified, and measured for this report. Length, width, thickness and weight data are reported for all artifact types. When metric data are tabled, the mean and range are reported and the inter-quartile range (IQR) is also reported for types with more than 10 members. Tabled data for maximum length and maximum width are based only on specimens that are complete in that dimension unless stated otherwise. Thickness data are based on all specimens including fragments. Weights are based on specimens that are complete in both length and width unless stated otherwise. All dimensions are given in millimetres and weights are in grams. Lithic raw material types are described as granite, chert, obsidian, and fine-grained or coarse-grained basalt. Bakewell and Irving ( 1994) have reported that petrographic definitions without geochemical analyses can result in confusion in the classification and sourcing of stone; we suspect they would classify the Eagle Lake area \"basalts\" as \"dacites\". May 18, 2003-227 Artifact numbers used in this report are those on the artifacts and, therefore, may include the temporary site numbers assigned in the field rather than Borden site numbers. Table V-1 provides a cross reference of temporary site numbers, Borden site numbers, and site names. Although many artifacts have been previously illustrated, both in publications (various Matson and Magne, Magne 1985; Matson and Pokotylo 1998; Pokotylo and Mitchell 1998) and in 'Grey Literature' research reports (Matson et al. 1980; Magne and Matson 1984), all artifact illustrations were re-photographed for the current volume. Where appropriate, references are made to Paull' s ( 1984) residue analysis, which is reported in more detail in Magne and Matson (1984 ). Paull used the Sudan IV test for fats and the Hemastix test for hemoglobin. Only loci that looked like they might be dried blood visually were tested. We found that the most detailed and useful artifact classification for comparisons was Stryd's 1973 dissertation, which included a detailed analysis of several large assemblages dating to the last half of the Plateau Pithouse Tradition in the Lillooet region of B.C. May 18, 2003-228 Table V-1 Field Assigned Site Numbers, Borden Site Numbers and Site Names (For lithic artifacts described in \"Artifact Description\" and labelled on Plates and Tables.) \u00EF\u00BF\u00BD (n.a.) EkSa 13 Shields Site ELP 1 EkSa 17 ELP 16 EkSb 5 ELP 19:1 EkSa 27 ELP20 EkSa 28 ELP22 EkSb 6 ELP26 EkSa 31 ELP 27:1 EkSa 32 Boyd Site ELP32 EkSa 36 Bear Lake Site ELP 44:2, G20:2 EkSb 13 ELP 44:3, 020:3 EkSb 14 ELP 44:4, 020:4 EkSb 15 ELPO2:4 EkSa 130 ELP 05:1 EkSb 21 ELPO7:1 EkSb 24 CR 1 EjSa 11 CR3 EjSa 10 CR9 EkSa 115 CR 12 EjSa 13 CR28 EkSa 98 CR32 EkSa 39 CR40 EkSa 89 CR50 EkSa 118 CR64 EkSa 34 CR 73 EkSa 35 CR89 EkSa 55 CR92 EkSa 33 Brittany Creek Site CR98 EkSa 62 T84-27 EkSb 37 Fish Trap Lake Site E1Rw4 E1Rw4 Quiggly Holes/ Bid well Creek Site ELPG2:04 Off site May 18, 2003-229 Flaked Lithic Assemblage Bifaces 1. Side-notched Points and Multi-notched Side-notched PointsSide-Notched Points SNPT n = 55 The side notched point assemblage includes 49 fine-grained basalt and 6 obsidian specimens. Table V-2 Side-Notched Points: Metric Data # of Specimens Range IQR Mean Length 13 16.7-32.5 18.8-27.4 22.5 Width 27 10.2-20.2 11.8-14.8 13.5 Thickness 55 1.6- 6.6 2.7-3.6 3.2 Weight 11 0.43-2.39 0.55-1.25 I.IOThe simple or single side-notched points and fragments (Plate 1, a-a') recovered in the Eagle Lake project area are equivalent to Stryd's Side-Notched Arrow Points (Stryd 1973:330-332, Plate 24 a-t) and Sanger's Group 10 Side Notched Points with small, narrow notches (Sanger 1970:42-44) which are commonly referred to as Kamloops side-notched points (Richards and Rousseau 1987; Pokotylo and Mitchell 1998). Blades are triangular with straight (rarely convex) lateral blade margins, side notches are (usually) bifacial, narrow, and shallow, and bases are generally straight, although some are concave. Kamloops side-notched points were strongly associated with the Kamloops \"Phase\" components (Stryd 1973:332) of the Kamloops Horizon (1200 -200 BP) as defined by Richards and Rousseau (1987). They first occur about 1200 years ago and are used to define the beginning of the Kamloops culture. Small, side-notched arrow points are found over much of western North America in the last lO00years. The dimensions and weights of Eagle Lake area side-notched points are similar to those May 18, 2003-231 a b C d e f g h k m 0 p s t u V w X y z a' b' c' d' e' f' g' o 1 2 3 4 5cm \u00EF\u00BF\u00BD\u00C2\u00B7 Side-notched and Multi-notched Points Side-notched Points (SNP1): a. ELP32:1122 b. ELP 19-1:454 c. CR 98:8T84-27:491 e. ELP 19-1:163 f. CR 73:40g. CR9:1 h. ELP 19-1:641 i. CR98:7j. ELP 19-1:24 k. EkSa 33:3895 I. CR98:lEIRw4:I677 n. CR 92:1892 o. ELP 19-1:8p. EkSa 13:4092 q. ELP 19-1:1461 r. ELP 19-1:244s. CR 1:2 t. ELP 19-1:175 u. EkSa 33:3047v. T84-27:21 l9 w. EkSa 36:4981/2215 X. ElRw4:1357y. EkSa 36:3072 z. CR 92:1868 a'. CR 32:1 Multi-notched Points (MNPT):b'. EkSa 13:1058 c'. E1Rw4:1356 d'. ELP 19-1:1493 e'. CR 28:2 f. CR 92:2194 g'. ElRw4:1675 ]\"\" '+ recorded by Stryd (1973) and Matson et al. (1984). Stryd noted that up to 30 % of the Kam.loops side-notched points, primarily the smaller specimens, were created by bifaciai marginal retouch o n a unifacial primary flake (Stryd 1973:331). This reduction process was apparent in only 3 ofthe Eagle Lake Project side-notched point assemblage (Plate 1, s-u). One point (Plate 1, v), exhibits the bilateral basal flaring and rapidly expanding stem described by Stryd (Stryd 1973:331, Fig 24 q-t) but has no evidence of \"the beginning of small side notches\" that he found on three of the four Lillooet area specimens.EkSa 36:3072 (Plate 1,y) was tested by Paull (1984) for fat and blood residues at two loci and both came up positive. This finding is in agreement with the traditional interpretation that these bifaces are arrow points. Two specimens have single side notches only (Plate 1, w-x): the one from the Bear Lake site (EkSa 36) (Plate 1, w) is thin with unifacial flaking along both lateral margins and one deep side-notch. It was recovered in two pieces (blade and base) which had snapped apart from the inside centre of the notch, so we can assume it broke during manufacture. The other single side\u00C2\u00ADnotched point (Plate 1, x) is flaked on alternate faces and was intact when found. A detailed, multivariate analysis of the variation within this point class is provided in the chapter on ethnic identification where, subtle variations within this class are shown to correspond with ethnic origin. Table V-3 Single-side Notched Points: Metric Data Artifact No. Length Complete Width Complete EkSa 36:4981/2215 31.2 Y 20.2 Y ElRw 4:1357 23.0 Y 12.2 Y Thickness 3.2 3.8 Weight 1.45 1.00 Large side notched points (Plate 1, y-a') include one, of obsidian, from the Bear Lake site (EkSa 36) that has a convex base with a spuL A specimen from the Brittany Creek site (EkSa May 18, 2003-232 33) (Plate 1, z) has an elongated, asymmetrical tip and crushing on one lateral margin: it mayhave been used as a drill. Table V-4 Artifact No. ELP 32:3072 CR 32:1 CR 92:1868 Large Side-notched Points: Metric Data Length Complete Width Complete 32.5 Y 15.7 Y 35.3 N 17.0 Y 27.4 Y 15.3 Y Multi-notched Points MNPT Thickness 5.4 6.6 6.2 n=6Weight 2.10 3.41 2.39 Five multi-notched side-notched points from the Eagle Lake Project area are fine-grained basalt, one, from the Shields site (EkSa 13), is obsidian (Plate 1, b'); and all are bifacially flaked (Plate 1, b'-g'). Stryd (1973:331-32) placed the multi-notched point within the Kamloops culture but Richards and Rousseau (1987:43-44) believe the multi-notched variant of the Kamloops side-notched point type dates between ca. 400 and 100 B.P. This kind of point is associated with PPT and not with Athapaskan assemblages. Five specimens exhibit multiple notches on the blade and one has muitiple notches on the stem only (Plate 1, g'). In all cases, the notches are asymmetrical, occurring only (or in larger numbers) on one lateral margin. The point with stem notches (Plate 1, g') and one specimen with blade notching (Plate 1, f') exhibit \"small, nan-ow notches\" described by Stryd (1973:331, Fig 24, i) for the single specimen in the Lillooet assemblage; the other four specimens have broad, shallow notches. The two specimens from the Bidwell Creek/Quiggly Holes site (ElR w 4) were recovered during excavation, not on the surface. The sizes reported here are similar tothose given by Matson et al. (1983:163) for MOC. May 18, 2003-233 Multi-notched Points: Metric Data Table V-5 Artifact No. EkSa 13:1058 EkSa 33:2194 ElRw 4:1356 ElRw 4:1675 ELP 19-1:1499 CR 28:2 2. Length Complete Width Complete Thickness Weight Kavik Points 29.1 Y 15.4 Y 4.3 1.44 16.4 N 12.6 N 3.1 0.68 21.3 Y 13.5 Y 3.1 0.69 24.4 N i3.3 N 3.7 1.01 17.9 Y 14.0 N 3.1 0.67 16.2 N 11.9 Y 2.8 0.58 KAPT n=4 The Kavik point (Campbell 1968), also referred to as Klo-kut (Shinkwin 1979: 117, 154; Morlan 1973:480, Campbell 1968:41), is a small, stemmed point with triangular blade, maximum width at the shoulders, and a well-defined stem that can be straight or contracting. The specimens from the Kavik site, described as crudely finished che1t and chalcedony, had pointed bases and sharp, unground edges (Campbell 1968:37). The Kavik point style is associated with northern Athapaskan culture (Campbell 1968:39-41) and is a relatively recent style (Campbell 1968:40; Boudreau 1974:11). The size range of the four Kavik points recovered in the Eagle Lake Project area is most similar to the assemblage of five \"Klo-kut-like points\" from the Dixthada site (Shinkwin 1979:117) although in form both assemblages are like the slightly larger points from the Klo-kut site (Morlan 1973:241-249). The Eagle Lake points are fine-grained basalt (Plate 2, a-d). Only items \"a\" (ELP 32-1:188; the Bear Lake site) and ''d\" (EkSa 33: 1007) on Plate 2 fully fit the description of Kavik or Klo-kut points. The other two points have both notches and tapering stems that might be considered to be combinations of Kamloops and Kavik point attributes. Nothing like them was found in PPT descriptive material, indicating that they are not typical of the Plateau Pithouse Tradition, and thus they are tentatively included along with more typical Kavik points in this report. May 18, 2003-234 a b C d e f k a n 0 p O 1 2 3 4 5 cm h g m Plate 2 Kavik, Comer-notched, Stemmed and Miscellaneous Points Kavik Points (KAPT): a. ELP 32-1:188 b. Ek:Sa 33:2650 c. CR 73:39d. CR 92:1007Comer-Notched Points (CNPT):e. EkSa 32:6200 f. EkSa 13:1474 g. EkSa 32:5712h. CR 89:1Stemmed Points (STPT):i. EkSa 32:1046 j. ElRw 4:1358 k. EkSa 13:2418L CR 40: 128 m. ELP 27 OS:8Miscellaneous Points (MIPT):n. F.kSa 33:2477 o. EkSa 33:3330 p. CR 92:208Table V-6 Kavik Points: Metric Data Artifact No. Length C Width C Thickness Weight CR 73:39 16.5 N 11.0 N 2.5 0.50 ELP 32-1:188 22.8 y 9.1 y 3.0 0.47 EkSa 33:1007 11.5 N 12.2 N 2.4 0.28 EkSa 33:2650 16.6 N 11.0 N 3.1 0.68 3. Comer-notched Points CNPT n=4Two complete comer-notched points and two base fragments, all fine-grained basalt,were iecovered in the Eagle Lake Project area (Plate 2, e-h). The largest complete point (Plate 2, e) is from the Boyd site (EkSa 32). The fragment of a corner-notched point (Plate 2, f) recovered from the Shields site (EkSa 13) is almost identical in size and shape. The small, complete corner-notched point (Plate 2, g), also from the excavation at the Boyd site (EkSa 32), is very fine-grained basalt with unifacial thinning of the base. These three points appear to fit Type A Corner-Notched Atlatl Point<; as described by Stryd for the Lillooet area of the British Columbia Plateau (Stryd 1973:338-39, Fig 26 a-h). Stryd reports (1973:339) that 12 of the 23 Type A points he recovered were from Kamloops \"Phase\" components and most of the rest were from the equivalent of Richards and Rousseau's (1987) Plateau Horizon (2400 -1200 BP). This distribution indicates they are probably not atlatl points and are, in any event, associated with the last half of the Plateau Pithouse Tradition. The presence of these two corner-notched points (and the stemmed points described below) and the lack of Kamloops points at the Boyd site were interpreted in the field as indicating that the Boyd site did not have a Kamloops component present. Since one of the main reasons for excavating the Boyd site was to obtain a terminal date for the Eagle Lake PPT occupation, this inference ended excavation there and led to the excavation of the Shields site. The identification of this point type as common in the Kam loops culture invalidates this May 18, 2003-235 interpretation, as did the resulting Kamloops-age radiocarbon date obtained from the Boyd site. Further, the third example of this point type was from the Shields site, another late PPT component. Paull (1984) tested EkSa 32:6200 (Plate 2, e) in three loci for blood and one for fat One test was positive for fat and negative for blood and two were positive for blood, as expected for projectile points. The fourth comer-notched point fragment from the Eagle Lake area was collected on the surface of EkSa 55 (Plate 2, h). It is similar to Sanger's Group 3 Projectile Point'> ( 1970:38, 105, Fig 20 k-p), attributed to the Lower Middle Period (5000 BP to 4000 BP), however, it has no evidence of basal grinding. Table V-7 Artifact No. EkSa 32:6200 EkSa 32:5712 EkSa 13:1474 CR 89:1 Corner-notche9 Points: Metric Data Length C Width C Thickness 54.1 Y 29.5 Y 8.2 32.6 Y 22.8 Y 5.3 35.6 N 22.4 N 8.6 28.6 N 24.9 Y 4.7 4. Stemmed Points STPT Weight 12.69 3.39 5.74 2.57 n=5The five stemmed points recovered in the Eagle Lake Project area are all fine grainedbasalt (Plate 2, i-m). Three (Plate 2, i-k), including one each from the Shieids (EkSa 13, k) and Boyd (EkSa 32, i) sites, are relatively thick, with neck and stem of nearly equal width. Only the specimen from the Bidwell Creek site (ElRw 4) (Plate 2, j) tapers slightiy from neck to base; one edge is worn to a polish. The other two specimens (Plate 2, 1-rn), including one from the Boyd site (EkSa 32), have unifacial or bifacial thinning of blade and stem: the thinned base of the Boyd site specimen (Plate 2, m) is concave. May 18, 2003-236 The stemmed points recovered in the Eagle Lake Project area are similar in shape to Stryd's Stemmed Atlatl Points (1973:327-28) but are smaller and fall within the general size (but not thickness) parameters of his Stemmed Arrow Points (1973:325). Stryd reports that both these types are found in Kam.loops components. In terms of size, all the points here appear to be shorter than any of the atlatl points reported by Stryd, and the widest one (22.8 mm) from Eagle Lake does not quite match the narrowest (23.4 mm) reported by Stryd; only one from Eagle Lake is as thick as any from Lillooet. Thus these points appear to be significantly smaller than Stryd' s Stemmed Atlatl Points. One example from the surface of the Boyd site (Plate 2, m) might be better compared with Sanger's (1970:44) Group 12 \"Single Basal Notch\" class that includes a variety of outlines and is found in the pre-PPT component of the Lochnore Creek site (EdRk 7). Figure 22, \"d\" in Sanger shows a very similar object to the Boyd site point. With this exception, the stemmed point class from the Eagle Lake Project area appears to date to the last half of the PPT Paull (1984) tested EkSa 32: 1046 (Plate 2, i) for blood at one loci and it was positive. Table V-8 Artifact No. EkSa 32: 1046 EkSa 13:2418 ElRw 4:1358 ELP 27:OS:8 CR 40:128 Stemmed Points: Metric Data Length C Width C Thickness 36.3 Y 15.6 Y 6.7 30.9 N 21.0 N 8.0 37.0 N 12.0 Y 5.5 28.6 N 16.8 Y 5.0 21.7 N 22.8 N 6.4 5. Miscellaneous Points MIPT Weight 3.66 4.72 2.46 2.30 3.26 n=3Two unnotched points and one fragment, all fine-grained basalt, were recovered from theBrittany Creek site (EkSa 33) (Plate 2, n-p ). The two complete specimens (Plate 2, n-o ), although dissimilar, both fall into Stryd' s Unnotched AITow Point Type that is affiliated with the May 18, 2003-237 j b C d e f k m n V p q r s t u 0 1 2 3 4 5cm Plate 3 Point Fragments Fragments of Small Points (PTFRs) a. CR 92:2284 b. EkSa 33:3483 C. ELP 19-1:158d. ELP32:1242 e. CR 92:590 f. T84-27:1270g. ELP 19-1:252 h. T84-27:2013 I. ELP 19-1 :880J. T84-27:l017 k. ElRw 4:1959 1. EkSa 33:3378m. EkSa 33:3422 n. CR 92:1199 o. EkSa 33:2863Framents of Large points (PTFRl):p. CR28:44 q. CR 28:47 r. ELP 16-1:25s. ELP26 OS:l t. ELP 14 OS:l u. EkSa 32:6208? '{_, Kamloops culture (1973:323-24, Fig 22 a-g). Both may be arrow point preforms. The fragment (Plate 2, p) may also be a preform, in this case for a small comer-notched point with squared shoulders and an expanding stem. Similar points are described by Stryd 1973:334-35, Fig 24) as Corner Notched Arrow Points Type C and are primarily affiliated with the Kamloops Horizon. Table V-9 Artifact No. EkSa 33:208 EkSa 33:2477 EkSa 33:3330 Miscellaneous Points: Metric Data Length C Width C Thickness 19.5 N 15.6 N 2.8 24.3 Y 13.1 N 4.1 32.3 Y 13.8 Y 5.0 6. Point FragmentsFragments of Small Points PTFRs Weight 1.04 I. 171.86 n = 13 Fragments of small, bifacially flaked points (Plate 3, a-o) include thirteen tip fragments (Plate 3, a-m), one medial fragment (Plate 3, n), and one medial fragment with base (Plate 3, o). Eleven tip fragments are from small, triangular, symmetrical points (Plate 3, a-k): three are obsidian; the remainder are fine-grained basalt. Two tip fragments (Plate 3, 1-m), both from the surface of the Brittany Creek site (EkSa 33), are slightly asymmetrical: one, (m), is fine-grained basalt, the other, (1), is chert. The medial/base fragment (Plate 3, o ), also from the Brittany Creek site, is probably from a side-notched point. Fragments of Large Points PTFRl n=6Fragments of large points include four medial fragments (Plate 3, p-s) and two tip fragments (Plate 3, t-u). The four medial fragments are fine-grained basalt and were surface collected (Plate 3, p-q). One tip fragment of obsidian (Plate 3, t), was on the surface of ELP May 18, 2003-238 b O 1 2 3 4 5 Quad 14. The other fragment of fine-grained basalt (Plate 3, u), complete except for the missing base, is from the excavation at the Boyd site (EkSa 32). Its blade is similar to Stryd's Notched Atlatl Point Type A affiliated with the Kamloops culture (1973:336-37, Fig 25 g). It has a bilaterally symmetrical blade with maximum width at the shoulders and shallow, semi-circular notches. The largest number of point fragments recovered at one site is seven, collected on the surface at the Brittany Creek site (EkSa 33) (Plate 3, b,i,m,o). The Fish Trap Lake site (EkSb 37) and ELP Quad 19:1 (EkSa 27) each yielded three specimens (Plate 3, f, h, j) and (Plate 3, c,g, i) respectively and one was recovered at the Bear Lake site (EkSa 36) (Plate 3, d) and the Bidwell Creek/Quiggly Holes site (ElRw 4) (Plate 3, k). 7. Large Formed BifacesComplete Large Formed Bifaces LFBI n=4 The complete, large formed bifaces include an igneous tci-thofrom the Bear Lake site (EkSa 36) (Plate 4, a), two bifaces of basalt with attributes similar to the tci-tho(Plate 4, b-c), and a large, basalt, stemmed biface, also from the Bear Lake site (EkSa 36) (Plate 4, d). The tci-tho,sometimes defined as a bifacially retouched scraping tool, is a distinctive Athapaskan artifact class (Dumond 1978:55, Wilmeth 1978, Clark 1976:68-69) that includes cortex spalls. On the specimen from the Bear Lake (EkSa 36) excavation, (Plate 4, a), one lateral margin exhibits bifacial retouch whereas the other lateral margin has been worn to a rounded contour. Both faces from the worn edge to the midpoint are smoothed and reddish, possibly the result of animal residues. Paull (1984) tested this object at three loci for blood and fat, and at one of the loci, for Lignin. One loci was negative for blood and fat, one positive for May 18, 2003-239 all three, and one positive for blood and negative for fat. Thus the inference that these objects are used in hide processing appears to be supported by these tests. The biface from ELP Quad 16 (EkSb 5) (Plate 4, b) is coarse-grained basalt. Although all flaked margins show evidence of fine polish from use, polish predominates on the tip and the thinnest margin, suggesting the specimen is homologous to the tci-thodescribed above. The other fine-grained basalt specimen, recovered at CR 64 (EkSa 34), (Plate 4, c), is asymmetrical and resembles a \"miscellaneous\" biface described by Sanger (1970:73-74, Fig 31-p) but is more crudely made. The curved edge exhibits finer, discontinuous, unifacial retouch or use-wear which suggests a function similar to the tci-tho. The stemmed biface from the Bear Lake site (Plate 4, d) is beautifully made and has a slightly steeper angle on the lateral margin with the deeper shoulder notch. The specimen was recovered in seven pieces, six from Unit 55 and one from Unit 58. Oniy the very tip of the blade is missing. Compared to the PPT bifaces reported by Stryd ( 1973) and Sanger ( 1970), this specimen is very large, and has a unique shape. It is probably not an accident that this shape is the same as seen in the Kavik point as it is from the prehistoric component (Layer Bf) at the Bear Lake site. Table V-10 Large Formed Bifaces (tci-tho--likt:1: Metric Data Artifact No. Length C Width C Thickness ELP 32:2802 97.1 Y 75.5 Y 22.7 CR64:OS:l 94.4 N 59.7 Y 17.1 ELP 16-1:24 58.8 Y 69.1 Y 21.0 ELP 36:4971, 122.8 Y 44.4 Y 13.4 4972,4973,4974,4975,5533 Large Formed Biface Fragments LFBF Weight 181.30 62.9 93.17 60.76 n=5 Five large biface fragments were recovered (Plate 5, a-e). Two, one from the Boyd site Mav 18, 2003-240 .I (EkSa 32) (Plate 5, a), and one from the Bear Lake site (EkSa 36) (Plate 5, b ), are morphologically similar to the proximal ends of the tci-tho-like bifaces described above. Two smaller fragments (Plate 5, c-d), both from the Bear Lake site (EkSa 36), also exhibit simiiar flaking patterns. EkSa 36:3004 (Plate 5,c) was tested by Paull (1984) at three loci for blood and fat residues. One loci was negative for both, one positive for both and one positive for blood but negative for fat, indicating probable use for flesh or hide processing. EkSa 36:3984 (Plate 5,b) was also tested by Paull at 4 loci for blood and three of these were positive, indicating that these specks were blood. One biface fragment from CR 28 (EkSa 98) (Plate 5, e) is very fine-grained basalt with an unusual form: one face exhibits a long, concave flake scar (probably a hinged flake removal from an attempt to thin the other lateral margin). The intact lateral margin exhibits unifaciai use-wear. u. Small Formed Bifaces SFBI n =43 Twelve complete small formed bifaces and 30 fragments were recovered in the Eagle Lake Project area and four forms were noted: Oval (Plate 5, f-k); Leaf-shaped/Ovate (Plate 6, a\u00C2\u00ADh); Triangular (Plate 6, i-n); and Miscellaneous (Plate 6, o). Fragments were assigned to these types where possible; the remaining fragments (not illustrated) are described (following Stryd 1973:348) as: End sections; Medial sections; and Unclassifiable fragments. Table V-11 Complete Small Formed Bifaces All Fo1ms: Metric Data # of Specimens Range Mean Length 5 32.4 - 50.7 42.0 Width 11 9.9 38.4 22.5 Thickness 12 3.9 - 17.2 8.6 May 18, 2003-241 d e f k o 1 2 3 4 5 cm V Plate;5 Large Formed Biface Fragments and Small Oval Formed Bifaces Largt? Formed Biface Fragments (LFBF): a. EkSa 32:5830 b. EkSa 36:3984 c. Ek:Sa 36:3004d. EkSa 36:4402 e. CR 28:5Small Formed Bifaces, Oval (SFBI):f. CR 40: 1 g. Ek:Sa 13:4337 h. Ek:Sa 13: 1889i. T84-27:1070 j. EkSa 33:3325 k. ELP 44-3:200a b C d e f g h I 'm 0 1 2 3 4 5cm Small Formed Bifaces (SFBI) complete: ELP 19-1 :980 b. CR 50:1 C. EkSa 33:3205EkSa 13:1385 e. EkSa 13:2419fragments: EkSa 33:4348 g. EkSa 13:4339 h. EkSa 36:4355Triangular complete: i. CR 98:5 j. EkSa 32:6295Triangular fragments:k. ELP 19-1:96 1. CR 92:1425 m. EkSa 33:3489Miscellaneous Small Formed Bifaces:n. Ek.Sa 36:4841 o. CR 92:1906Oval Small Formed Bifaces n=5 Three c omplete small, oval bifaces of fine-grained basalt (Plate 5, f-h), including two from the Shields site (EkSa 13), and two fragments of basalt (Plate 5, i-j) were recovered in the Eagle Lake project area. Cortex is visible on four of the basalt specimens: both complete specimens from the Shields site (EkSa 13) (Piate 5, g-h), the complete specimen from CR 40 (Plate 5, f), and one fragment (Plate 5, i). One small biface of green stone (Plate 5, k) also has some cortex remaining. The Eagle Lake area small, oval bifaces are comparable to Stryd's Biface Type Group E, (1973:231-32, 348), and as Stryd suggested, may be blanks for secondary flaking to produce various tools. Leaf-Shaped/Ovate Small Formed Bifaces n=8 Eight small bifaces recovered in the Eagle Lake area (Plate 6, a-h) are similar to Sanger's Leaf-shaped Biface Group 5 (1970:73, Fig 31-m,n) and Stryd's Ovate Bifaces Group A (1973:345-346, Fig 27a,b and Fig 28 a-d). The five complete specimens (Plate 6, a-e) and three end fragments (Plate 6, f-h) are basalt, have an ovate form, and are lenticuiar in section. The two complete bases, both from the Shields site (EkSa 13), have asymmetrical notching at the proximal end and evidence of retouch on the edges. Four medial fragments, described below but not illustrated, may also be from leaf-shaped bifaces. Triangular Small Formed Bifaces n = 11 Eleven small biface specimens (Plate 6, i-m) are similar to Sanger's Triangular Biface Group 3 (1970:73, Fig 31, f-i), although one (Plate 6, i) has an elongated form and wear attributes similar to Stryd' s Triangular Perforator ( 1973:Fig 28r). This specimen, from CR 98 May 18, 2003-242 (EkSa 62), is obsidian with a thick cross-section and a flat ventral face. The lateral margins exhibit fine retouch/ grinding which may be similar to what Stryd refers to as '' characteristic .. . light use modification of the lateral shaft margins\" (1973:350). However, unlike Stryd's examples, wear is not at the tip of the Eagle Lake specimen, but is equally distributed along both lateral margins from the tip to the mid-point of the specimen. The remaining triangular small formed bifaces are represented by one \"complete\" specimen (Plate 6, j) and 9 distal end (tip) fragments, one of obsidian (Plate 6, k) and eight of basalt (Plate 6. 1-m). Although some of these specimens are lenticular in cross-section all are thinner than the possible perforator described above (Plate 6, i). Lateral margins are primarily slightly convex and only one specimen, from Brittany Creek (EkSa 33) (Plate 6, 1) has straight margins. No intact bases were recovered. Miscellaneous Small Formed Biface n = lA smali, crescent-shaped, basalt biface with continuous flaking on both margins, no polish or use wear, and with both ends snapped off was recovered during excavations at the Bear Lake site (EkSa 36) (Plate 6, n). The crescent-shape lateral margin was bifacially flaked to produce a steep, sharp edge. Neither Sanger ( 1970) nor Stryd ( 1973) illustrate artifacts of this type. Medial Fragments of Small Formed Bifaces n=5Five medial biface fragments were recovered in the Eagle Lake Project area although none were recovered from the Bear Lake (EkSa 36), Shields (EkSa 13), or Boyd (EkSa 32) sites. All are from either oval or ovate (leaf-shaped) small bifaces and all have lenticular cross-May 18, 2003-243 a b e O 1 2 3 4 5cm Plate 7 Formed Scrapers (FOSC) k a. EkSa 13:5431 b. EkSa 13:2169d. EkSa 13:1391 e. EkSa 13:1564g. EkSa 32:2513 h. CR 92:564j. CR 40:49 k. CR 40:2h C. f. i. L EkSa 13:5257 EkSa 32:2099 CR 92:826 EkSa 33:4390 I it, sections with one of obsidian thicker than the four basalt specimens. End Fragments of Small Formed Bifaces n=5Five basalt end fragments, probably from triangular Kamloops side-notched points, include four thin proximal ends and one thicker fragment from the Brittany Creek site (EkSa 33) (Plate 6, o). This specimen is possibly a distal end (tip) and has some cortex and fine continuous, unifacial flaking on all four margins. The notch visible on one lateral margin appears to have been an unintentional by-product of a flake removal. The other four proximal end fragments have flat ends and straight lateral margins. Unclassifiable Fragments of Small F01med Bifaces n=6Six unclassifiable fragments of bifaces were recovered: one, from the Brittany Creek site (EkSa 33), is obsidian; all others are basalt. Retouched Flake tools 9. Formed Scrapers FOSC n = 12Following Stryd's formed scraper typology based on the primary location of retouch(1973:352-361 ), two types of scrapers were recovered from the Eagle Lake Project area: Continuous Scrapers (n = 6) (Plate 7, a-c, f, h,j) and End Scrapers (n = 6) (Plate 7, d-e, g, i, k-1). All were made from flakes and have marginal retouch of relatively uniform height forming an angle of 45 degrees or steeper with the ventral face. No spuned scrapers were recovered in the Eagle Lake Project area. May 18, 2003-244 Continuous Scrapers The six continuous scrapers have unbroken retouch around the edge, exclusive of the striking platform, and all retouched edges exhibit use wear. All are made of fine-grained material: chert, obsidian, or fine-grained basalt. Only one specimen (from the Boyd site, EkSa 32) has cortex on the ventral face (Plate 7, f). Three continuous scrapers were recovered at theShields site (EkSa 13) (Plate 7, a-c) including the only specimen with a complete oval outline (Plate 7, a). Another specimen from the Shields site (Plate 7, c) has unifacial retouch on alternating faces: the dorsal face on two edges and the ventral face on the third edge. The \"incomplete\" continuous scraper specimens have oval distal ends, however, as noted by Stryd, the proximal ends are missing, perhaps by intent ''to facilitate hafting\" or by accidental breakage (1973:360, Fig 30-u). Table V-12 Continuous Scrapers: Metric Data Artifact No. Length C Width C Thickness EkSa 13:5431 23.8 y 20.0 y 7.9 EkSa 13 :2169 24.6 N 18.3 y 5.2 EkSa 13:5257 24.2 N 20.5 y 5.6 EkSa 32:2099 17.0 N 10.7 y 4.0 EkSa 33:564 15.9 N 15.1 N 4.7 CR 40:49 19.7 N 21.8 N 3.5 End Scrapers WeightMaterial 4.02 Obsidian 2.03 Chert 3.33 FG Basalt 0.80 Chert 1.48 Obsidian 1.93 FG Basalt End scrapers have one rounded end produced by retouch (Magne 1985: 168), the primary w orking edge is located transverse to the long axis of the tool and the lateral edges exhibits less intense use-wear (Stryd 1973:353). All end scrapers recovered in the Eagle Lake Project area are based on flakes. Two end scrapers were recovered at the Shields site (EkSa 13) (Plate 7, d-e) and one, of chert, has cortex (Plate 7, d). One was recovered at the Boyd site (EkSa 32) (Plate May 18, 2003-245 7, g). All end scrapers have unifacial retouch on the dorsal face (Plate 7, d, e, g, i, k, l) and are similar to Stryd's Convex End Scraper Types (1973:354-55, Fig 30a-n) which he notes are similar to Sanger's Formed Uniface Groups 1, 2, 3, and 4 (round to oval unifaces, stemmed or tanged unifaces, and round to oval thick unifaces respectively) (Sanger 1970:78, Fig 33a-h). Table V-13 End Scrapers: Metric Data Artifact No. Length C Width C Thickness Weight Material EkSa 13:1391 30.1 N 23.8 y 8.0 3.98 Chert EkSa 13:1564 23.4 y 26.8 y 8.4 5.15 FG Basalt EkSa 32:2513 30.7 y 22.8 N 7.0 4.17 FG Basalt EkSa 33:826 28.7 y 27.3 y 5.4 5.89 CG Basalt CR 40:2 28.1 N 24.7 y 5.7 3.96 FG Basalt EkSa 33:4390 34.4 N 29.7 N 4.3 6.52 CG Basalt 10. Spurred Scrapers n=O No spurred scrapers were recovered in the Eagle Lake project area.lL Bifacially Retouched Flakes n = 81 Like Sanger's Non-formed Biface Type (1970:76), the bifacially retouched flakes from the Eagle Lake Project area have little deliberate shaping, are primarily basalt [ over 90% J ( obsidian n = 7, chert n = 1 ), and many specimens are pieces of larger artifacts. Sanger's bifaces were recovered in largest quantities in the Upper Middle Period (3500 -2000 BP) but were present, in decreased quantities, in the Late Period (1500 -500 BP) (1970:76, 105). They are similar to Stryd's Retouched Flakes (both unifacial and bifacial) which he describes as \"Flakes of irregular form with limited marginal retouch .. [that] .. display either unifacial or bifacial retouch ... usually ... restricted to one edge or part of one edge. Characte1ized by little or no wear, these flakes probably served as short use all-purpose cutting and scraping implements\" (1973:365). May 18, 2003-246 a d e o 1 2 3 4 5cm Plate 8 Large Bifacially Retouched Flakes (BIREl) a. EkSa 32:2097 b. ElRw 4:1355 c. d. CR 12: 1 e. EkSa 33:3386 f. ELP 20 OS:1 ELP 44-3:196 Both Sanger and Stryd note that, given the fragmentary nature of the specimens, measurements have little meaning, although Sanger reports that lengths range from 10 to almost 70 mm with the majority in the 20 to 40 mm range (1970:76). The Eagle Lake area specimens are reported as two sub-sets: large (complete length greater than or equal to 35 mm), and small ( complete length less than 35 mm) with length measured as the maximum dimension from the striking platform of the flake to the distal end. Large Bifacially Retouched Flakes BIREl n = 17 All large bifacially retouched flakes are basalt (fine-grained basalt n = I 0, coarse-grained n = 7) (Plate 8, a-f). Six specimens are complete (Plate 8, a, d-f), including one from the Boyd site (EkSa 32) excavation (Plate 8, a) and one of fine-grained basalt with continuous retouch on the end and lateral margins (Plate 8, d). With the exception of the Boyd site specimen and two from the Bidwell Creek (ElRw 4, Plate 8b) excavation, all specimens were recovered on the surface (fable V-14). Table V-14 Large Bifacially Retouched Flakes: Metric Data # of Specimens Range IQR Length 11 38.5-79.6 45.9-57.5 Width 8 23.3-58.7 32.1-37.0 Thickness 17 6.0-15.5 8.8-12.7 Weight 6 8.36-70.57 6.57-32.23 Small Bifacially Retouched Flakes BIREs Mean 56.9 37.4 10.5 27.81 n =64 The small bifacially retouched flakes (Plate 9, a-p) are basalt with the exception of seven obsidian specimens and one of chert (Plate 9, k). Fifteen specimens are complete (Plate 9, f,h,j,l,o-p) and five of these have cortex (Plate 9,f,1). Six fragments also have cortex. Nine May 18, 2003-247 b d \u00E2\u0080\u00A2 g h k p 0 O 1 2 3 4 5cm Plate 9 Small Bifacially Retouched Flakes (BIREs) a. EkSa 13:1422 b. EkSa 13:3959 C. EkSa 32:1049d. EkSa 32: 1054 e. EkSa 32: 1068 f. EkSa 32:6287g. ELP 32:104 h. ELP 32:184 1. E1Rw4:1347j. ElRw 4:1680 k. EkSa 33:3236 1. EkSa 33:3522m. ELP 19-1:159 n. ELP 19-1:660 o. ELP 19-1:318p. ELP 19-1:700specimens were recovered from the excavation at the Shields site (EkSa 13) (Plate 9, a-b) including one of obsidian; five at the Boyd site (EkSa 32) (Plate 9, c-f), including one of obsidian; and six at the Bear Lake site (EkSa 36) (Plate 9, g-h), again with one of obsidian. Thirteen were recovered at ELP Quad 19:1 (EkSa 27) (Plate 9, m-p) including a specimen with a very worn end that may have been used as a scraper (Plate 9, n). Table V-15 Small Bifacially Retouched Flakes: Metric Data # of Specimens Range IQR Mean Length 28 18.6-34.2 21.9-31.5 26.9 Width 33 8.5-55.9 12.3-23.7 20.7 Thickness 64 1.9-14.3 3.4-6.7 5.37 Weight 15 0.54-9.23 1.41-4.92 3.72 12. Unifacially Retouched FlakesLike Sanger's Non-formed Unifaces (1970:80) and Stryd's Unifacially Retouched Flakes (1973:366), the specimens from the Eagle Lake Project area exhibit no deliberate shaping and many are fragmentary. For the Eagle Lake Project assembiage, this type is defined as having retouch along a single edge only and is, therefore, similar to Sanger's Group ! to Group 6 Non\u00C2\u00ADformed Unifaces (1970:80-81). (See also multiple-edged unifaces and sinuous, multiple edge unifaces described below for other types of non-formed unifaces.) Large u nifacially retouched flakes have maximum length of complete specimen equal to or greater than 44 mm (Plate 10, a\u00C2\u00ADc) and small specimens have maximum length of complete specimens less than 44 mm (Plate 10, d-k).Fifteen (of sixteen) large specimens have cmtex, whereas only seven (of 60) small specimens do. Large unifacially retouched flakes are primarily basalt with two each of chert (Plate 10, b) and igneous material (Plate 10, c). The 60 small unifacially retouched flakes are May 18, 2003-248 --------------v// a b C d e a f g O 1 2 3 4 5cm Plate 10 Unifacially Retouched Flakes Large (UNREl): a. T84-27:l b. EkSa 13:3038 C. ELP G5-l:5Small (UNREs):d. EkSa 36:4478 e. EkSa 36:5934 f. EkSa 13:5351g. EkSa 13:3720 h. EkSa 13:1575 1. EkSa 13:1386j. EkSa 32:1053 k. CR 92:131/Dt,._,, also primarily basalt, with four of chert (Plate 10, j) and nine of obsidian (Plate I 0, f-g). No large, unifacially retouched flakes were recovered in the Bear Lake site (EkSa 36) excavation, although eight small ones were (Plate 10, d-e). The Shields site (EkSa 13) yielded five large (Plate 10, b) and twelve small unifaces (Plate 10, f-i), and two large and eight small unifaces (Plate i 0, j) were recovered at the Boyd site (EkSa 36). 13. Utilized Flakes UTIL n =43Utilized flakes are chipped stone flakes with macroscopic use-wear polish, striations, andirregularly spaced or continuous flake scars along at least one edge (Ludowicz 1980:98-10 I). The specimens from the Eagle Lake Project area are primaiily basalt (Plate 11, j) but include six flakes of obsidian, two chert, and one of quartz (Plate 11, a). Sixteen are complete flakes with utilization; the remainder are fragments of utilized flakes. Flake size varies considerably: weight of the twelve complete basalt flakes ranges from 0.22 grams (Plate 11, d) to 146.45 grams. One quartz complete flake is 42.87 grams and three obsidian complete flakes range from 0.50 to 2.65 grams. The assemblage has been separated into thirty-five small (Plate 11,c-i), seven medium size (Plate 11, a-b, j-k), and one Iai\u00C2\u00B7ge size utilized flake. Small flakes have a maximum length of less than 40 mm, medium flakes have maximum lengths between 40 and 100 mm, and the large utilized flake has a maximum iength over I 00 mm. Table V-16 Utilized Flakes: Metric Data # of Specimens Range Length 22 10.5-112.2 Width 25 8.4-63.1 Thickness 43 1.2-21.9 Weight 17 0.22-146.56 Weight* 16 0.22-42.87 * not including the largest flakeIQR 15.7-27.2 11.7-23.1 2.9-6.4 5.6 0.65-3.17 0.65-2.65 May 18, 2003-249 Mean 29.5 21.2 13.6 5.24 d b C e f g 0 1 2 3 4 5cm Plate'll Utilized Flakes (UTIL) Medium: a. CR 73:84 b. E1Rw4:1877 j. ELP 19-1:1172k. ELP 19-1:612Small:c. E1Rw4:2030 d. T84-27:1431 e. EkSa 32:5737f. EkSa 33:3181 g. Ek:Sa 33:3494 h. EkSa 33:4330i. EkSa 33:4393JJ/2 Small amounts of conex remain on eight of the thirty-five small, utilized flakes, often on the striking platform or one lateral margin. Cortex remains on alJ medium size flakes (n = 7) and includes four with full dorsal cortex (Plate 11, a, k), two with platform cortex (Plate 11, b, j), and one of fine-grained basalt with cortex on the distal end (CR 64:48, not illustrated) that may be a utilized c ore rejuvenation flake. The single large specimen also has a small amount of cortex on the dorsal surface (- 5%). Most utilized flakes were collected on the surface although eighteen were recovered during excavation at various sites including the Boyd site (EkSa 32) (Plate 11, e). However, none were recovered at either the Shields (EkSa 13) or Bear Lake (EkSa 36) sites. The surface of the Brittany Creek site (EkSa 33, CR 92) yielded fifteen utiiized flakes (Plate 11, f-i)including four of obsidian and two of che11 in addition to the fine- grained and coarse-grained basalt specimens. Most of the specimens from this site are small. weighing less than four grams, but include a fragment of a coarse-grained basalt utilized flake weighing 18.8 grams and a complete, fine-grained basalt utilized flake weighing 146.46 grams. Two flakes with utiiization on two edges were recovered at ELP Quad i 9: 1 (EkSa 27) (Plate 11, j-k). One (Plate 11, j) hasutilization flaking on the ventral side only on both edges (and a large flake removed on the ventral face to facilitate holding). The other (Plate 11, k) has use-wear flaking on the dorsal side (which also has full cortex). 14. Multiple-edged Unifaces MUUT n = 21Multiple edged unifaces have continuous retouch (or alternate edge retouch) on morethan one edge and retouched edges are relatively straight. If edges are S-curved, the unifaces fall into the next category, Sinuous Multiple-Edged Unifaces. Multiple-edged Unifaces, May 18, 2003-250 ,,. b C d a e f O 1 2 3 4 5 cm g k Plate 12 Multiple-edged Unifaces (MUUT) a. Ek.Sa 13:3965 b. EkSa 13:3964d. Ek:Sa 13:2763 e. EkSa 13:2348g. Ek.Sa 13:5414 h. EkSa 32: l 028j. EkSa 32: 1042 k. EkSa 32: 1047h C. f. 1. 1. EkSa 13:4611 EkSa 13:5411 EkSa 32:2550 EkSa 36:4474 Sinuous Multiple-Edged Unifaces, Gravers, and Perforators all typically have multiple working edges and points, and are usually made on thin flakes of fine-grained basalt or obsidian. In these characteristics they fit what Judge ( I 973: 108-110) called \"Utility Flakes\" in his analysis of Paleolndian assemblages. If these objects lack graving or perforating tips they fall into the two multiple-edged uniface classes; most Gravers and Perforators also have multiple unifacial retouched edges. Multiple-edged Unifaces were recovered only from excavations at the Shields (EkSa 13) (n = 13) (Plate 12, a-g), Boyd (EkSa 32) (n = 7) (Plate 12, h-k), and Bear Lake (EkSa 36) (n = 1) (Plate 12, 1) sites. Two specimens, both from the Boyd site (EkSa 32), exhibit alternating retouch (dorsal on one edge, ventral on the other) (Plate i2, h, k). Mate1ials are predominantly fine-grained basalt, with two obsidian unifaces recovered at the Shields site (EkSa 13) (Plate 12, a-b). The concentration on the PPT housepit sites indicates they may have value in indicatingethnicity, a subject investigated further in the next chapteL Table V-17 Mutliple-edged Unifaces: Metric Data # of Specimens Range IQR Mean Length 9 19.8-41.4 27.7-34.9 30.7 Width 17 8.2-26.4 15.5-20.0 18.4 Thickness 21 2.0-6.3 3.0-4.1 3.5 Weight 8 0.79-3.77 1.47-2.30 2.00 15. Sinuous, Multiple-edged Unifaces SINU 11=8Sinuous, multiple-edged unifaces are a variant of multiple-edged unifaces (continuousretouch on more than one edge) that exhibit a pronounced sinuous outline (incurvate-excurvate) (Plate 13, a-h). Like the multipie-edged unifaces, sinuous edged unifaces may have retouch (or use wear) on alternating faces. Sinuous edged specimens were recovered only at the Shields May 18, 2003-251 b C a 4 e 0 1 2 3 4 5cm Plate 13 Sinuous, Multiple-edged Unifaces (SINU) d a. EkSa 13:1054 b. EkSa 13:1888 c. d. EkSa 13:3950 e. EkSa 32:2095 f. g. EkSa 32:2510 h. EkSa 32:6197h EkSa 13:2063 EkSa 32:2096 (EkSa 13) and Boyd (EkSa 32) sites. All are basalt and all are complete flakes, except one (Plate 13, e). The Shields site (EkSa 13) specimens (Plate 13, a-d) are larger than those recovered at the Boyd site (EkSa 32) (Plate 13, e-h). As with the previous category, these may also be indicative of ethnicity, and are discussed further in the next chapter. Table V-18 Sinuous, Multiple-edged Unifaces: Metric Data by Site Boyd Site n=4 Shields Site n=4 Range Mean Range Mean Length 32.0-35.7 34.0 37.1-57.6 47.8 Width 13.7-24.4 18.6 26.0-35.5 3Ll Thickness 3.0-5.6 4.0 5.2-6.0 5.5 Weight 1.08-3.01 1.79 3.54-8.43 6.88 16. Gravers GRAV n = 31Gravers are flakes that have a pronounced projection, refeITed to as a shaft and tip (Stryd1973:361-364) or a spur (Sanger 1970:83-84), and must have use-wear on the tip (Stryd 1973:361-364). Stryd distinguishes between \"sharp\" gravers with sharp tips blunted through use (1973:361-364,Fig 31, a-g), and \"rounded\" gravers with rounded, unifacially retouched shaft tips (Stryd 1973:361-364, Fig 31, h-m). The gravers collected in the Eagle Lake Project area (Plate 14, a-j) are all \"sharp\" gravers with blunted tips, all are basalt, and all, except one, were recovered in excavations. The surface-collected specimen is from the Fish Trap Lake site (T84:27, EkSb 37) (Plate 14, a). Eighteen gravers were recovered at the Shields site (EkSa 13) (Plate 14, c-f) and two specimens (Plate 14, d-e) are almost identical. Eleven are from the Boyd site (EkSa 32) (Plate 14, g-j): one specimen (Plate 14, h) has more than one graving tip; and and only one at the Bear Lake site (EkSa 36) (Plate 14, b), a somewhat dubious member of this class. As with the previous two flake tool classes, these may also have value in identifying ethnicity, as discussed further in the next chapter. Although each specimen has the defining attributes of May 18, 2003-252 a f k C b g m q p r O 1 2 3 4 5cm d e n 0 s t Plate 14 Gravers, Alternating Pert'orators, Non-alternating Perforators Gravers (GRAV): a. T84-27:14d. EkSa 13:4327g. EkSa 32:5708J. EkSa 32:2512b. e. h. Alternating Perforators (PERFa):EkSa 36:4978 EkSa 13:4686 EkSa 32:6202 k. EkSa 13:3205 1. EkSa 13:4793n. EkSa 32: 1057 o. ELP 19-1 :324Non-alternating Perforators (PERFn):p. EkSa 36:1059 q. EkSa 13:2357s. Ek.Sa 32:2519 t. EkSa 32:6199c. f. I. m. r. EkSa 13:2347 EkSa 13:5685 EkSa 32:6204 EkSa 32: 1056 EkSa 13:3039 gravers, many have use-wear and retouch on lateral margins as well, which indicates that the tools were used for more than one type of function. Bone and antler artifacts recovered at the Shields site (Ek:Sa 13), including leister points, harpoon points and decorated fragments, may have been produced using gravers. Table V-19 Gravers: Metric Data # of Specimens Range Length 17 19.3-38.0 Width 25 14.0-28.5 Thickness 31 2.4-19.2 Weight 12 1.03-4.72 17. Perf orators PERF IQR 24.3-29.2 17.9-21.8 3.4-4.1 1.27-1.88 Mean 26.9 19.8 4.3 1.79 n = 19Sanger grouped perforators and drills together (1970:84, Fig 34 i-1) as only eight wererecovered in his study. Sanger's perforators and drills share a number of characteristics: elongated points on bifacially chipped flakes, usually basalt, and a length range of 19 mm to 63 mm. Sanger distinguished drills by the presence of wear polish in a pattern created by rotaryaction; those without wear polish were categorized as perforators. In the Eagle Lake Project area, perforators were recovered that are similar to Sanger's type (Plate 14, k-t) and have the attributes defined by Stryd as well (1973:350-52): a long, sharply tipped shaft that is generally unifacially flaked (although some have alternate and bifacial retouch ), piano-convex shaft cross\u00C2\u00ADsections, light use modification of the lateral shaft margins with most wear at the tip, and (commonly) broken tips. Only ELP Quad 19:1 (EkSa 27) yielded obsidian perforators. The majority of specimens from the Shields (EkSa 13), Boyd (EkSa 32), and Bear Lake (EkSa 36) sites are very fine-grained basalt. No clearly bifacially retouched \"drills\" were recovered. The Shields site (EkSa 13) yielded ten perforators; one is chert and the others are May 18, 2003-253 fine-grained basalt. At the Boyd site (EkSa 32) six perforators, all fine-grained basalt, were recovered and only one perforator was recovered at the Bear Lake site (EkSa 36), a crude specimen of large size (Plate 14, p). Two perforator fragments were recovered at ELP Quad 19: 1 (Ek:Sa 27). The Eagle Lake Project area perforators are separated into types based on the location of retouch. Alternating perforators (Plate 14, k-o) exhibit unifacial retouch on alternate faces and non-alternating perforators (Plate 14, p-t) exhibit unifacial retouch on one face only. Reference is also made, where applicable, to butt modifications as defined by Sttyd: oval ( 1973: Fig 28-s), irregular, triangular or unclassifiable (1973:351-2). Alternating Perf orators PERFa n=8 Alternating perforators exhibit unifacial retouch on alternate faces (Plate 14, k-o). Only one specimen is obsidian (Plate 14, o); it is from ELP Quad 19: 1 (EkSa 27) and the butt was unclassifiable. All other alternating perforators are basalt and are from excavations at the Shields site (EkSa 13) (n = 4) (Plate 14, k-1) and Boyd site (EkSa 32) (n = 3) (Plate 14, m-n). The specimens from the Shields site include one oval butt perforator (Plate 14, k), one triangular perforator, and four irregular butt perforators that have incorporated the iITegular shape to create a haft or handle on one lateral edge (Plate 14, l). The three alternating pe1forators from the Boyd site are all oval butt specimens. Table V-20 Alternating Perforators: Metric Data # of Specimens Range Mean Length 5 28.1-45.5 32.9 Width 7 18.0-32.7 23.1 Thickness 8 3.5-8.3 4.8 Weight 5 1.59-6.94 3.32 May 18, 2003-254 Non-Alternating Perforators PERFn n = 11 Non-alternating perforators from the Eagle Lake Project area (Plate 14, p-t) are all fine\u00C2\u00ADgrained basalt with two exceptions: an obsidian perforator from the surface of ELP Quad 19: 1 (EkSa 27) that is fragmentary and unclassifiable and a large, oval butt perforator of chert from the Shields site (EkSa 13). Five other non-alternating perforators recovered at the Shields site include one triangular butt perforator and four irregular butt perforators that range from the smallest recovered in the Eagle Lake Project area (Plate 14,q) to large ones including one (Plate 14, r) with a unique, long shaft. At the Boyd site (EkSa 32), three non-alternating perforators were recovered: one triangular (Plate 14, s), one with an ovai butt (Plate 14, t) and one unclassifiable fragment (missing both base and tip) that is similar to the oval butt perforator. Only one large, non-alternating perforator was recovered at the Bear Lake site (EkSa 36) (Plate 14, p): it has an irregular butt similar to those described by Stryd (1973:352). Table V-21 Non-Alternating Perforators: Metric Data # of Specimens Range Mean Length 4 20.3-50.8 34.8 Width 8 8.2-42.9 24.6 Thickness 11 3.1-9.6 5.4 Weigh 4 0.61-6.02 3.51 18. Drills DRIL n=ONo drills of lithic manufacture, with wear indicating rotary motion (Stryd 1973:349),were recovered in the Eagle Lake Project area. Bipolar Reduction Lithic Assemblage The original report on the 1979 Eagle Lake Archaeological Project described fifty-two May 18, 2003-255 a b e g h k m n 0 p O 1 2 3 4 5cm Plate 15 Piece Esquillee (PEEQ) a. CR40:48 b. ELP 32:1428 C. E1Rw4:1678d. T84-27:4 e. T84-27:2073 f. EkSa 13: 1585g. EkSa 13:4783 h. EkSa 13:4334 l. EkSa 32: 1064j. EkSa 32:1444 k. EkSa 33:1301 L EkSa 33:3208m. EkSa 33:4323 n, ELP 19-1:79 0. ELP 19-1:1245p. ELP 44/020:228bipolar cores and wedges (Ludowicz 1980: 101-105). With specimens from later field seasons added, the number of artifacts created by bipolar flaking techniques has increased to one hundred and sixty, classified as pieces esquillees, wedges, and bipolar cores and fragments. Bipolar reduction involves striking the raw material from above while it rests on an anvil, therefore, the resulting fracture has evidence of forces coming from opposite directions. Artifacts created using bipolar techniques exhibit battering and negative bulb scars on opposite ends (Ludowicz 1980:101) and a double wedge shape (Loy and Powell 1977:58). Sanger noted evidence of bipolar flaking in the Lochnore-Nesikep assemblages, however, believing the artifacts were the detritus of lithic manufacture, he combined what others had referred to as pieces esquillees and wedges into one group of \"bipolar flaked artifacts\" (1970:84). For the Eagle Lake assemblage analysis we follow Macdonald (1968), Stryd (1973), and Magne (1985) in interpreting bipolar artifacts as implements. The large quantity of bipolar artifacts found dming the Eagle Lake Archaeological Project probably results from the use of small pebbles of fine-grained basalt and obsidian as the raw material for lithic manufacture. Both of these raw materials are usually available as small pebbles, the fine-grained basalt locally, and obsidian from Obsidian Creek, north of Anahim Lake. 19. Pieces Esquillees PEEQ n = 86The Eagle Lake Project area pieces esquillees (Plate 15) are similar to those described byStryd (1973:369, Fig 31, s-v) and Magne (1985:168). They are small, utilized artifacts with a rectanguloid shape and all four edges about equal in length (Stryd 1973; Magne 1985: 168). They exhibit paired crushed margins (usually the lateral margins) and bipolar flake scars (not extending across the entire face) where iITegular bipolar flakes were driven from both faces at May 18, 2003-256 the crushed margins (Stryd 1973 after MacDonald 1968:85-90; Magne 1985:168). Stryd's assembiage amounted to only thirteen specimens, all of them basalt. In the Eagle Lake Project area most of the eighty-six specimens recovered are basalt, except for ten obsidian pieces esquillees collected on the surface, six from the Brittany Creek site (EkSa 33) (Plate 15, m) and four at ELP Quad 19:1 (EkSa 27) (Plate 15, n-o). The Eagle Lake area specimens are somewhat smaller than those described by Stryd: he reported a maximum length of 51 mm, and illustrates one of approximately 42 mm and another of 40 mm, of the four illustrated ( out of the total of 13). In contrast, the longest Eagle Lake area specimen is 37.2 mm in length from platform to flake termination (Plate 15, a), in a sample of 86. The smallest complete piece esquillee, with a length of 16.1 mm, was recovered at the Shields site (EkSa 13) (Plate i5, f). Thirty-three pieces esquillees, all fine grained basalt, were recovered at the excavated sites: four at the Bear Lake site (EkSa 36) (Plate 15, b), one at Quiggly Holes/Bidwell Creek (ElRw 4) (Plate 15, c), eight at the Fish Trap Lake site (T84:27/ EkSb 37) (Plate 15, d-e), fourteen at the Shields site (EkSa 13) (Plate 15, f-h), and six at the Boyd site (EkSa 32) (Plate 15, i-j). The largest collection of pieces esquillees from a single site, numbering thirty-two, is a surface collection from the Brittany Creek site (EkSa 33); these are primarily fine-grained basalt (Plate 15, m), however, six obsidian specimens were also recovered (Plate 15, k-1). Only one other site yielded obsidian pieces esquillees; four were recovered at ELP Quad 19: 1 (EkSa 27) (Plate 15, n) along with ten fine-grained basalt specimens (Plate 15, o ). Surface collections at four other sites in the Eagle Lake area ranged from one to five specimens, including five from Grassland Quadrat 20/ ELP Quadrat 44 (Plate 15, p). Thirty-four of the eighty-six pieces esquillees are complete. All specimens illustrated in Plate 15 are complete except \"c\" and \"k\" and represent the size range and all sites where pieces May 18, 2003-257 b C a g I k p q O 1 2 3 4 5cm Plate 16 Bipolar Wedges and Cores Wedges (WEDG): m r d ' h n 0 s a. ELP 32:129 b. ELP 32:160 c. d. E1Rw4:1302 e. E1Rw4:1360 f. g. EkSa 13:2167 h. EkSa 13:2849 1. j. EkSa 13:4087 k. EkSa 32:5826 1. m. EkSa 33:3437 n. EkSa 33: 3808 o. Bipolar Cores (BICO) and Bipolar Core Fragments (BICOf): p. EkSa 33:2411 q. EkSa 33:3621 r. s. ELP 19-1:161 t. ELP 26 OS:2t ELP 32:1058 T84-27:352 EkSa 13:3178 Ek.Sa 32:630 l EkSa 33:4302 Ek.Sa 33:3745 /(, \"'-esquillees were recovered, except CR 1 (EjSa 11) which yielded one fragment. Table V-22 Pieces Esquillees: Metric Data # of Specimens Range IQR Mean Length 52 11.2-37.2 18.9-24.9 21.7 Width 55 11.4-33.9 15.1-21.3 18.7 Thickness 86 2.6-10.6 4.2-6.4 5.42 Weight 34 0.72-7.06 1.59-3.26 2.4 20. Bipolar Wedges WEDG n = 60Bipolar wedges exhibit two lines of bipolar reduction, perpendicular to each other,resulting in flaking and crushing on all four margins (Plate 16, a-o). These characteristics are well illustrated on a specimen from the Shields site (EkSa 13) (Plate 16, g). Materials recovered in the Eagle Lake project area include thirteen obsidian (Plate 16, b, i, m); five coarse grained basalt (Plate 16, o ); and one chert specimen (Plate 16, n) in addition to the forty-one specimens of fine-grained basalt. Cortex remains on four specimens: one each from the Bear Lake site (EkSa 36) (Plate 16,a), the Fish Trap Lake site (EkSb 37) (Plate I 6, f), and the Shields (EkSa 13) and Brittany Creek (EkSa 33) sites. Bipolar wedges have a distribution similar to the pieces esquillees: the Brittany Creek site (EkSa 33) yielded the largest number with twenty-one specimens, the Shields site (EkSa 13) had eleven, the Boyd site (EkSa 32) had nine, Bear Lake (EkSa 36) had eight, and Fish Trap Lake (EkSb 37) and Bidwell Creek site (ElRw 4) each yielded four. Three other specimens were recovered at three surface sites in the Eagle Lake Project area. The specimens from the Bear Lake (EkSa 36) (Plate 16, a-c) and Boyd (EkSa 32) (Plate 16, k-1) sites are more \"blocky\" than the thinner, wedge-shaped specimens from most other sites. Specimens from the Brittany Creek site (EkSa 33) also include many blocky wedge fragments May 18, 2003-258 of basalt and obsidian (Plate 16, o ). The finest examples of bipolar wedges were recovered at the Shields site (EkSa 13) (Plate 16, g-j) and include one of obsidian. Table V-23 Bipolar Wedges: Metric Data # of Specimens Range IQR Mean Length 32 12.1-43.1 19.7-29.1 23.8 Width 33 6.2-32.8 12.3-21.4 17.4 Thickness 60 3.3-17.5 5.9-10.0 7.5 Weight 25 0.24-18.80 1.37-2.99 3.33 2L Bipolar Cores & Fragments BICO and BICOf n = 14 Bipolar cores have cortex and, like pieces esquillees, exhibit crushing and negative flake scars on opposite ends as a result of the core being held on an anvil whiie being struck from the top (Plate 16, p-t). The core may have been inverted 180 degrees and struck again from the opposite end, however, there is no evidence of crushing along the sides of the core, and the resulting shape is a long and narrow core (Ludowicz 1980: 103). Bipolar cores may be an early stage in the production of pieces esquillees or, if rotated 90 degrees and struck again, bipolar wedges. As with the other artifacts of bipolar flaking, the largest number of cores was recovered on the surface of the Brittany Creek site (EkSa 33) (n = iO, Plate 16, p-r) including the onlyexamples of coarse-grained basalt (2) and chert ( l) (Plate 16,q) in addition to three obsidian (Plate 16, p) and four fine-grained basait specimens (Plate 16, r). No bipolar cores were recovered at the Bear Lake (EkSa 36), Shields (EkSa 13), or Boyd (EkSa 32) sites. Three specimens were recovered on the surface at other sites: ELP Quad 19: 1 (EkSa 27) yielded two bipolar cores, both obsidian (Plate 16, s), and the Fish Trap Lake site (EkSb 37) and ELP Quad 26 (EkSa 31) (Plate 16, t) each yielded one fine-grained basalt specimen. May 18, 2003-259 Table V-24 Bipolar Cores and Fragments # of Specimens Range Mean Length 13 12.9-36.7 25.2 Width 10 8.9-28.1 20.0 Thickness 14 4.8-13.0 7.8 Weight 11 0.46-10.23 4.39 Cobble-based Lithic Assemblage 22. Cortex Spall Tools SPTO n = 11 Cortex spall tools are large, generally flat flakes, derived from cobbles with water worn cortex on the dorsal surface and naturally sharp edges that may show evidence of utilization or limited retouch (Sanger 1970:88-89; Matson et al. 1980). Spall flakes, often from granular basalt and other dense igneous cobbles, may exhibit retouch to provide a haft end and a blunted scraping end, although generally only one end will be worked (Magne 1985: 168). Polish is common at the scraping end of the tool (Magne 1985: 169). Stryd noted that in the assemblage of sixty-eight retouched spalls he described, all but one exhibited unifacial dorsal retouch; the one exception was bifacially retouched (1973:366-67). He also noted that retouch is primarily at the transverse edge, but also occurs on the lateral edges, and occasionally on the entire margin except for the bulb area. The modest collection of eleven cortex spall tools from the Eagle Lake Project area include three (one from the Shields site (EkSa 13), one from ELP Quad 19: 1 (EkSa 27), and one from a grassland quadrat) with no retouch but with use-wear, similar to those described by Sanger (1970:88-89) and specimens from the Mouth of the Chilcotin (Matson et al. 1980:25). Specimens with retouch include the variations described by Stryd: six exhibit unifacial retouch ranging from limited flaking of the transverse edge (all specimens from the Brittany Creek site, EkSa 33), to much flaking of either the transverse edge ( one specimen at the Bear May 18, 2003-260 C 1 2 3 4 5 :1 Lake site (EkSa 36) (Plate 17, a) or both the transverse and lateral edges (ELP G2:2). Two cortex spall tools with bifacial flaking were also recovered (Plate 17, b-c ), including one (Plate 17, b) with bifacial retouch on all edges except at the bulb. The other bifacially flaked specimen is unique (Plate 17, c ): it is the only one of fine-grained basalt and appears to be a thin cobble, with cortex on both faces, that has been shaped by bifacial flaking to create a rounded end and incurvate sides, perhaps for hafting. The other, curved end exhibits extreme polish and the lateral edges exhibit use-wear flaking. This artifact was likely hafted as a hide scraper (Teit 1900). All but one of the eleven cortex spall tools were recovered on the smface, including one from the surface of the Bear Lake site (Plate 17, a). The specimen recovered during excavation was at the Shields site (EkSa 13). The only fine-grained basalt specimen is from a grassland quadrat (Plate 17, c ). All other specimens are either coarse-grained basalt or igneous material. Most sites yielded only one specimen, however, the Brittany Creek site (CR 92/EkSa 33) yielded four, (two each of coarse-grained basalt and igneous material) and ELP Quad 19: 1 (EkSa 27) yielded two of coarse-grained basalt (Plate 17,b). Table V-25 Cortex Spall Tools # of Specimens Range Mean Length 9 80.0-151.5 100.3 Width 8 58.1-95.7 76.3 Thickness 10 7.6-26.7 16.5 Weight 8 72.89-318.00 154.94 23. Core Tools and Fragments CORE n = 12Core tools are cobbles with flake scars but with no prepared platforms (Matson1976:131). Core tools and fragments (not illustrated) were collected only at the Brittany Creek May 18, 2003-261 a o 1 2 3 4 5 cm Plate 18 Hammerstones and Cobble Tools Hammerstone (HAST): a. EkSa 33:3192Cobble Tool (COBL):b. ELP 32:1558site (EkSa 33) and all are from the surface. Ali specimens are coarse-grained basalt except one of igneous material. Two cores exhibit evidence of retouch and may have served as choppers: Ek:Sa 33:3940 has large, coarse flakes removed bifacially to create a steep edge ( opposite the rounded cortex-covered edge) and finer bifacial retouch of the steepened edge to sharpen it. The other retouched core tool (EkSa 33:4245) exhibits unifacial retouch on one sharp, steep edge and also retains cortex cover. Table V-26 Core Tools # of Specimens Length 9 Width 9 Thickness 12 Weight 8 24. Hammers tones Range 66.5-257.6 43.9-116.9 16.9-83.7 57 .83-1320.00 (HAST) Mean 98.6 74.8 39.6 390.33 n=3Hammerstones are unshaped, round to oval cobbles with indications of use in the form ofpitting or roughening, usually confined to limited and specific areas, but some show wear all over one or both ends (Sanger 1970:89). Sanger's hammerstones vary widely in weight, ranging from 29 grams to 3,305 grams. Stryd described hammerstones with pitted ends in three forms: oval, elliptical, and triangular (1973:380-81). In the Eagle Lake Project area, only three hammerstones were recovered. Two oval shaped, igneous cobbles were recovered: one from ELP Quadrat 26 (EkSa 31) has extensive pitting on one end and one from the Bear Lake site (EkSa 36) has extensive pitting on one end and the adjacent lateral margin. The specimen from the Brittany Creek site (EkSa 33) is unusual (Plate 18, a). It is also igneous rock but has a natural wedge shape at one end. At the other end, large bifacial flakes have been removed to create a similar, but alternate, wedge shape. Both May 18, 2003-262 ends are worn from use, forming two sinuous edges. Although this might be classified as a cobble tool (see below), its shape varies from Sanger's definition of cobble choppers as it is not a rounded cobble. Table V-27 Artifact Hammerstones: Metric Data EkSa 33:3192 EkSa 36: 1720 ELP 26-3:20 Length C Width C Thickness 81.4 Y 47.6 Y 26.1 160.0 N 86.9 Y 64.9 i06.3 N 70.8 Y 47.4 25. Polished Cobble POCO Weight 173.46 1393.6 559.39 n = 1One cobble recovered from hearth Feature H at the Bear Lake site (EkSa 36) has anunusual but natural outline resembling a hatchet-shape (not illustrated). The basalt cobble is 128.0 mm long, 67.6 mm wide and 27.4 mm thick, exhibits polish on both ends and on the curved lateral margin, and weighs 332.16 grams. 26. Pecked Cobble Tool PCOB n=lOne cobble tool, recovered on the surface of the Brittany Creek site (EkSa 33, notillustrated), is unique. It is a flat, round, porous igneous cobble that is worn around the entire margin. One surface exhibits polish and is pecked at the centre. It is 137 .0 mm long, 128.6 mm wide, and 20.7 mm thick and weighs 680.50 grams and is possibly an anvil stone. 27. Cobble Tools COBL n=3Cobble tools are unifacially flaked implements (with rare specimens having bifacialflaking) based on rounded cobbles that may have served as chopping tools or as scrapers (Matson 1976:141). Flaking is generally crude, often with only three or four large flakes May 18, 2003-263 removed to form an edge and materials vary widely, although basalts are common (Sanger 1970:84-87). Two artifacts, both from the Bear Lake site (EkSa 36), are similar to Sanger's cobbie choppers (1970:84): they are based on flat, round cobbles of igneous material crudely flaked with three or four large flakes removed. One is unifacially flaked to form a sharp edge on the transverse edge with the greatest curvature. The other (Plate 18, b) is unlike those described by Sanger. It is bifacially flaked on both ends. Based on the presence of extensive use-wear on the edge opposite the flaking it may have been used for scraping. One flat, round basalt cobble was recovered from the Shields site (EkSa 13). It exhibits two large flake removals at the broader end, and has extensive use wear including pitting, striations and large, unifacial flake removals on one lateral margin. Table V-28 Cobble Tools Artifact Length C Width C Thickness Weight ELP 32-1:1467 136.3 y 97.6 y 34.6 754.90 ELP 32-1: 1558 154.0 Y 100.3 y 33.0 738.80 EkSa 13:4039 169.0 Y 107.6 y 46.9 1299.4 28. Cobble with Use-wear UCOB n = 1One flat, round, igneous cobble was recovered from the Shields site (EkSa 13) thatexhibits use-wear but no flaking (not illustrated). It is worn smooth along one margin and along one adjacent face but also exhibits pecking along the same margin, indicating possible use as a grinding stone. The specimen is 86.2 mm long, 78.2 mm wide, 38.5 mm thick and weighs 417.59 grams. 29. Large Flake Tools LFLT n = 13Large flake tools, unlike the large formed bifaces, generally exhibit unifacial flakingMay 18, 2003-264 0 1 2 3 4 5 Large Flake 13:4330 13:3962 (Plate 19). They are based on two flake shapes derived from cobbles: one is oblong, triangular and blocky (Plate 19, a-b); the other has a half-moon orulu shape and is flatter (Plate 19, c-d). Many specimens of both types have cortex that often becomes the hand-held part of the tool. All have been shaped to create a sharp working edge and a hafting area, usualiy by unifacial flaking, but occasionally by bifacial flaking (Plate 19, c). Two specimens, one from the Bear Lake site (EkSa 36) and one from the Fish Trap Lake site (EkSb 37) (not illustrated) exhibit alternate edge flaking. The flat, ulu-shaped flakes generally have unifacial flake removals to create a sharp working edge and, along the opposite edge, a dulled edge for ease of holding. Use-wear flaking is generally unifacial, but on two specimens is bifacial (Plate 19, a). All tools exhibit use-wear on the sharpened edge in the form of macroscopic polish, crushing, and small, discontinuous flake removals. Three large flake tools have a unique sinuous edge with use wear: one is a flat flake type (Plate 19, d), and two are of the blocky type (Plate 19, b ). The greatest number of specimens, six, was recovered in the Shields site (EkSa 13) excavation (Plate 19, a, c-d), including one specimen, of fine-grained basalt (Plate 19, d) with extensive use-wear (in the form of crushing and small, unifacial flake removals) on one lateral margin and on the sinuous edge. One blocky, large flake tool of granite with alternate unifacial retouch was recovered at the Bear Lake site (EkSa 36). Three blocky specimens were recovered on the surface of Grassland Quadrat 20; all are coarse-grained basalt. One exhibits bifacial flaking on all edges. The tool has use-wear on more than one edge, as if the tool was rotated, but one edge has the most extensive (and bifacial) evidence of use. Another specimen from Grassland Quadrat 20 (Plate 19, b) is bifacially shaped at the platform end and has a sinuous, naturally steep edge with unifacial use-wear flaking. May 18, 2003-265 I I C I d a b a f g h e O 1 2 3 4 5cm Plate 20 Microblade and Ground Stone Assemblages Microblades (MIBL): a. EkSa 32:2091d. EkSa 33:2766Adze flakes (ADZEf):e. CR 73:61Notched Slate (NOSL):g. CR 73:113Waisted Stones (WAST):b. EkSa 32:2941f. CR 73:109Polished Pebble (POLP):h. CR 73:100i. CR 73:83 j. CR 73:41Pebble with Ochre (OCHP):k. CR 73:38k C. EkSa 32:6133One large flake tool was also recovered at each of the following sites: at Fish Trap Lake (EkSb 37) a flat, curved specimen of fine-grained basalt; at ELP Quadrat 19: 1 (EkSa 27) a blocky specimen of coarse-grained basalt; and at ELP Quadrat 22 (EkSb 6) a blocky specimen of fine-grained basalt. Use-wear flaking and patination indicate this tool was used, discardedand reused more recently. Both usages are evident on the curved edge, but earlier use-wear flaking includes the opposite, slightly sinuous edge. Table V-29 Large Flake Tools: Metric Data # of Specimens Range Mean Length 10 71.8-131.1 101.7 Width 11 49.8-84.7 62.4 Thickness 13 15.0-44.7 23.8 Weigh 9 77.71-292.94 159.42 Microblade Assemblage 30. Micro blades MIBL n=4 Microblades are relatively thin, naiTow flakes with straight, parallel sides (and ridges) and relatively constant thickness/width ratio (Sanger 1970:60). Striking platforms have flat, unfacetted surfaces and the ventral side may be a smooth, straight line or may be crescent shaped (Sanger 1970:62). Obsidian microblades were recovered in the Eagle Lake project area at two sites: three at the Boyd site (EkSa 32) (Plate 20, a-c) and one on the smface of the Brittany Creek site (CR 92/EkSa 33) (Plate 20, d). Only one (Plate 20, a) is complete and it is 18.3 mm long. The fragmentary specimen from the Brittany Creek site is a little wider and thicker than the microblades from the Boyd site. May 18, 2003-266 t a C f O 1 2 3 4 5cm o 1 2 3cm PLATE 21 Incised and Decorated Bone (INCB) a. EkSa 36:6010 whittled b. EkSa 36:5787 grooved/incised c. EkSa 36: 1948 incised tooth d. EkSa 13:5717 decorated bone e. CR 73:46f. CR 73:42g. EkSa 36:60llah. EkSa 36:601 lb1. EkSa 36: 1902I d Table V-30 Artifact Ek:Sa 32:2091 EkSa 32:2941 EkSa 32:6133 Ek:Sa 33:2766 Microblades: Metric Data Length C Width C 18.3 Y 4.6 Y 10.3 N 4.4 Y 6.2 N 3.6 Y 8.4 N 6.0 Y 3L Microblade Cores MICO Thickness 1.5 0.9 0.4 1.9 Weight 0.12 0.05 0.01 0.11 n=O No microblade cores were recovered in the Eagle Lake Project area. Bone and Antler Assemblage 32. Incised Bone INCB n = i 1 In the E agle Lake Project area incised or decorated bone fragments and one incised tooth were recovered at three excavated sites: six fragments at the Bear Lake site (EkSa 36) (Plate 21, a-c, g-i), one fragment at the Shields site (EkSa 13) (Plate 21, d), and two fragments at ahousepit site CR 73 (EkSa 35) (Plate 21, e,f). In all, seven specimens show clear evidence of incising and one worn fragment may have been incised. The decorative patterns created by the incised lines are similar to those reported by Sanger (1970:92-95) and Stryd (1973: Fig 40, k-1) for beads and pendants. One fragment from the Bear Lake site (EkSa 36) was whittled, rather than incised (Plate 2L a), and one from the Shields site (EkSa 13) was decorated with notches (Plate 21, d). An incised tooth (Plate 21, c) was recovered at the Bear Lake site (EkSa 36). One specimen from CR 73 (EkSa 35) (Plate 21, f) is a flat rectangle (with an outline similar to the neck of a bottle, with three parallel lines carved onto the upper lip of one edge and two lines on the opposite edge). In f01m it is similar to the \"sap scraper\" (Stryd 1973:392; Pl 35), \"gaming pieces\" (Morlan 1973 :Pl 19), cut antler rectangles (Campbell 1968 :40) and incised bone objects (Shinkwin 1979: 115, Pl 31) that form part of the \"rather distinctive Kavik artifact May 18, 2003-267 d b .\u00EF\u00BF\u00BD\u00C2\u00B7\u00EF\u00BF\u00BD :,. : \u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00EF\u00BF\u00BD, i J 0 -Y-1 e. j :\u00E2\u0080\u00A2.\u00E2\u0080\u00A2 I I I \u00C2\u00B7,\r .. , .'' Figure iB::. Illustrations of decorated artifacts from Bear Lake and Shields sites. a - h: Shields site i - k: Bear Lake site type\" (Campbell 1968:40). Stryd's inference of sap scrapers is based on morphological similarity to ethnographic specimens (Stryd 1973:392; Teit 1909a, Fig. 235c; 1909b: Fig. 275). One other bone specimen (Plate 21, e) recovered from CR 73 (EkSa 35), is small and fire\u00C2\u00ADblackened, 15.0 mm long, 6.0 mm wide, and 3.0 mm thick. It has one deep, longitudinal groove incised on one face. One incised bone specimen from the Bear Lake site (EkSa 36, Plate 21, i) was recovered in seven fragments from the historic component and has been reassembled. All fragments are soot-blackened on some areas. The re-assembled specimen (Plate 21, i, Figure Y-1, k) is both perforated and incised and measures 96.5 mm long, the width varies from 1 1.5 to 14.0 mm, and thickness is 3.0 mm. It has a circular perforation at the nanower end and one face has diagonal lines intersecting to create a series of \"diamond shapes\" that are filled-in with incised lines running parallel to the length of the bone. Figure V-1, k illustrates this fully reconstructed with front, back, and side views. Plate 21,i shows the assembled front view. FIGURE Figure V-1. Illustrations of decorated artifacts from Bear Lake and Shields sites. The whittled specimen (Plate 21, a) of bone from the Bear Lake site (EkSa 36) is a small, rounded fragment (15.0 mm long, 5.0 mm wide, and 4.0 mm thick) that is whittled on one end to taper but not to a point. The incised tooth from the Bear Lake site, EkSa 36, (Plate 21, c; Figure V-1, j) is hollow and incised with one curved line with two diagonal lines within its curve. The tooth is 14.0 mm long, 6.0 mm wide, and 3.0 mm thick and, as reported in the faunal analysis, is either a porcupine or beaver incisor. 33. Bone Fish Spear Points (Leister Prongs) BPNT n=6 May 18, 2003-268 a PLATE 22 Bone Fish Spear Points (Leister (BPNT) a. EkSa 13:5346 (in 2 pieces)b-L EkSa 13:2161, 2141, 2144 (glued in 1and EkSa 13:2143, 2149, 2155, 2148(glued in 1 piece) b-2. Ek:Sa 13:2153, 2158, 2159, 2156, 2145 (glued in l piece) and EkSa 13:2146 b e d C O 1 2 3 4 5cm c. EkSa 13:2142, 2152, 2151 (glued in 1 piece) and EkSa 13:2162, 2154 (glued in 1 piece)d. EkSa 13:2166, 2157, 1563 (glued in 1 piece)e. EkSa 13:3605Fragments of bone leister prongs were recovered at only one site, the Shields site (EkSa 13), and represent at least five points (Plate 22, a-e; Figure V-1, a-h). Leisters are fish spears with two flexible, barbed side prongs designed to pierce and retrieve fish f ro m above and were used from platforms, holes in the ice, or from canoes, especiaily at night aided by torches (Rostiund 1952:105-112, 293-295; Teit 1900:252). Specimens from the Shields site are unilaterally barbed and many show evidence of burning. One unique point (Plate 22, a) exhibits two unilateral barbs and 4 notches (not unilateral) and has a design of incised dot<; spiralling around the shaft. Three of the other specimens exhibit extended barbs (with the larger part of the barb extending beyond the lateral edge of the point) (Plate 22, b-d) and one has two enclosed barbs (with the larger part of the barb within the lateral edge of the point) (Plate 22, e). One specimen with nine extended barbs (Plate 22, b ), recovered in fragments, has been re-assembled and, although there may be fragments missing, the length of all four sections is 316.5 mm. The tip is straight and exhibits facets on six sides coming to a point. Another point with very similar attributes was also recovered. The proximal end tapers below three small notches, one pointing up, one down and one perpendicular to the lateral edge of the shaft. The points are similar to the unilaterally barbed points reported by Stryd (1973: Fig. 41, a-d) and, like the point<; reported by Morlan (1973: 278-287) at the Klo-kut site, they are variable in form and size. They do not have the \"lenticular\" barbs repmted by Campbell (1968) for the Kavik site. Table V-31 Bone Points: Metric Data Artifact# Plate # Length C Width C Thickness Weight EkSa 13: 5346 22-a 172.0 N 7.0 y 6.0- 5.0 6.79 EkSa 13: 2161 et al, 2143 et al 22-b 316.5 N 5.0 -12.0 Y 5.0 - 6.5 13.40 EkSa 13: 2151 et al, 2154,2150 22-c 173.3 N 5.8 -10.0 Y 4.5 to 6.0 5.36 EkSa 13:1563 et al 22-d 83.0 N 9.0 y 5.0 2.64 EkSa 13:3605 22-e 51.5 N 8.0 y ,., \u00EF\u00BF\u00BD .).) 1.51 May 18, 2003-269 C d O 1 2 3 4 5 cmPLATE 23 Bone and Antler: Harpoon Points, Awls, and Beamers Harpoon Points (TPNT): a. EkSa 13:2274b. EkSa 13:5747c. EkSa 13:3203Bone Awl (BAWLT):d. EkSa 13:5713Scapula Beamer (BEAM):e. EkSa 13:3260point composite harpoon valve frag. Haft end of antler harpoon? With bilateral line guards. bone awl scapula beamer The six bone point tips, including the two mentioned above, all vary in method of manufacture. One tip (Plate 22, c) has five facets, creating a flat-sided tip rather than a rounded one. The tip of the specimen with enclosed barbs (Plate 22, e) is bifacially ground to produce a flattened point with rounded tip. Two point fragments were also recovered. 34. Harpoon points and fragments TPNT n=3Three fragments of bone or antler harpoons were recovered at the Shields site (EkSa 13),including two composite toggling channelled valves (Plate 23, a-b ). EkSa 13:5347 is the best example (Plate 23, b). EkSa 13:2274 (Plate 23, a) is only a likely example, that may be the broken tip of some other object. Plate 23, c (EkSa 13:3203) is the base of a harpoon with bilateral line guards, such as found on \"Marpole-Style\" unilaterally barbed harpoons (Matson and Coupland 1995:205). Composite toggling harpoons were fastened with a line around the middle of the valve (below the bulge in Plate 23, b) and \"toggled\" or turned when the fish pulled the harpoon off the shaft. Plate 23, c would not \"toggle\". The missing upper part of Plate 23, c would have barbs and the line would have been tied between them and the line guards which are the two protuberances visible. Harpoons used for freshwater fishing had heads that detached from the harpoon shafts but were attached by a line so that fish could be retrieved (Rostlund 1952:105). The cun-ent use of \"gaffs\" at Henry's Crossing continues this technology for salmon fishing today (Burnard-Hogarth, this volume). Table V-32 Harpoon points and Fragments: Metric Data Artifact Number Plate # Length C Width C Thickness EkSa 13: 2274 23-a 6Ll N 13.2 Y 5.8 EkSa 13:5347 23-b 45.1 N 8.5 Y 5.0 EkSa 13: 3203 23-c 96.9 N 27 .2 Y 8.9 May 18, 2003-270 Weight 3.42 1.43 14.21 35. Bone Awl BAWL n-1 LOne bone awl with broken tip and use-wear polish but no modification was recovered at the Shields site (EkSa 13) (Plate 23-d). It is 160 mm long, 21.5 mm at the widest part, 7.1 mm thick, and weighs 12.32 grams. 36. Scapula Beamer BEAM n=3Three scapula beamers were recovered at the Shields site (EkSa 13) (Plate 23-e) that aresimilar to the unornamented beamers reported by Morlan at the Klo-kut site (Morlan 1973:305-307) which functioned as hide scrapers. Stryd (1973:391-2) also repo1ts six such items madefrom deer scapulas. All specimens from the Shields site are very worn, broken at the working edge and distal end, and two are broken along the lateral margin. Lengths (incomplete) range from 16.0 to 20.5 mm, widths (incomplete) range from 26.0 to 44.4 mm, and thickness ranges from 3.8 to 17 .5 mm. Paull (1984) tested EkSa 13:3260 (Plate 23, e) at four loci for blood and fat and at one of those four for fat and lignin. All of the loci looked like blood specks and three were positive for blood and fat. The fourth loci was negative for these and starch but positive for lignin. Whether the blood and starch are the result of the function of the tool or because the object is bone is unclear. CR 73 / EkSa 35 Assemblage Unique Items The following artifact types were recovered only at EkSa 35 (CR 73 ), a housepit site near Brittany Creek that has been truncated by the Chilko River. The site is about three hundred and fifty years old (SFU 15: 360+/-80 BP). In addition to the waisted stones, adze flakes, notched May 18, 2003-271 slate, polished pebbles and pebble with ochre described below (Plate 20, e-k), the excavation also yielded two specimens of incised bone, one Kavik point. one side-notched point, one utilized flake of quartz, and one small, bifacially retouched flake of obsidian (previously described). 37. Waisted Stone WAST n=3Three flat, oval cobbles exhibit bifacial flaking at the mid-section to create a waist andhave the appearance of crnshing but not polish at the waist. Two complete specimens (Plate 20, i-j) and one, smaller size, fragment were recovered and it is thought these may be net-sinkers.On one granite specimen (Plate 20, i) large, coarse, flakes are removed from both edges to create the waist; the flake removals are unifaciai on one edge and bifacial on the other. The other complete specimen (Plate 20, j) is fine-grained basalt and exhibits bifacial flaking of both edges to create the waist. Table V -33 Artifact CR 73:41 CR 73:83 CR 73:108 Waisted Stones: Metric Data Plate Length C Width C 20-j 62.7 Y 43.3 Y 20-i 87.9 Y 41.6 Y 21.2 N 20.5 N 38. Adze Flake ADZEf Thickness 13.5 16.0 8.7 Weight 60.30 96.56 5.16 n=2At CR 73 (EkSa 35) two fine-grained green.stone flakes with ground and polished dorsalsurfaces were recovered during the excavation and have been identified as adze flakes (Plate 20, e-f). In his report on the Lochnore-Nesikep area, Sanger reported fragments of nephrite celtswith grinding, bevelling and grooving (1970:89). May 18, 2003-272 Table V-34 Artifact CR 73:61 CR 73:109 Adze Flakes: Metric Data Length C Width C 26.1 N 12.0 Y 13.7 Y 9.0 Y 39. Polished Pebble POLP Thickness 2.5 2.1 Weight 0.86 0.14 n=2Two small, highly polished pebbles of fine green stone were recovered during theexcavation of the CR 73 (EkSa 35) house pit. One is ovoid in shape (Plate 20, h); the other is triangular. Table V-35 Artifact CR 73:37 CR 73:100 Polished Pebbles: Metric Data Length C Width C Thickness 30.1 Y i3.5 Y 6.8 19.0 Y 14.8 Y 7.0 40. Pebble with Ochre OCHP Weight 3.42 3.6i n=lOne rounded igneous pebble with ochre on one end was recovered in the excavation ofCR 73 (EkSa 35) (Plate 20, k). The pebble is 35.9 mm long, 26.5 mm wide, 14.8 mm thick and weighs 21.02 grams. 4L Notched/Incised Slate NOSL n=l One fragment of slate was recovered in two pieces on the surface of CR 73 (EkSa 35) (Plate 20, g). It exhibits a continuous zig-zag pattern incised into the edges. 42. Historic Artifact Assemblage Historic Artifacts: Bear Lake Site (EkSa 36) HIST The following details on the historic artifacts from the Bear Lake site (EkSa 36) are May 18, 2003-273 a b C {f 11 0 h 0 I k m \u00E2\u0080\u00A2 f \u00E2\u0080\u00A2 q s O 1 2 3 4 5cm PLATE 24 Historic Artifact Assemblage (HIST) a. EkSa 36: 1895 religious medalButtons:b. EkSa 36: 1890/2779 white glassc. EkSa 36: 1960 metal, Scovill eagled. Ek.Sa 36:1918, 1923, 1924 metal buttonbacke. EkSa 36: 1944 copper face, iron backBeads:f. Ek.Sa 36:1961 & 3008seed beadsg. Ek.Sa 36:2778 blue wire wound, largeh. EkSa 36:2742 blue wire wound,medium 1. EkSa 36:3964 blue wire wound, mediumj. EkSa 36:1790 redk. ELP 32-1:159 blue hexagonal beadd e 0 0 ' n r 0 u Miscellaneous Metal: L EkSa 36:397 5 can key and strip m. EkSa 36: 1893 metal cut outn. EkSa 36: 1994 metal cut outo. Ek.Sa 36:1932 metal cut outGun Related:p. EkSa 36:3006 gun tlint fragq. EkSa 36: 1885, 1886, 1887 3 dropsof lead sprue Horse Related: r. EkSa 36:583 7 round brass with stars. EkSa 36:1902 & 1942buckle, 2 partsOther:t. EkSa 36: 1897 threaded gray metal,unidentified u. CR 28:40 rolled copper tube largely taken from the report prepared by Rod Heitzmann (2001). All artifacts date to the mid\u00C2\u00ADnineteenth century (the historic lodge has bark 1ing tree-ring dates of AD 1851 and AD 1877) and relate to clothing and personal items, cooking and food, guns, horse tack, and hardware. Most items exhibit considerable wear or damage. Clothing and Personal Artifacts n = 13 Clothing and personal items consist of one religious medal, five buttons, and seven beads. The religious medal is a round, thin, silver-plated medal 15.4 mm in diameter with a broken cast metal hoop for attachment at one end (Plate 24, a). It appears that when the hoop broke, a small hole was drilled close to the rim at the opposite edge. Conosion has effaced the image and the inscription around the margin, however, the image appears to represent Christ with an upraised hand across the chest (The Sacred Heart symbol). The five buttons represent four types: two are white glass, typically used on shirt-; or underwear (Plate 24, b ): one three-part metal button has a cast eagle on the front and a single loop on the back for attachment, a style commonly used for coats (Plate 24, c); one metal button back may be from a cloth-covered button likely used for a coat (Plate 24, d); and one flat, two\u00C2\u00ADpart button with copper face and iron back was common on pants and dresses (Plate 24, e). The three-part metal button (Plate 24, c) consists of a front, back and a single loop. The obverse of the front is cast with an eagle, facing left; the reverse is impressed \"SCOVILLS & ... \", identifying this button as having been produced after 1840 when the Scovill Manufacturing Company of Waterbury, Connecticut became known as Scovills & Company. Buttons with this label were probably made between 1840 and 1850. The seven glass beads include two white drawn seed beads (Plate 24, f) usually used on Mav 18, 2003-27 4 -' moccasins or c lothing, and five larger beads including three dark blue wire wound beads: one large (11.1 mm in diameter) and two medium (7.4 and 7.9 mm in diameter) (Plate 24, g-i); one red, wire wound bead of medium size (7.1 mm diameter) (Plate 24, j); and one dark blue, drawn, faceted, hexagonal bead (5.9 mm diameter) (Plate 24, k). Cooking and Food Related Artifacts n = 63 Most of the artifacts in this category are fragmentary. Cooking implement fragments include parts of a minimum of four thin walled, tinned iron vessels. Fragments of two tinned iron pots were recovered: one pot had an outward rolled rim and handle lug, the other had an inward rolled rim. The crimped 1im for a container with slip-on lid was also recovered. Other kitchen related items include a can key and strip from a sealed food container (Plate 24, 1), 43 strips of metal from tin cans, cut into varying lengths and widths, possibly for other uses such as tinkling cones or repairing other metal objects (Plate 24, m-o ), and several pieces of a wider, circular metal band, possibly used to secure stove pipes. Gun Related Artifacts n=4 One fragment of a gunflint (Plate 24, p) is likely from a British source dating from the nineteenth century. Three droplets of lead sprue (Plate 24. q) are possibly the result of the manufacture of lead balls or shot. One unidentifiable gun part, consisting of a threaded base of grey metal was also recovered. Horse Tack and Related Artifacts n=4 Two artifacts may be decorations from horse tack or saddles: one is a fragment of an May 18, 2003-275 oval-shaped leather cut-out with stitching around the outer rim and two large holes on opposite sides near the rim, likely for the attachment of leather tassels or strips. The other is a round, dome-shaped brass decorative piece with an etched star on the face (Plate 24, r, Figure V-1, i). It has folded-over tangs and small holes along the lower edge for attachment to leather saddles or harness. A third artifact, of rectangular cross-section grey metal, was recovered in two pieces and is probably a buckle from a horse harness or pack (Plate 24, s). Hardware Related Artifacts n=4 One common machine-made nail with square cross-section and length of 32 mm was recovered (not iliustrated). These nails were commonly used in the mid to late nineteenth century as wood fasteners. Three segments of chain links were recovered: two are from heavy iron chain link and one is from a light gauge iron chain link. Unidentified Artifacts n=l One unidentified artifact (Plate 24, t) is the base of a round, threaded grey metal object with a diameter of 22 mm. It has a circular, 14.2 mm diameter hole in the base and the other end has been cut. Its function is unknown. 43. Historic Artifacts: Other Sites HISTOne rolled copper tube (Plate 24, u) was recovered at CR 28 (EkSa 98). Made from aflattened piece of copper coiled to form a tube, it measures 61 .7 mm in length and 8.9 mm in maximum diameter. The method of manufacture is not like two \"rolled copper tubes\" from sites in southwest Yukon (Workman 1978: 347; Pl 15) or one copper tube, identified as possibly May 18, 2003-276 a bead, from the Dixthada site (Shinkwin i 979: 104, Pl 28), all of which have long sides meeting and hammered together. However, it may have functioned similarly as a bead or tinkler for attachment to clothing (Workman 1978:347). Mav 18. 2003-277 ., ' -sz=-c 440 .. \"' ... >, C _g..... u .\u00EF\u00BF\u00BD \"'0 ;_ I 80 40 0 40 80 120 160 200 240 280 320 360 400 440 AD 19SO AD ISOO Calendar yr IP Figure 'ii!. Relationship between radiocarbon dates and actual dates. From Stuiver (1978). Radiocarbon and Dendrochronological Dating Dating of archaeological materials at Eagle Lake was undertaken with both radiocarbon and dendrochonological methods. The archaeoiogicai problem of determining the time of Athapaskan arrival at the Eagle Lake locality is complicated by the irregularity of radiocarbon dates for the last 500 years. This problem brought to our attention by Stuiver (1978), who has analyzed tree-ring samples from several parts of North America, including Vancouver Island. Stuiver's study shows, for example, that a radiocarbon date of 160 years BP corresponds to calendar years of 0 BP, 150 BP, 220 BP, and 280 BP (see Figure V-2). A calendar date of 240 BP may result in radiocarbon ages of 80 to 100 BP, while a calendar date of 160 BP can be radiocarbon dated at 200 to 220 BP, thus reversing the actual sequence of dates. FIGURE Figure V-2. Relationship between radiocarbon dates and actual dates. From Stuiver (1978). Dendrochronology was proposed as a feasible solution to these problems for several reasons. First, Marion Parker had dendrochronologically dated samples from the Mouth of the Chilcotin, Pavilion Lake, Potato Mountains, and with a single Eagle Lake sample. Secondly, there was good reason to suspect that a living tree chronology May 18, 2003-278 extending back 500 years could be achieved with some of the large Douglas Fir stumps in the Eagle Lake region. Furthermore, once a master chronology is established, future dendrochronological analysis can be reiatively inexpensive, and thus of great value to future research on the Interior Plateau. So as part of the project proposals (Matson 1978, 1982), to the Social Sciences and Humanities Research Council of Canada, funding was obtained to develop a local tree-ring chronology and to date any archaeological samples obtained. This endeavor was successful and the dendrochronology and dating is reported in detail in Parker's appendix. Dating the Athapaskan Entrance; the dating of Bear Lake, Boyd and Shields sites Standard radiocarbon samples were submitted to the Department of Chemical Engineering of Washington State University, Beta Analytic, and Brock University to provide comparative data for the dendrochronological work, and to provide ages for components expected to date beyond the capability of the dendrochronology. The Bear Lake site (EkSa 36) contained abundant charcoal and six charcoal samples were submitted for radiocarbon assessment. The Boyd (EkSa 32) and Shields (EkSa 13) sites contained very little charcoal, and as a result only one sample was sent from the Boyd site and three (one, a bone sample) from the Shields site were submitted. The results of the Bear Lake, Boyd and Shield Site radiocarbon analyses are presented in Table V -36. May 18, 2003-279 Sam le # Site Date o Modern C14 Probable A e WSU # Feature La er1 EkSa 36 106% <150 BP, pre-bomb 2 EkSa 36 295 + /- 75 BP AD 1545 or AD 1760 (2901) EU54 L ayer B3 (2902) Fea. D 3 EkSa 36 4 EkSa 36 5 EkSa 32 6 EkSa 13 7 EkSa 36 8 Ek.Sa 36 104.3% Post <150 BP pre-bomb (2903) EU28 Layer B3 113 % AD 1958,post-bomb (2904) Fea G 505 +/-70 BP AD 1410 (2905) HP 1 EU 2 46- 49 cm 107% <150 BP, pre-bomb (2906) HP 1 EU 10 68-72 cm Layer B3a 415 + I - 115 BP (BGS 2263) Fea. E (Unit 10) 210 +/-40 BP (Beta 148106) Fea. I (Unit 54) 9 EkSa 13 750 + / - 90 BP 10 EkSa 13 1650 + /- 85 BP (WSU 3379) HP 2 U nit 5 (1985) Layer Bl a (Bone) (BGS 2264) HP 1, Unit 2 Layer B3, 41-45 cm Table V-36. Radiocarbon age estimates, major excavations. The radiocarbon results obtained in 1984 from Washington State were somewhat disappointing, but not entirely unexpected. It is not surprising that samples 1, 3, and 4 date to within the last 150 years, since these charcoal samples are from the historic lodge structure at the Bear Lake site. Sample 3 is from a charcoal sample within the historic lodge, thought to be part of the structure, Sample 4 is from above the hearth, Feature G. Sample 1 is from Layer B3 which was interpreted as the historic deposit above Feature I, apparently correctly. Sample 2, which was the sole charcoal sample that was obtained from the May 18, 2003-280 550 500 450 400 350 300 250 200 150 1 Fea.11 BETA 148106 Fea.D WSU2902 Fea,E BG 52263 Averaged Date 550 500 450 400 350 300 250 200 150 Radiocarbon Years BP roasting pit (Fea. D) at the Bear Lake site, yielded a date of 295 + / - 75 BP (WSU 2902), which is useful since it probabiy dates the earlier occupation at the site. The Fea. D date translates to AD 1546 or AD 1760, according to the calibration curve available in 1984, and we favoured the AD 1760 date, and suggested that this was the age of the lithic component at the site in 1984. Since then we have obtained two more radiocarbon dates, one from Feature E, the outdoor hearth ( 415+ / - 115 BP, BGS 2263), and an AMS site from Flotation sample 18 from Feature I (210 + / - 40 BP, Beta 148106), the hearth of the inferred prehistoric iodge. This latter sample should be a direct date of the inferred prehistoric lodge. The flotation sample had numerous very small basalt flakes also present indicating its prehistoric nature. The sample was carefully picked through and the largest piece of charcoal present (0.03 g) was sent off for dating. The three dates of the prehistoric Bear Lake component are shown plotted with one sigma in Figure V-3. FIGURE Figure V-3. Bear Lake Prehistoric Radiocarbon Oates. Note that at one sigma we have a continuous distribution, but that the earliest and latest dates do not overlap at one sigma but do at two. Given that we infer a single prehistoric occupation, and the fact that these three dates come from three very separate features and were run by three different laboratories, the relatively consistency of the dates supports the inference of a single prehistoric occupation. Remembering that the wood used for fires, which is true for all three samples, would be dead wood, much of it could be recently dead, but Feature I could include heart wood that might be 100 years May 18, 2003-281 340 320 300 280 260 a.. 240 CD Q.) 220 C: 200 0 180 0 160 140 a: 120 100 80 60 1550 1600 1650 1700 1750 1800 1850 1900 1950 Calendar Years AD \u00EF\u00BF\u00BDA\u00EF\u00BF\u00BD\O LV\\u00EF\u00BF\u00BD\u00EF\u00BF\u00BDor'-) E \/ (. \u00EF\u00BF\u00BD L IP hJ IA \"-(\_ \u00EF\u00BF\u00BD \,t. f.\ \u00EF\u00BF\u00BD '-old (if downed trees were used as firewood in winter dwellings) when the cultural event, the fire, occurred. Since the Feature I date is the most recent, this is not the case, but the amount of variation among the three dates may well be the result of this factor. On the face of it, this occupation appears to be between 200 and 400 radiocarbon years ago. Since we believe the dates are dating the same event, we can average them, using the procedures described by Berry (1980). This procedure is the standard one used when averaging sample estimates with different dispersions where one weights the date inversely to the amount of dispersion, although this is not always clear from the radiocarbon literature. Since the AMS date has the least dispersion, it is weighted the heaviest and we end up with an average of 245 + /- 34 BP (Figure V-3). This is our best estimate for the prehistoric occupation of the Bear Lake site in radiocarbon years. FIGURE Figure V-4. Averaged radiocarbon date for the Bear Lake Prehistoric component. The question is what does this mean in terms of calendrical years ago, since the radiocarbon curve is so irregular in the last 500 years? Figure V-4 is a plot of radiocarbon years against calendar years using Stuiver et aL 1998. This figure shows that the averaged radiocarbon date of 211-279 BP calibrates to AD 1645-1660, a narrow range of only 15 years, and with no other intercepts at one sigma! Thus our best estimate of the prehistoric occupation of the Bear Lake site is AD 1645-1660, and this gives a minimal date for the first Chilcotin occupation of the Eagle Lake area. May 18, 2003-282 Given the various possibilities resulting from the use of old wood and the statistical probabillities involved in dating, this elegantly precise date should not be given too much weight. The more reliable result is that we have three dates from three different features all indicating the same general date for this occupation of 200-400 radiocarbon years ago, or AD 1550-1750. Note that we rejected no dates to obtain the averaged date, all the dates are consistent for that time range, and if we are correct that there is only a single prehistoric occupation--as is suggested by the consistency of the dates--averaging is an appropriate technique. A more formal procedure to see if it is appropriate to treat all dates as estimates of a single event is to use the F-test. The F-test is usually thought of as an analysis of variance procedure which one uses to test if the observed differences between samples (or sampling strata) are statisticaily significant. In this situation one is in effect using the F-test in reverse, to see if the observed sample parameters (mean and variance) (read individual radiocarbon dates) are likely to occur through repeated sampling of the same population (read date the same event) where the number of different samples is 3 or more. Given that the information associated with the radiocarbon dates is differentthan is usually associated with samples, there is more than one procedure to calculate F\u00C2\u00ADtests from radiocarbon dates; we have chosen again to follow Berry (1980) in this case using Long and Rippetreau's (1974) procedure. There are also normal theory assumptions involved which apparently are approximated, given that radiocarbon date distributions are said to be Poisson distributed, along with the assumption that the variances are equal (which is not likely in this case). The result of this procedure for the three Bear Lake site dates is an F-test ratio of 3.106 with 2 and infinite degrees of May 18, 2003-283 freedom, compared with the 0.05 ratio of 3.00 and the 0.01 ratio of 4.60 for those degrees of freedom. Thus, if one believes that the assumptions of this test are met and that the 0.05 level is appropriate, there is a formal statistical reason to reject the assumption that these three dates are of the same event. Given that we have no other information suggesting that there are two prehistoric occupations at this site, and the assumptions involved, along with the observation that the 0.05 level is just met, we believe otherwise and therefore averaging is appropriate which results in the above radiocarbon date of 245 +/- 34 BP or AD 1645-1660 when calibrated. Having developed a minimum date for the initiation of the Athapaskan occupation of Eagle Lake and thus of the Eagle Lake Phase (given that the ethnic identification is correct), what is the latest date for the PPT occupation? The date of 505 + I - 70 BP (WSU 2905) (which calibrates to AD 1425, 1395-1445 according to Stuiver etal. 1998) for sample 105 from the Boyd site, Housepit 1 (EkSa 32) is close to what we would expect for the terminal age of pre-Athapaskan occupation of the Eagle Lake region, but does not fit very well with the presence of medium-sized stemmed and corner-notched points and an absence of side-notched points, as well as microblades. Further analysis, though, reviewed in the artifact description section, showed that the corner-notched point is a style found most abundantly in Kam loops Horizon components in Lillooet. Thus, despite the lack of Kamloops sidenotched points in our modest assemblage from the Boyd site, we believe that this is a valid date and analyses reported later support the \"late PPT\" nature of this assemblage. There may well be earlier material mixed in as suggested by the three microblades and the stemmed point recovered from the surface (Plate 2,m). Thus our best point estimate (uncorrected) for May 18, 2003-284 the termination of the PPT occupation is the AD 1425 date, about 250 years before the Bear Lake site date for the prehistoric Athapaskan component. As discussed in the review of the excavations at this site, this Boyd site sample is from a large, decayed, burned piece of wood that is interpreted as part of the roof structure that fell into the pit after this structure was abandoned and burned. Since there is no floor noted above this sample, and the one clear floor was immediately below it, this should date the last house constructed in this pit. Unlike the Bear Lake prehistoric radiocarbon dates, this is probably construction wood, and thus the radiocarbon date should predate the construction and burning by some time. Given, though, that this beam is not really very large in diameter, and probably from short\u00C2\u00ADlived lodgepole pine, only a few decades should be subtracted from this date, given a date of perhaps AD 1475. This is thus our best \"corrected\" estimate for termination of the PPT at Eagle Lake and thus a maximum date for the initiation of the Eagle Lake Phase. Our \"best\" sample for the termination of the PPT pre-Athapaskan occupation at the Shields site (EkSa 13) yielded a date of <150 BP (WSU 2906). We presume that either rodent disturbance or a burned root explains this anomaly. As reported in the section describing the excavation, Layer B3a appeared to be a floor layer, or material on top of the floor, so we thought this sample would date the best floor present at this site, and possibly the last construction of a pithouse there. Although this layer begins at around 60 cm below unit datum (Figure IV-22), this datum is high on the side of the rim of the housepit, while the sample is from the southeast corner of the unit, out in the middle of the house pit (Figure IV-21 ). The ground surface is 30-40 cm above the sample making Mav 18, 2003-285 .I disturbance or rodent disturbance more likely than the depth below unit datum might indicate. There are several layers above this layer, that may also be floors, so this would not necessarily date the last house constructed there. The other two dates of 750 + / - 90 BP (WSU 3379) and 1650 + / - 85 BP (BGS 2264) are stratigraphically consistent and both indicate PPT occupations, the first of mid\u00C2\u00ADKamloops Horizon and the second of mid-Plateau Horizon. The difference in the dates indicates the presence of more than one occupation which agrees with the field interpretation of stratigraphy. The 1650+ / 85 BP date which calibrates to AD 412, AD 260-532 (Stuiver et al. 1998) is from a charcoal sample also from Layer B3 in Housepit 1,not all that f ar from WSU 2906 and the layer immediateiy above Layer 3a (Figure IV-22) which was run in 200L WSU 3379 is a date based on 253 g of bone from Layer B1a from Unit 5 in Housepit 2 run in 1985. It will be remembered from the excavation description that we were not certain that the excavated deposits in Housepit 2 was actually from a pithouse, and Layer Bla had a large amount of burned bone. This layer begins between 10 and 15 cm below surface and has a maximum thickness of 12 cm. In any event this date should be of the last main use of this depression, whether as a pit house or not. It may be that this protected area was used as an outdoor work area when not all the housepits at the Shields site had houses erected in them. Neither prehistoric date from the Shields site, though, is consistent with the age Richards and Rousseau (1987:44) argue for the multi-side notched point and which was found there, albeit during wall scrapping. That the multi-sidenotched point generally dates to the later part of the Kamloops Horizon is clear, but we are unaware of any convincing evidence for the Richards and Rousseau 400 BP limit. On the other hand, May 18, 2003-286 we don't have a date that we think indicates the most recent house floor of the Shields site, either. The calibrated date for 750 BP is AD 1265 (one sigma range of AD 1215-1375, using Stuiver et al. 1998) if that is to be preferred to the Boyd site date, still well within the Kamloops culture which is thought to begin at 1200 BP or AD 750 (Richards and Rousseau 1987). Two Kamloops side-notched points were recovered during excavations at Shields site. In summary, the radiocarbon dating of our core problem is in accord with other archaeological evidence that the Eagle Lake locality was occupied well into the Karnloops Horizon of the PPT, with the most recent point date estimate being AD 1425, for the death of construction wood used at the Boyd site. Taken into account that this is construction wood, and that the fire and abandonment of this house must have been after this, results in our estimate of A.D. 1475. All three radiocarbon dates from the Bear Lake site prehistoric component are consistent in showing that all three features had wood which died between 200 and 400 radiocarbon years ago. When averaged and calibrated, a date of A.D. 1645-1660 results, giving a minimum date for the Athapaskan entrance into Eagle Lake. The gap of about 200-250 years between the two occupations is modest given the following factors. First, there is no claim that we have dated the most recent PPT occupation--our \"best\" charcoal sample for this purpose turned out to be \"recent\". Neither of the other two dated PPT samples from the Shields sites were interpreted as dating the last occupation. The single Boyd site date, though, can be credibly interpreted as dating the last occupation there. Similarly, there is no claim that the best prehistoric Athapaskan site identified during the survey, the Bear Lake site, necessarily is the earliest such occupation. There are aiso, always, the various May 18, 2003-287 uncertainties associated with radiocarbon methods. Finally, it is to be expected that the PPT would not leave one year and be replaced by the Chilcotin the next. Instead one would expect the PPT to leave, the Chilko River e nvirons to become a neutral area, used by both, then mostly by Athapaskan, and finally, winter habitations such as the Bear Lake site would be located in it. Would 100 years for this process be a reasonable guess? So some sort of gap would be expected, although we expect further research will reduce the current 200- 250 years. Dendrochronological Dates. Tree-Ring dates are to be preferred over radiocarbon dates, when they are available. Although like radiocarbon dates, they essentially date a biological event, the death of a tree, they are far more precise. Further, the statistical probabilities involved with radiocarbon dates are not present, and the last extent ring is given the date rather than the average with radiocarbon dates. Finally, dendrochronological dates are calendric, and the calibration step is unnecessary. These features, in combination with the uncertainties associated with the last 500 years of the radiocarbon record, led to the investment to develop this procedure for the southern interior. The dendrochronological program was very successful in providing firm dates for the historic occupation at the Bear Lake site and is discussed in detaii in the separate report provided by the late Marion Parker in Appendix 1. About 100 samples of charcoal were removed from the Bear Lake site, and 11 of these provided dendrochronological dates. The living tree chronologies were developed using X-ray May 18, 2003-288 densitometer techniques and tree-ring widths. The archaeological samples were dated using X-ray densitometer techniques but simpler ring-width measurements also proved to be quite adequate for the task Table V-37 summarizes the results obtained on the 11 archaeological samples and on an additional sample taken from the decaying lodge structure near the Chilko River at Lingfield Creek that was mentioned in the ethnographic section above. Sample # Site 11 EkSa 36 29 EkSa 36 30 EkSa 36 31 EkSa 36 3 3 EkSa 36 35 EkSa 36 54 EkSa 36 72 EkSa 36 77 EkSa 36 93 EkSa 36 96 EkSa 36 none Lingfield Creek Lodge Date 1822 +/-p -1873 B 1817 +/-p -1860 v inc. 1832 p -1877 B 1848 p -1877 r 1822 + I -p -1877 rB 1832 + / -p -1877 r 1823 inc -1851 B inc. 1823 + /-p -1851 VY1830 inc -1877 +/-vv 1848 inc -1877 rB 1813 inc -1858 + /-vv 1808 p -1890 vv inc. Feature Fea. BFea. H Fea. H Fea. H Fea. H Fea. H Fea. B Fea. B Fea. B Fea. B Fea. B Table V-37. Summary of dendrochronologicai analyses. (p=pith, B=Bark present, r=outermost ring is continuous, v=judgement that date is close to cutting date, vv=no way of estimating how close the cutting date is) The most reliable dates for the death of the tree are those with 'B', indicating the presence of bark on the sample (see Appendix I for full explanations of symbois), and r being the next most, indicating the outermost ring is continuous around the available section. Thus, with the clustering of these and other dates around AD 1877, we have a good date for the construction of the lodge structure at the Bear Lake site. An interesting finding is that samples from the iodge itself (Fea. B) cross-date with the structure to the immediate east of the lodge (Fea. H), which was constructed at the same May 18 , 2003 -289 time. The three bark dates of AD 1851 and AD 1873 most likely indicate the use of wood that had been previously cut or of trees that had been dead at the time of their procurement. Given that the time space between the bark dates of 1851 and those in the 1870s is about the time span for a generation, it might be that the lodge was reconstructed after this time. This might expiain the presence of the two sets of rafter support posts. It is unlikely, though, that this structure was used for more than two decades given the very sparse historical remains, or that it was reused after twenty years, given the comments earlier about the thermodynamic inefficiency of this type of structure. If one burned up all the easy downed wood in the area, it would probably take more than 20 years for this to replenish itself, unless there was a pine beetle infestation. For these reasons, given that the best dates are those of 1877, we believe that this is the date for the initial construction of the lodge, and that older wood was incorporated into the structure and into Feature H. Perhaps the most significant result of the dendrochronological study is that we can begin to ask questions about events occurring within such small time periods. The date of 1890 vv for the Lingfield Creek lodge, aithough not a cutting date, is thought to be within a few years of the actual cutting date, given the cross-section (see appendix) and the age (82 years) of the tree. Lane (1953:145) refers to another style of rectangular lodge which starts out with three or four rows of logs cribbed on the ground. The Lingfield Creek lodge appears to be a variant of this form with notched logs at the end, and thus a variant using an axe. This is likely a transitional form between the older type of rectangular lodge and the log cabins which replaced them for May 18, 2003-290 winter habitations by 1900 (Teit 1909:759;776). The date of the Lingfield Creek lodge thus provides a good estimate of the time when this style of Athapaskan iodges was still being built in the region. It should be noted here that a date of AD 1561 for site CR 73 excavated in 1979 that was previously provided by Parker, is no longer considered reliable. All the dated samples are of lodgepole pine and, as is obvious from the pith dates, the trees were relatively young. Our coring of living lodgepole pines also demonstrated that few lodgepoles in the area today are older than 100 years. In summary, it appears that trees growing in the Eagie Lake region are of adequate quality for dendrochronological purposes, and that while Douglas fir trees live to 600 years of age in the region, lodgepole pine can be used for dendrochronological purposes. This is fortunate, for almost all of our samples were of lodgepole pine. Bear Lake site (EkSa 36) samples exhibit cutting dates of 1851 to 1877, with the most common cutting date being 1877. None of the four samples collected from the Boyd site (EkSa 32) was of dateable quality. Six new living tree chronologies were built for the Eagle Lake region, which provide a foundation for future tree-ring dating in the area, and which also should allow crossdating with other regions using the computer crossdating methods developed by Marion Parker. Radiocarbon Dating of other Eagle Lake sites. In 1979 we received radiocarbon age estimates for samples from three other sites. Site CR 73 (EkSa 35), an isolated, small, circular housepit on the edge of the Chilko River yielded a date of 360 + /- 80 BP (SFU 15) and lithic scatter site CR 92 (EkSa May 18, 2003-291 33) yielded a d ate of 860 + /- 80 BP (SFU 14). Notice that the ''tentative\"dendrochronological date, since rejected, discussed above, and the radiocarbon date from CR 73 are consistent. Our excavations into a lithic scatter area at site EIRw 4 yielded charcoal with an age of 280 + / - 80 BP (SFU 16). Given the dates of the Bear Lake, Boyd and Shields sites, the CR 92 date should indicate a PPT component, which is supported by analyses reported iater in this voiume. \u00EF\u00BF\u00BD 1 2 3 Site CR 73 (EkSa 35) EkSa 33 (CR 92) E1Rw4 Probable Age 360 +/-80 BP 860 +/- 80 BP 280 +/-80 BP (Lab#) (SFU 15) (SFU 14) (SFU 16) Table V-38. Radiocarbon dates from other Eagle Lake Sites. Potato Mountain Radiocarbon Dates. Besides testing at Eagle Lake, testing also occurred during the Potato Mountain survey mentioned earlier (Alexander and Matson 1987). The charcoal (fable V-39) was recovered two contexts; first, as series of smaii roasting pits, and second, from the stratified Mountain Fan Site. Sa ple Site Date 1 RP T84-3 (EjSb 3) 103.55% (Modern) 2 RP P 6-1 100+/-60 BP 3 RP P 2-3 450 + / -70 BP 4 RP P 2-9 615 + / - 80 BP 5 RP P 8-3 1680 + / - 90 BP (Mountain Fan Roasting Pit) 6 RP P 6-1 1710 +/-90 BP 7 RP P 6-1 1910 + / - 50 BP 8 P8-3, Layer 3 960 + / - 80 BP (Mountain Fan, EU 4, 20-35 cm bsd) 9 P8-3, Layer 5 2220 + / - 80 BP (Mountain Fan, EU 3, 34-38 cm bsd) Table V-39. Potato Mountain Radiocarbon Dates. May 18, 2003-292 Lab. No. wsu 3377 wsu 3378 wsu 3376 wsu 3373 wsu 3380 wsu 3372 wsu 3381 wsu 3374 wsu 3375 The most recent date is from a sample collected in 1984, prior to the iarger Potato Mountain Project, which is interpreted as a contaminated sample. The next is from a very iarge site, P 6-1 (EjSb 12), at 100 +/- 60 BP (WSU 3378). This charcoal is actualiy from a burned rock pile next to the roasting pit, interpreted as a clean out from the pit It may, of course, actually be from a forest fire after the use of the site. WSU 3376, 450 + /- 70 BP is from P 2-3 (EjSb 26), a large (140 cm in diameter) roasting pit, in a site withfew pits and within the forest edge. The roasting pit appears to have been re-used and there appears to be no reason to question this date. WSU 3373, 615 + /- 80 BP from a roasting pit in P 2-9 is a similar roasting pit, and also with evidence of reuse. A roasting pit in P 8-3 (Mountain Fan site) yielded a date of 1680 + / - 90 B.P. (WSU 3380). This site is well away from any closed canopy forest, on an old stream fan which also had the only stratified site in the area. The next two dates are also from an open site, with large numbers of roasting pits, P 6-11710 +/-90 B.P. (WSU 3372) and 1910 +/- 50 B.P. (WSU 3381). These are dates from two different roasting pits. These roasting pit dates can be interpreted in a number of different ways. One way is to interpret this as continuous use of the area for mountain potato collecting and processing over the last 2000 years. Another way is to point to the three relatively old dates and say they indicate the use of large sites out in the open circa 1500 to 2000 BP, and a use in the last 700 years of smaller sites located along the forest edge. This would involve rejecting the date of 100 + /- 60 BP as contaminated. Further dates would be needed to demonstrate either interpretation convincingly. A simple simulation, however, demonstrated that the gap is likely a statistical \"artifact''. Matson wrote a simple program in BASIC (in 1986) to produce a 100 random May 18, 2003-293 dates ranging from Oto 2000 BP then to sample 7 of these 100 dates. This program ran 50 cycles. On 18 of the 50 cycles gaps of 720 years or greater occurred. The current set of 7 roasting pit dates has a gap of 720 years. Therefore about a third of the time such a gap will occur by chance. The probability is that there is a continuity in use of roasting pits over the last 2000 years, although whether the open areas and edge of forest areas were used over the entire period remains open. In general, the roasting pits indicate the extensive use of spring beauty over the last 2000 years. Most other roasting pit dating evidence indicates an initiation a little more than 2000 years ago (Pokotylo and Froese 1983; Peacock 1998), albeit that information is mainly for mid-level elevations and the likely use of balsamroot sunflower. Thus extensive use of roots appears to occur at the beginning of the Plateau Horizon of the PPT, and includes the Potato Mountains, as well as the area surrounding Eagle Lake. Two other dates (the last two in Table from the Potato Mountain are from the Mountain Fan site (P8-3, EjSb 39), the only stratified site in the area. In addition to the 50 x 50 cm test in a roasting pit, two other 50 x 50 cm tests were placed, followed by two 1 x lm units along the edge of the fan. Although now Lingfield Creek now flows by the fan in a steep ravine 5 m deep, in the not too-distance past it deposited the fan. The two larger units revealed three cultural layers with charcoal and lithics, separated by fan deposits. The middle layer (3) in Unit 4 produced a date of 960 + /- 80 BP (WSU 3374), at 20-35 cm below surface, while the lowest layer (5) was dated to 2220 + / - 80 BP (WSU 3375), 34-38 cm below the surface of Unit 3. The older date probably indicates the use of roasting pits as well at this time, while the roasting pit date from this site (1680 + /- 90 May 18, 2003-294 0 P.P.T. Dates A1hapaskan Dates Probable Athapaskan Dates Bear Lake Dendrodates I I wsu 3374 I ISFU 14 I wsu 3379 I IWSU3373 I I Iwsu 2905 \u00EF\u00BF\u00BD w\u00EF\u00BF\u00BDu 3376 .. --+--\u00EF\u00BF\u00BD SFU 16 EB H-1L Bear Lake\u00EF\u00BF\u00BD t--i--4 \u00E2\u0080\u00A2\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7+\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7IWSU 3378 SFU 15 I I I 250 500 725 I I 1000 1250 RCYBP (1950) t---H WSU3375 wsu 3381 I I WSU3372 I I Iwsu 3380 I I I80S2264 I I I I 1500 1750 2000 2250 BP) indicates a use that did not result in a preserved cultural layer between the second and third such preserved. The younger date falls into the gap in roasting pit dates and points to the strong likelihood that there are probably roasting pits on Potato Mountain that also date to about 1000 years ago. It is interesting to combine the Eagle Lake and Potato Mountain dates and look at the combined distribution (Figure V-5 ). Even though we can not directly identify prehistoric material from Potato Mountain as Athapaskan or PPT, it is safe to assume that when Eagle Lake was Athapaskan Potato Mountain was also, and the same for PPT Thus most of the dates in Figure (V-5) are PPTwith the identified Athapaskan ones and likely Athapaskan one separately identified. FIGURE Figure V-5. Distribution of Eagle Lake and Potato Mountain dates. This distribution shows a relatively continuous distribution of dates over the last 1000 years, consistent with a relatively small hiatus between PPT and Athapaskan occupations. The apparent gap between 700 and 1500 years ago seen in the Potato Mountain date distribution now becomes a gap between 1000 and 1500 B.P. There then is a relatively continuous distribution from 1500 until 2200 B.P. With the larger number of dates (n of 16) the question becomes whether this gap can be expected by chance. This was investigated in a similar fashion to the earlier question about the date distribution from Potato Mountain. For these circumstances, Matson wrote a less simple program in Modula 2 that simulated date distributions. lt returned similar May 18, 2003-295 results as the earlier BASIC program when the Potato Mountain dates were simulated. When simulating the distribution of 16 dates from both Potato Mountain and Eagle Lake it was run four times, each time simulating 100 samples. The number of times gaps larger t han that seen in the actual date distribution (690 years) were found in each run of 100 samples were, 1, 2, 7, and 7. The number of times gaps larger than 600 were found in these runs were 5,5, 11, and 16. In sum, in no case were more than 7 percent of the sample runs was a gap as large as that observed, and in one case, only 1 out of 100 runs was the gap as large as the one observed. The sampling simulation of dating distribution indicates that it is unlikely, but not impossible, that the observed gap in radiocarbon dates is an artifact of the number of sites that have been dated. The obvious question then becomes if there is any kind of archaeological remains that have been avoided, or is this gap apt to represent a true hiatus in occupation. And, in fact, we have avoided dating the very large housepit villages of very large depressions that we thought members of the Lillooet phenomenon, which elsewhere dates to this gap, between 1000 and 1500 B.P. We return to this subject when the dated Lillooet site (EkRo 18) from the Mouth of Chilcotin is discussed below. Two dates have been recently reported from EjSa 11 aka CR #1, a large pithouse site of large pithouses near the outlet of Chilko Lake (Klassen 2002) which do not close the \"Gap\". These are 2160 + / - 40 BP (Beta 165873) and 2010 + / - 40 BP (Beta 16587 4) and presumably date a Plateau Horizon occupation before the hypothesized Lillooet phenomenon at this site. These dates extend the known PPT occupation in the Eagle Lake area, and are nearly the same age of the oldest date reported from Potato May 18, 2003-296 Mountain. If these dates were included in the simulation (either as an averaged date, as a single occupation may be involved, or as two dates, if one argues for two different events) the probability of the gap occurring by chance would even be less than that demonstrated above. Taseko Lake Dates Seven charcoal samples were submitted for dating (fable V-40 ) as part of the Taseko Lakes Project (Magne 1985 b) of which four returned prehistoric dates. Three of these were from EhRv 2 , the very large site with 100 or more cultural depressions at Fishem Lake. This site includes very large housepits that we interpret as examples of the Lillooet phenomenon (Magne 1985 b). The other prehistoric date is from a roasting pit in Yohetta valley (EhRw 12 ). Sample No. Date Site Provenience WSU3015 1975 + I -70 BP EhRv2 Eroding Housepit wsu 3026 (Unacceptable sample, too small to date) wsu 3027 1660 + /-80 BP EhRv2 Housepit 1, Floor wsu 3028 1165 + I -70 BP EhRv2 Housepit 2, Floor wsu 3029 55 +/-70 BP EhRw 11 Feature A ( Roasting pit) wsu 3030 210 +/-55 BP EhRw 12 Feature A ( Roasting pit) WSU2907 (Less than 150 BP) EhRw 12 Feature E ( Roasting pit) Gak5325 1290 + I -80 BP (MOC) EkRo 18 (36 cm bsd, Housepit) Gak5326 870 +/-80 BP (MOC) EkRo 48 (8 cm bsd; Housepit) Gak5327 1450 +/-75 BP (MOC) EkRo 48 (27 cm bsd; Housepit) Table V-40. Radiocarbon Dates from Taseko Lakes and the Mouth of the Chiicotin. May 18, 2003-297 The dates from EhRv 2 indicate an extensive history, extending well beyond the expected dates of the Lillooet Phenomenon (Lenert 2001 ). WSU 3025 is from an eroding housepit not of the largest size, showing occupation of this housepit village began at least 2000 years ago, a date essentially equivalent to the two dates from EjSa 11 (aka CR #1). Both housepits 1 and 2 from EhRv 2 are very large (circa 15m in diameter) structures that were tested by single 1 x 1 m units. In both cases (WSU 3027 and 3028) the s amples were from deposits that were inferred to be floor surfaces, but only WSU 3028 from housepit 2 provided the circa 1200 BP date expected for the Lillooet Phenomenon. The 1660 +/- 80 BP of WSU 3027 from housepit 1 is siightly early for the Lillooet Phenomenon, but may date a pre-Lillooet occupation or be the result of the re\u00C2\u00ADuse of construction wood. The dates from the three roasting pits in Yohetta Valley are from pits of substantial size, with EhRw 11 Feature A and EhRw 12 Feature A being in the 4m in diameter range, both excavated by a singie 1 x 1 m unit. All three showed extensive fire-cracked rock and abundant charcoal (Magne 1985). The inconsistency in dates probably has to do with the variations in carbon 14 in the last few hundred years (Stuiver 1978) rather than intermixture of modern materiaL Thus these dates can all be interpreted as showing the last use being in the last 100-250 years. Mouth of Chilcotin Dates. As briefly reviewed in an earlier chapter, three radiocarbon dates also exist for some of the Mouth of Chilcotin test excavations, as shown in Table V-40. These are May 18, 2003-298 from two different house pit sites. Ek.Ro 18 is a pithouse site of very large and deep pits which was interpreted as a member of the Li1looet phenomenon. At the time of excavations the date of 1290 + /- 80 BP appeared to be recent for this, but according to Lenert (2001) this i s at the peak distribution of such dated large house pits around Lillooet. Lenert (2001) shows that the existing dates of the very large house pits range between 1000 and 1500 RCYBP. So this date supports the interpretation of EkRo 18 as a member of this phenomenon, but extends the range at least to the Mouth of the Chilcotin. The dates of the Lillooet phenomenon (including EkRo 18 and EhRv 2 from Taseko Lakes) fit into the dating \"Gap\" (Figure V-5) found at Eagle Lake and Potato Mountain. The obvious assumption is that if the apparent Lillooet components in the Eagle Lake area, such as those inferred to be present at EIRw 4 (Quiggly Holes), EkSa 5 (Canoe Crossing), and EjSa 11 (CR #1), were dated they would date to this time. This raises the question whether during the Lillooet time if all the inhabitants in the Eagle L ake area were concentrated in a limited number of very large sites. Further, the absence of dated material from Potato Mountain may indicate that the alpine area was not extensively utilized at this time. We raise these as interesting hypotheses to be tested, not as ideas that are supported by a significant amount of empirical information. The other MOC site dated is EkRo 48 which has smaller housepits from which two Kamloops side-notched points excavated and many more from the surface. The later radiocarbon date is in accord with this assemblage and even the one of 1450 BP, from the floor is not necessary out of line, given that this is likely structural wood, and m ay have inner rings laid down hundreds of years before the housepit was built. May 18, 2003-299 A.D.1000A.D.500Eagle Lake Eagle Lake ? ? A.D.1650? A.D.1650------------ A.D.1475 ------------ A.D.1475Kamloops Horizon Kamloops Horizon A.D.1000Ullooet A.D.800 Ullooet Phenomenon Phenomeoon A.D.500Plateau Horizon Plateau Horizon A.D.0 A.D.OOPPT [21 Athapaskan Alternative Regional Sequences Ft c--oR\u00EF\u00BF\u00BD r G, ri.i_ \u00EF\u00BF\u00BDAN A \u00EF\u00BF\u00BDr-, V rS L..A- i<:,\u00EF\u00BF\u00BD St. < ... .. . ... nain . \u00E2\u0080\u00A2 . . ;_. '; . \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD,\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD8:;\u00EF\u00BF\u00BD\u00EF\u00BF\u00BDai . }\u00EF\u00BF\u00BD.8.\u00EF\u00BF\u00BD.\\ .. Kutdlin\u00C2\u00B7::\u00EF\u00BF\u00BD:=:i\u00EF\u00BF\u00BD:;\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\ \u00EF\u00BF\u00BDd' \u00EF\u00BF\u00BD .... , ! \u00C2\u00B7\u00C2\u00B7\u00C2\u00B7/.\u00C2\u00B7\u00C2\u00B7 \u00C2\u00B7 ........... .:' r Tutct>ooe \u00E2\u0080\u00A2\u00E2\u0080\u00A2 - Boundary between Padfic : and Arctic Drainages Bear Lake \u00EF\u00BF\u00BD\u00EF\u00BF\u00BDH;I. ,..... .-\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 f \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \ \u00C2\u00B7 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \ Dogrib.\u00C2\u00B7'.,,.,uowt }p ,;-' .. \u00EF\u00BF\u00BD SekMi ' Beaver .\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7 D \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \\u00EF\u00BF\u00BD i-. .\u00E2\u0080\u00A2\u00C2\u00B7 '\. i v ........ . -,1 ! . \u00C2\u00B7,.,., \u00C2\u00A3 . Carner \u00C2\u00B7,.,.._ .. \u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00EF\u00BF\u00BD =J..N\".JA U.S.A \ \ APACHEAN \u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2 J1car\u00EF\u00BF\u00BD1\u00EF\u00BF\u00BD. \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00E2\u0080\u00A2\u00C2\u00B7\u00E2\u0080\u00A2\u00E2\u0080\u00A2, : Kiowa \u00E2\u0080\u00A2 \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD.;'Navajo\. \u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7: ! Apadle : \.\u00C2\u00B7\u00C2\u00B7 ... -::::t./\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\-:.::\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7 .. J \u00EF\u00BF\u00BDwe:\u00EF\u00BF\u00BD; .. /_\" . \u00EF\u00BF\u00BD .. '(-P / : Mesca1ero \u00E2\u0080\u00A2., \u00C2\u00B7-. :._;<- '-:..: \., ...! I Chiricahua:.= \u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2 \u00EF\u00BF\u00BD-\u00EF\u00BF\u00BD .\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7\u00C2\u00B7,' Lipan \u00E2\u0080\u00A2 \u00EF\u00BF\u00BD \u00E2\u0080\u00A2\" . - ;\u00E2\u0080\u00A2v!'!\"!..\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00C2\u00B7 : tr.lEXICO : \u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2\u00E2\u0080\u00A2 !'-Toooso\u00C2\u00B7 Figure VII-L Distribution of Athapaskan Speakers. The Plateau Athapaskan groups are the Nicoia and Chilcotin with the Carrier just to the north in centrai B.C. They are all in the interior Pacific drainage, well disjunct from the southern Northwest Coast groups (PCA) such as the Hupa and Toiowa in northwestern C alifornia and southern Oregon. An area just inland from the northern British Columbia Coast once was home to the Tsetsaut (Duff 1981), who are barely documented (as is true for the Nicola). As Figure VII-1 shows, the B.C. Athapaskans are part of a continuous distribution which in the northern interior of B.C. and adjacent northwestern Alberta includes the Sekani, Tahltan, and Kaska with the distantlv .I related Tlingit on the Alaskan Panhandle. North of the Tlingit are the Eyak, the other language widely recognized to be related to Athapaskan, but not usually included within it. Further to the east in the Subarctic, the Beaver, Slavey, Mountain Dene, Dogrib, Yellowknife, and Chipewyan are found, while further to the north and west in the Yukon and Alaska there are the many groups inhabiting what, according to linguistic diversification, must have been the Athapaskan homeland (fagish, Tutchone, Ahtna, Han, Kutchin, Tanana, Tanaina, Kolchan, Koyukon, Holikachuk, lngalik). On the Plains, contiguous to, and northeast of the Apacheans are the Kiowa Apache. The Tsuu Tina (Sarsi) are found imrnediately south of the Beaver, although as discussed below both these groups have moved significantly in historic times. In the Southwest the Navajo and Apache, the Apacheans, are spread over a wide area, but are not (or were not) the exclusive inhabitants, as many other linguistic groups also are located in the Southwest. Figure VII-1 shows that both the Pacific Coast Athapaskans and the Apacheans are well separated from the northern Athapaskans. May 20, 2003-387 With this brief survey in hand, we now wish to provide an overview of the nature of the ethnographic, linguistic, and archaeological evidence as well as arguments regarding recent Athapaskan movements, not forgetting the Chilcotin. Then we will move onto a consideration of how and why the migrations occurred. Linguistics The linguistic evidence of the unity of the Athapaskan language despite their geographical spread has been evident for more than 150 years. Goddard (1996:294) reports that the connection between Northern Athapaskan and Apachean languages was first reported in 1852 and the Athapaskan nature of the PAC languages in 1855. The reason for these separations has been the subject of voluminous study for over 100 years (e.g., Morice 1890, 1893, 1895). ln fact the most overriding evidence for Athapaskan migrations is linguistic, which Foster (1996) reports was first convincingly demonstrated by Sapir in a number of publications in the first half of the 20th century. The most definitive treatment has been that of Dyen and Aberle (197 4) who applied a rigorous methodology to linguistics and kinship to reconstruct Proto-Athapaskan kinship and its evolution. Among many details of concern primarily to linguists, the linguistic evidence provides important generalizations regarding the nature of Athapaskan movements. The following inferences come from the linguistic research (see Campbell 1997:110-113; Dyen and Aberle 1974; Kincade and Poweli 1976; Hoijer 1963, 1971; Krauss 1973, 1979; Krauss and Golla 1981; Foster 1996). L The direction of general movement was from north to south, with some evidence of May 20, 2003-388 later northward movement by Kiowa-Apache; 2. The Pacific Coast Athapaskans appear to have diverged from the northern groupsslightly earlier than the Apacheans. Kincade and Powell (1976: 89) state that \"Because of their greater divergence from the northern languages, the Pacific coast language speakers must have migrated southward earlier than the south-western Athapaskans.'' (see also Hoijer 1956 and Krauss 1973). 3. The Plateau and Pacific interior drainage Athapaskans are linguistically most closelyrelated to the Pacific Coast group, while the southern Apacheans are more closely related to the eastern Subarctic (Chipewyan, Beaver), and Interior B.C. (Can-ier) groups; 4. The Tsuu Tina (Sarsi) are linguistically more closely reiated to Sekani than to Beaver,but share substantial linguistic traits with Chipewyan and Carrier; Although most of these appear to be robust patterns, the weak part of the linguistic analyses concerns attempts to date the times of divergence, usually though lexicostatistics/giottochronology. There is a substantial literature devoted to dating linguistic divergence, much of it using lexicostatistics. There is, however, a deep current of dissatisfaction with lexicostatistics (e.g., Hoijer 1956; Hymes 1957; Campbell 1998; Crowley 1997) since the dates obtained for linguistic divergence can often be shown to be unreliable. Such studies as the lexical study of Athapaskan kinship by Dyen and Aberle (1974), demonstrate that methodologically sound linguistic studies can reveal May 20, 2003-389 much about culture change in the relative sense without the need to fix absolute dates on the changes in question. Furthermore, recent historical linguistic work (e.g., Hill 2001) appears to often accept lexicostatistical information as a useful first approximation, which was not the case a few years ago. A brief summary of our understanding of ''glottochronology\" is in order. The basic theory b ehind lexicostatistics is relatively simple and widely accepted; the devil is in the details. The main idea is that over time there is a loss of core words in a vocabulary, a postulate that is accepted by all. The details are the assumptions that: 1) a single loss rate is common to all languages; 2) that there is a cross-cultural core vocabulary that is adequately represented by either the 200 or 100 Swadesh list; and 3) that these lists can be reliably compared by identifying cognates. All of these assumptions have been heavily criticized (Campbell 1998; Crowley 1997; Lehmann 1962). First, it has been demonstrated that the loss rate is variable (Campbell 1998:184; Lehmann 1962:110) casting doubt on the usefulness of assuming a single rate. Second, the use of at least three core lists of different sizes indicates that a core vocabulary is not easy to identify, and there are strong theoretical reasons to believe that cores will vary cross-culturally (Campbell 1998:180-183; Crowley 1997:173-175; Lehmann 1962:110). Finally, the identification of cognates is not easy or well standardized (Crowley 1997:183-184). Depending on your view of these problems, this method is absoluteiy useless (actually worse than useless, as it is misleading) except in a few unusual situations (Campbell 1998:186); has very serious problems in practice (Crowley 1997); is one of a series of tools to be used (Foster 1996; Lehmann 1962:110-111 ), or is a very useful tool (Krauss 1973,1979, our interpretation). May 20, 2003-390 Some of these problems are reduced if a single linguist is making the core lists and the identification of cognates (Crowley 1997:184) and if the languages being compared are in similar settings and at a similar level of technology. (Iceland's rate of change is different from other Germanic languages, but that would be expected because of its isolated setting.) Furthermore, although there are obvious problems in the reliability of the absolute dates produced by glottochronology, there are problems with any alternative linguistic dating techniques which often vary widely from those obtained by archaeology when the two can be compared (Bellwood et al. 2002). In our case, we will be using lexicostatistics as useful hypotheses, pointing out some of the weaknesses when they are apparent to us. Dates given by us without citations are calculated from Hoijer's lists (1962 with corrections) as published in Dyen and Aberle (1974:12-13). Since these lists and identification of cognates were all done by Hoijer, some of the potential problems with lexicostatistics are reduced. Furthermore, for the most part, we are dealing with groups in similar settings and at a similar level of technology. An additional problem with glottochronology is that it does not extend very far into the past, but that is not a concern here as all the material we are most interested in is well within both the theoretical range of circa 6000 years and practical range of about 4000 years. Although the calendric dates derived from glottochronology may be in error, it is clear the Pacific and Apachean Athapaskan divergences occurred in late prehistory. The relative orders of divergences are widely accepted and that lexicostatistical use is seen as more reliable than trying to produce absolute dates. According to Foster (1996:76) the initial split between Eyak and Athapaskan in May 20, 2003-391 Alaska is 3400 BP+/- 500. Within Athapaskan the earliest split in the Alaskan-Yukon homeland is thought to have occurred about 2400 BP + / - 400 years ago. The separation of southern C anadian Plateau and Plains groups from the Apacheans (Navajo and Apache) Foster (1996:75) dates to A.D. 1000 and we calculated the divergence between Chipewyan (the northern Athapaskan language most similar to Apachean) and Navajo at 909 years. Hoijer (1971) places the Apachean separation slightly earlier at ca. 1200 to 1300 BP. The coincidence of Hoijer' s (1971) timing of the initial split with the date (as it was known in the middle 1970s, Lerbekmo et al. 1975) of the White River Ash fall which covered much of southern Yukon about 1200 years ago was what prompted Workman (1978,1979) to speculate that this volcanic event was the prime cause of the initial migration towards the U.S. Southwest. Clague et al. (1995) have recently redated the eastern lobe of the White River tephra at 803 A.D. ( + / - 60 years), supporting this correspondence between the timing of these two events. The Carrier have been where they are for a long time. According to Dyen and Aberle (1974: 250), Carrier dialects have deep subdivisions which suggest long occupancy and a lengthy differentiation in the area. The Chipewyan/Carrier split is more recent than the Apachean split from Canadian languages, at a calculated 620 years. Since the Northern Athapaskan languages tend to form linguistic dines or dialectic chains, lexicostatistics probably underestimates their time of separation, while the Apacheans are too far away to be part of any chain, thus making the estimate of that split apt to be more reliable. Thus the separation of Chipewyan and Carrier may in fact be closer in time to the Apachean split than indicated by lexicalstatistics. Interestingly enough, given the long recognized linguistic similarity between the May 20, 2003-392 eastern Subarctic Athapaskans and the Apacheans, the Carrier are the second most similar northern group to the Navajo (1042 years) in Hoijer's table. Given the iinguistic similarity of the Chipewyan and Carrier cited above, we have three possible scenarios, that the Apachean either split off from: 1) Carrier; 2) Chipewyan; 3) a recent common ancestor of all three. Davis (1975:624) reports somewhat different lexicostatistical dates for some of the same relationships. He has the Chipewyan-Navajo divergence at 788 years, greater than that of the Navajo-Carrier at 663. Chilcotin-Navajo is measured at 889, and Chilcotin-Carrier at 60L The same general pattern arises from these numbers, though, \" .. that languages like Navajo, Chipewyan, Chi\cotin and Carrier are representatives of a divergence that took place approximately 1000 years ago, perhaps beginning around northern British Columbia, and that PCA ianguages resulted from an earlier split.\" (Davis 1975:624). The earliest divergence of Pacific Coast Athapaskan is dated at 1000 to 1100 years ago (Chipewyan and Hare vs Galice at 1042; Carrier vs Galice at 1134). Interestingly enough the calculated split between Apachean and Pacific is little different with the calculated divergence between Navajo and Galice at 1042, although the average difference is higher between Pacific Athapaskans and Apachean than with Northern Athapaskans. Tyhurst (1984:349-354), as reviewed earlier, presents evidence that Chilcotin is most closely related to Ulkatcho Carrier. This indicates that the 601-620 year divergence dates for the Carrier and the Chi\cotin are apt to be overestimates and that the separation between the Ulkatcho Carrier and the Chilcotin happened more recently. May 20, 2003-393 The linguistic data in general show Athapaskans (or Eyak-Athapaskan) being in the Yukon/ Alaska homeland approximately as far back as lexicostatistics can be used (around 4000 years) agreeing with almost everyone who has studied this question. Most archaeologists believe ancestors of the Athapaskans were in that area for at least double that time. Moving away from the Alaska/Yukon border, the Northern Athapaskan languages tend to form dialect chains, but show separation times of approximately 2000 years (Slave /Tanaina is 1920 years, Tanaina /Dogrib is 2165) between Arctic drainage and Alaskan Athapaskans, again indicating that Athapaskan speakers have been in that general area for some time. Although the relationship between Pacific drainage and Arctic drainage Athapaskan appears to be close, as the Carrier /Chipewyan figures cited above indicate, Carrier also shows substantial distance from the Alaskan Athapaskan languages (1640 years with Tanaina) which should be indicative of substantial residence in the Southern Yukon, Northern B.C. area. In contrast with this lexicostatistical information, which is in general agreement with other linguistic information (Krauss 1979; Foster 1996:74,75) demonstrating substantial local residence for subarctic Athapaskans, the Pacific and Apachean groups show recent separation from the north. That the Apachean homeland was in the north is unassailable (Foster 1996:75). The lexicostatistics cited above apparentiy fit with other information to show that there are larger differences between Pacific and Northern Athapaskan languages than between Apachean and Northern Athapaskans (Foster 1996:75). This leads to the inference that the Pacific Athapaskans left the north first, with a linguistic separation of 1200 years ago, and the Apacheans at a few hundred May 20, 2003-394 years later. We wonder, though, if the move south for both was really at the same time. Except for the Kiowa-Apache, the Apache/Navajo group has remained as a singie dialect chain, and a relatively large community, spread across the relatively open Southwest. In such a setting would not the group retain much of the linguistics traits it brought with them? In contrast the Pacific Athapaskans have divided into at least four separate languages, and are hemmed in a much denser populated area, with a much more sedentary lifestyle. In such a setting, would not the languages change more, over the same length of time? As we reported above, the Hoijer lists result in dates of 909 and 1042 years, not a very large difference. Hoijer (1971) hypothesized a Proto-Canadian-Pacific-Apachean (PCPA) system, a proto language shared by some northern Athapaskans and the Apachean and Pacific Athapaskans. This is what exactly would be expected if the latter two were the end results of a singie migration. If we assume that a single linguistic group left the north in both cases, the maximum separation existing today among the Pacific Athapaskans and Apachean would give us minimum dates for this event -- in contrast to the separation dates with Northern Athapaskans which should give maximum dates. For the Apacheans, we only have two languages, and the maximum separation is between Kiowa-Apache and Navajo at 865 years (and Kiowa-Apache and the other Apachean dialects give statistically identical figures). For the Pacific Athapaskans the maximum separation is Galice and Mattole at 1134 (the Galice are a small isolated group in interior Oregon), although most of the variations within Pacific Athapaskans are in accord with the Kiowa-Apache differences. Interestingly enough, this 1134 year measure (71 shared May 20, 2003-395 cognates) is the same as Carrier/ Galice difference. Given the argument made above, this is not a large difference, allowing the hypothesis of common departure dates continued viability, alongside the traditional acceptance of a slightly earlier departure date for Pacific Athapaskans. Either idea is consistent with the linguistic evidence. Ethnographic Considerations One could expect that evidence of recent migrations would be included in origin beliefs, and a survey of northern Athapaskan oral tradition reveals traces of such. Moodie et al. (1992) have documented Tutchone, Mountain Dene, Hare, Slave and Dogrib oral traditions of volcanic events which they equate with the White River eruption. They also relate a Hare migration story recorded by Petitot (1886) of Dene people living first on the borders of the western ocean, plateaux and mountains, and then moving east of the mountains (Moody et al. 1992: 159). Unfortunately, however, there is no traceable origin legend between Pacific Athapaskans in Oregon or California and their nearest Plateau neighbours, the Nicola or Chilcotin (see Wyatt 1998; Lane 1981 ). There is an oral tradition of Nicola and Chilcotin relationship, this being that the Nicola descended from a Chilcotin hunting party. Teit (1979: 11) states that the Nicola - Similkameen Athapaskans were a southward- migrating party of Chikotins that became surrounded by Thompson and Okanagans. They made terms and were given territory and were allowed to settle in the upper Nicola and Similkameen valleys. Other traditions say that they had always been there, with territory from just above Spence's Bridge to just below the international boundary and that they were later pushed about by the Okanagan and Thompson. May 20, 2003-396 Bouchard and Kennedy (1984) date this event to ca.1770. Some Nicola interviewed by James Teit late in the late i9th century denied this story, maintaining that they had always been located in southern British Columbia. Today, people of the Upper Similkameen Band recognize partial Nicola ancestry. Interestingly, Teit (1979: 7) relates further southward movement of the early Nicola-Chilcotin: \"It is said that the Antelope afterwards shifted to what is now Montana, where he has many descendants, but some of his children remained in Nicola Valley as they had intermarried with Coyote's offspring.\" This comment has bearing on other historical evidence presented below. For the Chilcotin, Lane (1981 :402) states that many Chilcotin believe they used to live north and northwest of their location at contact. The Chilcotin are sometimes thought of as having moved into Salishan territory in late prehistoric times, but this may have been a long term process of mutual interaction. Teit (1909: 469), for example notes: \"The Canon division (of Shuswap), about fifty years ago, were strongly mixed with Chilcotin, so much so that the people of the North Canon band spoke chiefly Chilcotin in many houses; and the other bands had also a considerable amount of Chilcotin admixture.\" As we move further north on the Interior Piateau, the oral tradition evidence is weak, but there are well documented historic moves by Ulgatcho Carrier, leading to full occupation of former Chilcotin territory around Anahim Lake early in the 20th century (Donahue 1973). These historic Ulgatcho and Chiicotin moves are likely connected to fur trade period events which also appear to have been the principal cause of westward movements by groups such as the Sekani and Beaver. On the Pacific Coast, practically the only ethnographic traces of northern origin May 20, 2003-397 for the California Athapaskans, apart from their language, is a suite of social traits which Jorgensen (1980) labels cultural baggage: certain girls' puberty rites, individual curing rites, raiding, warrior status, and a general focus on individual rather than group accomplishments in life. Baumhoff's summary agrees (1958:158), but adds that they did not have the strong local organization characteristic of Central California. We have also noted, though, some similarities between the rectangular California Athapaskan house structures and those seen in interior B.C. (Baumhoff 1958). In particular, there is an emphasis on rectangular structures, with a single ridge pole, vertical planks or bark slabs at the gable ends and a large opening for smoke along the ridgepole (Baurnhoff 1958:202, Plate 10f, 11a). The Pacific Athapaskans live in two different environments and have two very different adaptations. Some, including the Tolowa and their neighbors along the southern Oregon coast, and the Hupa, participated in the Northwest Coast culture area, while the others (mainly in California) either were further south or inland, on the east side of the Coast range, did not have the large coastal winter villages and did not participate in typical Northwest Coast activities (Baumhoff 1958; Fredrickson 1984). These groups were more typical \"Californian\" with tribelets and less emphasis on wealth and status (Baurnhoff 1958:159). The Pacific Athapaskans in general appear to be very similar to their neighbors in both environments with little remaining of their Northern Athapaskan heritage beyond their language. Looking eastward at the record for Plains Athapaskans, we note that there is no documented interaction between Plains and Plateau or Pacific groups with either, but there were southward and westward movements for Navajo and Apache off the plains May 20, 2003-398 and into the Southwest documented by Spanish explorers in the mid-sixteenth century (Schaafsma 2002). Gulley (2000) has examined the Spanish sources and shows that those attributions can often be in error. Habicht-Mauche (1992), however, s hows that the identification of the bison-hunting-and-dog-using \"Querechos\" found by Coronado in A.D. 1541 in the southern Plains as Apachean is aimost certainly correct. Furthermoreshe points out that the Tierra Blanca complex found in the Texas panhandle is most likely the archaeological equivalent. The Tierra Blanca complex appears to begin at about A.O. 1450 (Habicht-Mauche 1992) making this the earliest widely accepted evidence of Athapaskans in the Greater Southwest. There are also early historical records of northward and westward movements by Kiowa and Kiowa Apache (Mooney 1898) which appear to have been motivated by hostile pressures from Comanche, who are a group of Shoshonean speakers, but this is later, after the horse was integrated into plains culture. There is documented evidence of contacts between Kiowa Apache and Tsuu Tina, in the form of intermarriage, which by Mooney's accounts seem to have taken place on a regular basis. Mooney states (1898: 247) that the Kiowa came from \"the extreme north'' and the Kiowa Apache came withthem, and that \" ... both tribes say they have no memory of a time when they were not together''. Mooney (1898: 246) writes \" ... they never had any political connection with the Apache proper and were probably unaware of their existence until about one hundred years ago\". As pointed out by Gunnerson and Gunnerson (1971), the Kiowa and Tsuu Tina knew each other rather well, even to the point where some prominent Kiowa were of Tsuu Tina blood, considering themselves in a measure related to the Kiowa-Apache (see also Mooney 1898: 247). Brant (1949, 1953), in opposition to May 20, 2003-399 Mooney's position asserts that the Kiowa-Apache split from ''an eastern Apachean prototype\", such as Lipan or Jicarilla (and thus likely descended from the Querechos), and moved northward. Brant supports his conclusions by linguistic and mythological evidence; Dyen and Aberle (1974) also agree. Gunnerson and Gunnerson (1971) on the other hand, hold that later anthropologists (meaning Brant) went too far in discounting Mooney's position. They provide evidence pointing to the Kiowa-Apache being a Dismai River people, who lived in and around the Black Hills of South Dakota, and who moved onto the Plains at about A.O. 1700 with the Kiowa, after the latter had abandoned their relationship with the Crow (Gunnerson and Gunnerson 1971 :19). Two independent maps bear directly on this issue: The first is Mooney's showing Kiowa movements to the Calgary area and from the Yellowstone area (redrawn here as Figure VII-2a). This may account for the origin of the Kiowa Apache in two ways. First by showing an obvious connection to Tsuu T'ina territory. Secondly, another map may relate directly to Teit's note that Antelope moved to Montana. This second map (redrawn here as Figure VII-2b) is that of Ac ko mok ki, a Blackfoot individual, drawn for Peter Fidler of the Hudson's Bay Company in 1801 (Ruggles 1991: 19). This map shows the locations of 32 Aboriginal groups with the numbers of tents in each. Curiously, one group, next to the warrnwater river, or Yellowstone River, is named Beaver. This is the heart of southern Blackfoot country and was brought to our attention bv Prof. Brian Reeves. There is not, nor has there been, any known Blackfoot band .I \u00E2\u0080\u00A2 known as Beaver. Ac ko mok ki was an informed person, well able to show Mand ans (mud house Indians), Flat Heads, and others. There is a very real possibility that the Beaver on Ac ko mok ki's map were Athapaskans. May 20, 2003-400 \ I L 0 I) 0 0 \u00EF\u00BF\u00BD if;, 0 0 0 FLATHEADS0 \u00EF\u00BF\u00BD 0 <;j#,\u00EF\u00BF\u00BD 0 \u00EF\u00BF\u00BD 00 ' I ;-'. : \ I ', . .Eart\u00EF\u00BF\u00BD liartield 1i!i1House 0 0 \=\u00EF\u00BF\u00BD\u00E2\u0080\u00A2w\u00E2\u0080\u00A2\"\u00E2\u0080\u00A2.' I I l io FIGURE Figure VIl-2. a) Mooney's map; b) AcKo mok ki's map. Other than the linguistic ties, the close link between Tsuu T'ina and their northern neighbours the Beaver takes the form of oral traditions, in two versions, describing a split between the two groups. Note, however, that the linguistic evidence shows clearer connections between Tsuu Tina and Sekani and Chipewyan than between Tsuu T'ina and Beaver. As Dyen and Aberle (1974:251) point out, this Tsuu Tina - Sekani connection makes good sense in relation to what is generally known about early fur trade group distributions in northern Alberta, that is, there was considerable east to west movement during that period (and this fits their analysis of shared innovations). As stated by Denniston (1981: 433), prior to Cree expansion, Sekani and Beaver were likely one people. Beaver occupied the territory north and west of Edmonton and eventually moved to the west side of the Rocky Mountains. Tsuu Tina were known to visit Fort Edmonton. Ugarenko (1979) places Beaver territory prior to 1700 south of Lake Athabasca, with later displacement westward caused by the Cree. Among the Beaver, Tsuu Tina and Chipewyan, legends describing past groups' co-associations exist in the form of direct interactions, in the area of Lake Athabasca. Wilson (1888: 11) records: 'Formerly,' said 'Bull's Head,' ' the Sarcee territory extended from the Rocky Mountains to the Big River' (probably the Peace River). Also (Wilson 1888:11): \"Another Indian told us how the Sarcee were at one time one people May 20, 2003-401 with the Chipewyan and gave us the myth which accounts for their s eparation. 'Formerly,' he said, 'we lived in the north country. We were many thousands in number. We were travelling south. It was winter, and we had to cross a big lake on the ice. There was an elk's horn sticking out of the ice. A squaw went and struck the horn with an axe. The elk raised himself from the ice and shook his head. The people were all frightened and ran away. Those that ran toward the north became Chipewyans, and we who ran toward the south are the Sotenna or Sarcees.\" Hale (1885) notes that Petitot says the southernmost tribe of Petitot's \"montagnards\" Tinneh are Tsa-ttinne, rendered as \"dwellers among the beavers\", derived from tsa, beaver and tinne, man. Hale (1885: 21-22) goes on: \"M. Petitot described the Tsa-ttinne or 'Beaver Indians' as comprising two septs - a northern tribe, who hunt along the Peace River, and a southern, who dwell about the head-waters of the North Saskatchewan, toward the Rockv Mountains. The latter, he savs, are the , ; Sarcis, who have separated themselves from the northern band. H. Their southward advance brought them into contact with the Blackfeet, with whom they confederated, not against their Tinneh kindred, as had been supposed, but against the Cree, who have from time immemorial been the common enemies of the Tinneh and Blackfoot tribes.'' According to Jenness (1938) Sir Alexander Mackenzie noted \"Sarsees'' at headwaters of the North Saskatchewan in 1792. They were known to Umfreville at Cumberland House in 1790, to McGillivray at Fort George in 1794, and to Alexander Henry the Younger in 1810-11, also at Fort George, who says that they occupied the area of the Beaver Hills. Therefore, Tsuu Tina in 1700s were in the upper headwaters of North Saskatchewan and Athabasca Rivers and by the early 1800s had moved down river May 20, 2003-402 (eastward), probably to be closer to the later established fur trade forts. This ethnohistoric information is consistent in that, despite of their geographic location on the map, the Sarci-Tsuu Tina should not be looked at as a likely source for the Apacheans. Their southernmost northern Athapaskan position is one that has only developed relatively recently in the post-contact environment (Dyen and Aberle 1974:251). The Tsuu Tina, however, definitely made southern treks in the historic era. For example, Hale (1885) states that in 1870 the Tsuu Tina were at Maria's River in Montana (see also Jenness 1938: 7). Jenness also writes: \"In the early part of the nineteenth century, apparently, seven Sarcee warriors who had travelled far to the southward discovered a camp of \"Utah\" Indians.\" One of the warriors ends up marrying a woman, then years later brings her back to the main Sarcee camp, accompanied by some of her brothers. \"The Sarcee joyfully welcomed the Utah Indians and entertained them for several weeks \" 0enness 1938:7). In general, then, the ethnohistoric data and known historic distributions for the Plateau/Pacific groups reveal that the Plateau and Pacific groups are unconnected ethnohistorically, while the Tsuu Tina only imply, via Kiowa Apache relationships southward with the other Apacheans, a long-standing continuous stretch of Athapaskan social connections along the east side of the Rocky Mountains. The linguistic evidence and the Dyen and Aberle analysis (1974:425) indicate the Kiowa Apache derived from the other Apacheans. Bamforth (1988:90) shows a similar interpretation for the distribution of ethnic groups along the middle and central portions of this distribution at A.D. 1700. Although Bamforth's proposed distribution is May 20, 2003-403 Paci/Jc OcHn \u00E2\u0080\u00A2 Microblade OKavik \u00E2\u0080\u00A2 Kavik and Microbfade 0 Atigun , , . Q ' {) . ! c:::;i l) .\u00E2\u0080\u00A2 ,, 0 ' Dixthada , KlOklll : . [S , '\u00C2\u00B7 0MdPs5 '\u00C2\u00B7 \u00C2\u00B7, \u00EF\u00BF\u00BD,-:::, It-:,. I ;\u00EF\u00BF\u00BD ' FtiVermillion 1 \u00C2\u00B7, ,\u00C2\u00B7 . -\u00EF\u00BF\u00BDtJ Peace POint : \u00C2\u00B7.\{ {, I Bezya.; . .. S \u00EF\u00BF\u00BDhmla\u00EF\u00BF\u00BD, Ft McMurray; y1gate110e ; _ MShfTI Lake\u00E2\u0080\u00A2 \u00C2\u00B7, ' \u00EF\u00BF\u00BD,ar La;;;\u00EF\u00BF\u00BD::\u00EF\u00BF\u00BDr. ,,, : \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD - \u00E2\u0080\u00A2\u00E2\u0080\u00A2 _ \u00E2\u0080\u00A2\u00E2\u0080\u00A2 1 Htgh R1\u00EF\u00BF\u00BD11r ! i \J! .' -. \u00C2\u00B7--- i ' 1 , - \u00E2\u0080\u00A2\u00E2\u0080\u00A2 - .. _ .. - .,. _. ( CANADA \"? \u00EF\u00BF\u00BD:.\u00EF\u00BF\u00BDy ' \u00EF\u00BF\u00BDI Terrace ; - \u00C2\u00B7u.iA - ;r.-i Standley . - .. - . '.-J \u00C2\u00B7 i \u00C2\u00B7-, \ (\u00C2\u00B7-. dependent on historical events, it points to a slightiy different arrangement of Athapaskan groups existing in a similar continuous chain along the eastern slopes of the Rocky Mountains. Nonetheless, the Tsuu T'ina's southward position appears to have occurred quite late in time. To the west, although there is no link between Pacific Coast and Plateau groups, there is some very limited evidence of linkage between Chilcotin/Nicola to the central Plains. Interestingly enough, despite their relatively recent arrival in the Southwest, Apachean origin stories show iittle awareness of northern genesis. Archaeological Evidence Archaeological research has been concentrated at the extremes of Athapaskan distribution, with far more work having been completed in the southern-most Apachean areas than in the northern Arctic drainage area of Yukon and Alaska, and even less work on the Plateau. Although two volumes have been devoted to archaeological examinations of the northern Athapaskan Question (Derry and Hudson 1975; Helmer et al. 1977), it is rare (but not unknown) among the archaeological studies to have reliable confirmation of northern Athapaskan presence in the form of several lines of material evidence such as we have demonstrated for the Eagle Lake Project area. We believe, however, that there are indeed clear associations of particular artifact types with Athapaskans (Figure Vll-3), but these are not what Kehoe, Wilcox, nor Perry have suggested. FIGURE May 20, 2003-404 \u00E2\u0080\u00A2Figure VII-3 Northern Athapaskan Sites. In the Yukon Territory and Alaska, Morlan, Wilson, and Shinkwin have with reasonable certainty traced Athapaskan assemblages into prehistory. In the Yukon, Morlan (1973) presents convincing evidence that the weli-stratified Klo-Kut site represents continuous Athapaskan occupation from ca. 1200 BP to the present, divided into three main periods. Klo-Kut, also with the distinctive Kavik points (Figure VII-3,4), generally has its \"strongest ties ... to the west along the Brooks Range\" (Morlan 1973:ii). At Dixthada in Alaska, Shinkwin (1979) recovered Kavik points (Figure Vll-3,4), in rectangular house depression deposits that she dates to A.D. 1100 to A.D. 1500. At the Alaskan Atigun site, occupied ca. A.D. 1400 - 1800 Wilson (1978) shows a strong Kavik component (Figure VII-3,4). Derry (1975) proposed that there was northward movement of Athapaskans in Alaska, hypothesizing that the northern lobe of the White River Ash fall ca.1500-1900 BP (Clague et al. 1995) was a major factor in forcing new adaptive strategies in what had been Inuit territory. In other words, at a point in time just before the circa A.D. 800 White River Ash fall, Athapaskans with identifiable lithic types are demonstrable in archaeological components in Yukon and Central Alaska. FIGURE Figure VII-4. Kavik points from Klo-kut (middle), Atigun (bottom) and Dixthada (top) sites. Shinkwin also discusses the long-debated question of late microblade use, since she recovered microblades and microblade cores at Dixthada. She notes West's (1967) rejection of the late dates at Donnelly Ridge (A.D. 120 t 200, B-649; A.D. 160 t300, B\u00C2\u00ADMay 20, 2003-405 650), and cites Cook's (1969) demonstration of early and late microblade sites in Alaska. Shinkwin maintains that the upper layers of Cook's Healy Lake are like the lower layers of Dixthada. As we note below, microblades are present in a number of assemblages otherwise identified as Athapaskan and dating to the last 1500 years. Turning from Alaska and the Yukon, there is quite a geographic gap in Athapaskan archaeological research until we arrive at the central British Columbia plateau. Donahue (1977a:261) argued that Athapaskan cultures most likely were present on the northern Plateau at Tezli (Figure VI-2) since about 4500 BP. Although the archaeological evidence of this is weak, there are indications of an Athapaskan presence in the northern areas of the Interior Plateau longer than in the south (Dyen and Ablerle 1974:276). In fact it can be argued that the ancestors of the Tlingit/ Athapaskans have continually occupied the northern interior and northern Northwest Coast since circa 8500 BP, given that many archaeologists associate the North Coast Microblade Tradition and interior variants with Athapaskan ancestors (Borden 1979; Matson and Coupland 1995:82-96). Given the linguistic evidence presented above, the centrai British Columbia archaeological record should show clear linkages with the Yukon and Alaska Athapaskan materials in more recent times. This is in fact the case. At the protohistoric site of Ulgatcho, Donahue (1973) recovered five small stemmed points that easily fall into the Kavik range (Figure VII-3,5) (as well as 15 side-notched points, 7 with concave bases) but also one microblade core, and seven microblades, these from the upper 10 cm of deposit in house deposits that reached 20 cm maximum depth. Note that Donahue used 1 /16 in. screens, enhancing recovery. Chinlac (Figures Vl-2, VIl-5), another May 20, 2003-406 protohistoric Carrier village of large rectangular structures (Borden 1952) that is said to have been attacked by a Chilcotin war party in about A.D. 1745 (Morice 1906), contains a microblade core fragment (Magne and Matson 1984, 1987). The Chinlac assemblage also includes Kavik points (Figure Vll-5), as does another assemblage from the area (Borden 1952: Plate 1). FIGURE Figure Vll-5. Kavik points from Ulgatcho (top), Potlatch (middle) and Chinlac (bottom) sites. All of these sites are outside of the range of the P.P.T. and the culture history of these areas prior to 1500 years ago is very poorly understood. lt is clear that the P.P.T. once extended further north and northwest than it did at contact times (Matson 1988) as indicated by the remains of concentrated pithouse villages along salmon streams north of its ethnographic extent. This former P.P.T. territory is now occupied by Athapaskans (either Carrier or Chilcotin), indicating a reiatively recent Athapaskan expansion to the Blackwater / West Road river and further south. Fladmark (1973, 1976) and Montgomery (1978) have reported on the Punchaw Lake site near, but north of, the confluence of the Blackwater and Fraser rivers (Figure Vl-2), and from which we used Side-Notched points as some of our \"known\" Athapaskan points. The most recent prehistoric occupation at this site is represented by a large number (43) of rectangular \"house platforms\" of which two were excavated and several others tested. These appear to be similar in size to the Bear Lake iodges, but of a different structure, possibly indicating that these were not occupied in the winter. Four radiocarbon assays May 20, 2003-407 (Fladrnark 1976) date the house platform occupation to the last 600 years (560 + / - 75 B.P. [Gak 4905]; 290 + I - 70 B.P. [Gak 4906]; 250 + / - 70 B.P. [Gak 4908]; and 240 + / - 70 B.P [Gak 4909]) which gives a minimal age to the Athapaskan occupation of this formerP.P.T. area. At Anahim Lake further south and west, Wilmeth's Tschandu and Bes Teo houses at the Potlatch site (Wilmeth 1978) also have Kavik points (Figure Vll-5). In addition, these houses, which Wilmeth ascribes to Athapaskan occupations, also have relatively abundant microblades (n=71) as do some of the other sites there (Daniktco, n=l5, dated to A.O. 700; BezYaz, n=14). He also suggested that Suzchet House, with a hearth date of A.O. 1240, (S-502, Wilmeth 1978:150) is unlikely to be Chilcotin, and proposed that the initial Chilcotin occupation of the Anahim Lake region within the last 500-600 years (Wilmeth 1977: 101), or at !east at some time relatively soon after theSuzchet House occupation. Our analyses of iithic assemblage and projectiie point ethnic variability across a broad region of Interior British Columbia (Magne and Matson 1982, 1987; this volume) support Wilmeth's identifications and demonstrate that indeed the Suzchet House materials do not appear to be Athapaskan, but are similar to other P.P.T. assemblages from the Interior Plateau, including some from the Eagle Lake area. In sum Wilmeth (1978) has good evidence for his conclusion that the Chilcotin probably arrived at Anahim Lake soon after A.O. 1300. Hobler and Bedard (1992) report on two adjacent sites in the upper Bella Coola \u00EF\u00BF\u00BD\G1J\u00EF\u00BF\u00BDC 7I--4-valley at Qwliutl (west of Anahim Lake) that appear to contain both P.P.T. and I\ Athapaskan occupations. Lower Qwliutl (FcSm 6) has two excavated rectangular structures, which we interpret (as do Hobler and Bedard) as likely Athapaskan May 20, 2003-408 structures, p ossibly with two or three courses of logs around the base of the walls somewhat similar to what we found at the Lingfield Creek lodge, and reported elsewhere earlier in this volume. The structures appear to be about 6 x 6 m and 6 x 8 m in size. The two radiocarbon dates obtained and reported as A.O. 1520 + /- 70 (Beta 49136) and A.O. 1700 + /- 90 (Beta 49138), are of the same order as the Eagle Lake Phase at the Bear Lake site and consistent with Wilmeth's inferences that the Chilcotin probably arrive after A.O. 1300. Upper Qwliutl (also recorded as FcSm 6) is only 150m away and consists of a pithouse site, with seven relatively small and adjacent house pits. A single Upper Qwliutl date reported as A.O. 1170+ / - 80 (Beta 49137) from inferred burnt roof fall is also consistent with our findings about the presence of P.P.T. further inland to the east on the Chilcotin Plateau. As we have reported earlier in this volume, at Eagle Lake our earliest Athapaskan assemblage (Eagle Lake Phase) at Bear Lake dates to A.O. 1645-1660 and the Athapaskan occupation could have begun at A.O. 1500 (iatest inferred date for Eagle Lake P.P.T. is A.O. 1475). This pattern is consistent with both Wilmeth's findings, and those reported above for the upper Bella Cool a valley. Although the Eagle Lake dates are somewhat later than Wilmeth' s, that is to be expected once one takes into account the more southeastern geographic location of Eagle Lake and historic changes in Chilcotin territory. In our multi-assemblage comparison to determine Athapaskan and P.P.T. ethnic identity we used primarily either side-notched points or entire assemblages (Magne and Matson 1984, 1987; this volume) to identify ethnicity. The Athapaskan side-notched point style has a concave base, often with an extended spur, high base, and a long blade May 20, 2003-409 a b f g h d 0 2 e as found from Chinlac and Anahim Lake (Figure Vll-6). Using the complete assemblage analysis, the direct discriminant analysis showed that the distinguishing artifact types were side-notched points, Kavik points, miscellaneous points, corner-notched points, stemmed points and large bifaces. Wilmeth (1971 :57) argued that spurred scrapers with \"concave sides and the presence of graver spurs and (or) spoke shave concavities\" were Athapaskan markers at Potlatch House. Donahue also found these at Ulgatcho, and they are present in some Dismal River sites (Gunnerson 1960; Gulley 2000) but are not found at Eagle Lake. The consistency among Wilmeth's results, Fladmark's, Hobler's investigations and ours gives support to our inferences about the arrival of the Chilcotin. FIGURE Figure VII-6. \"Athapaskan\" Side-Notched points (from Chinlac); The Fish Lake site near Taseko Lakes (fyhurst 1993) is a late prehistoric (though undated) shallow rectangular depression which could be produced by an Athapaskan\u00C2\u00ADstyle lodge. It also contained microblades and microblade cores. The analysis of the three side-notched points from the Fish Lake sites, though, as reported in the previous chapter, classified two as definitely P.P.T. and one as dubious P.P.T. (Figure Vl-8). Similar to the Fish Lake site, the also undated Hallett Lake site (FkSe 22) west of Prince George, consists of a shallow rectangular lodge with a microblade core and microblades (Carlson 2000). In the Nicola Valley Wyatt (1972) reports on several sites that show some similarities with the Athapaskan material identified above. Perhaps the most likely site May 20, 2003-410 is the Jenny's Flat Site (EbRa2) (Wyatt 1972:132 -145) which appears to be very recent in date and includes at least two side-notched points with spurs (Wyatt 1972:Plate 26 h, t). This site includes numerous bifaces, remains of lots of lake fish, and shallow, round depressions, interpreted as resulting from mat lodges. Although Wyatt (1998) did not believe that he had found good evidence of the Nicola in his research, we think that this assemblage deserves a closer look, and that additional directed research in Nicola territory would be rewarded. While Wyatt (1971) had earlier reported some points as similar to Wilmeth's Kavik points, these are actualiy larger and from much earlier assemblages and so are unlikely links. Elsewhere on the southern B.C. Plateau, in the Highland Valley, Stryd and Lawhead (1983) excavated a probable temporary structure (EcRg 2) with microblades and cores dated to 1120 + / -170 BP (SFU 230), 1490 + / -150 BP (SFU219) and 1920 + /-210 B (SFU 231). Between Highland Vailey and Eagle Lake lie Sanger's (1970) Lochnore\u00C2\u00ADNesikep assemblages, with some of the best-known rnicroblade assemblages for western North America. These are often held as the standard for what non-Athapaskan, early- to-mid Holocene microblade, non-pithouse assen1blages should look like. They were also found in relatively deep housepit sites, not associated with the housepits themselves. It is often forgotten, however, that Sanger (1970: 123) argued that microblades continue up to 2000 BP. More recently, others (e.g., Stryd and Rousseau 1996) have argued that microblades do not occur in the last 4000 years, and we do have many assemblages that date back to 4000 years that do not have rnicroblades. There are many substantial, but undated microblade sites on the Interior Plateau, with a strong upland, short-term camp association (Ludowicz 1983; Greaves 1991; Pokotylo and May 20, 2003-411 Mitchell 1998:97; Richards and Rousseau 1987; Vanags 2000). The issue of circular reasoning with regards to microblades has been addressed explicitly by Helmer (1977), when he said (and we quote at length because 25 years later it is apparent that Helmer's cautions were justified): (A) problem stemming from Sanger's analysis is the assumption that microbladesceased to be of any significance after A.D. 1 (Sanger 1967: 196). This observation has resulted in the definition of a number of components in the north-central Interior based on the segregation of microblades and associated old-looking artifacts from otherwise apparently unmixed historic and proto-historic materials (cf. Wilmeth 1969; Donahue 1972, 1973, 1975). The implication here is that because n1icroblades appear early in the Lochnore/Nesikep Locality and because they are phased out ca. 2000 years ago at the same location, the temporal parameters for microblades in the north-central Interior must be the same. By accepting the hypothesized early occurrence and disappearance of microblades the culture history of the northern Interior can be made to conform with the sequence from Lochnore/Nesikep and thus to the accepted view of early population movements into the Interior of B. C. The rejection of this interpretation, on the other hand, would throw the entire sequence open to review .... there is ample evidence in the north-central Interior to contradict this interpretation ..... (Helmer 19776: 94). Although none of our Athapaskan assemblages at Eagle Lake contain microblades, one of our P.P.T. sites (Boyd Site, EkSa 32) did! Our point here is not that microblades are definitely part of the Northern Athapaskan toolkit in the last 2000 years, but that there are too many of these occurrences to casually dismiss them. At May 20, 2003-412 some level, microblades clearly do not mean Athapaskans, even in the Northwest. For example, microblades are present in Marpole and Locarno Phase components (3300-2000 BP; Matson and Coupland 1995) on the coast, yet these are clearly Salishan (and these cores h ave a different form). Microblade technology goes back to 8600 years along the northern B.C. Coast and is termed there the North Coast Microblade Tradition. Almost all see interior manifestations as being closely re]ated, with both being genetically related to the Denali complex of Alaska (Matson and Coupland 1995; Coupland 1996). Many archaeologists (Borden 1979; Magne and Fedje 2002; Matson and Coupland 1995) see this early microblade manifestation as likely being ancestral to Athapaskans, but this is a time depth about twice as long as conventionally allowed by lexicostatistics - or other historical linguistics techniques (Campbell 1998). This North Coast Microblade Tradition and related material continues up to 5000 BP, not showing any easy connection with the last two thousand years and the iater Athapaskan cultures under discussion here. Much of the early Pre-Pithouse Tradition material of the southern interior of B.C. has been lumped into the \"Nesikep Tradition\" which dates to 4000-6000 BP (Stryd and Rosseau 1996). As originally defined by Sanger (1970) it includes microblades, basal ground corner-notched atlatl projectile points, well-made scrapers, and the absence of housepits. When Matson inspected some of Wilmeth's Anahim collections in the late 1970's he was struck how these microblade-containing assemblages did not have this other associated material--basal ground corner-notched atlatl points and scrapers--of the Nesikep Tradition (and typically present at housepit sites that did have May 20, 2003-413 microblades). This appears to be true also of many of the other microblade assemblages discussed above (although not of the highland microblade assemblages reported by Pokotylo 1978; Vanags 2000; and Greaves 1991). So we do not believe that these \"late\" microblade manifestations can be explained away in bulk as unrecognized intermixtures of the \"Nesikep Tradition\", although some of them may be. The North Coast Microblade Tradition, though, typically has neither scrapers or bifaces of any sort (Matson and Coupland 1995; Coupland 1998). Typically, microblade and core tools are the major constituents of this culture. If something like this culture was present i n the areas under consideration -- and it certainly was in northern B.C., the Yukon and Alaska-- it would be very difficult to recognize when intermixed with later material (and certainly was not recognized at the Boyd site). In conclusion, these microblades are either an occasional constituent of Athapaskan culture in the last 1500 years, or result of the intermixture from a much older culture, heretofore not recognized in southern B.C. It is a sad commentary on the pre-P.P.T. culture history of the plateau that our ignorance should be so great, that more than 35 years after microblades were first widely recognized as an important part of early Plateau prehistory, their temporal occurrence remains so unclear. On the Pacific Coast Gould (1966; Matson and Coupland 1995:248-58) excavated a late prehistoric Tolowa (Athapaskan) site at the Point St. George in northern California. The assemblages excavated by Gould indicate a fully coastal-adapted culture with no traces of northern Athapaskan affinity with the exception, possibly, of some details of house and sweat lodge construction and the separation of cemeteries from dwelling areas. It seems that much of what might be relevant to Pacific Athapaskans in May 20, 2003-414 Oregon and Northern Califmnia remains unpublished in the form of contract reports, other than Gould (1966), Pettigrew (1980), and Connolly (1991 ). Connolly (1986, 1991) does provide some useful information from this grey literature, however, especially concerning various reports by Pettigrew that allude to possible Athapaskan materiai culture traits in prehistory, especially microblades and small contracting stem points. Two nearby sites, the Looney and Standley sites, in southern Oregon Athapaskan territory (Figure VII-3) are relevant here (Pettigrew 1980; Connolly 1986, 1991). Both sites had items which had been identified as microblades and microblade cores, although Connolly's (1991) reanalysis indicates that these may actually be scrapers and fortuitously produced flakes. The Looney site is considered to be too old (circa A.D. 1-500) to be of concern here, but the similar Standley site has 3 radiocarbon dates of lessthan 500 BP, but also includes dates of more than 2000 BP and is thought to include even older material (Connolly 1991 :39). Stemmed (\"Coquille\") points are common in older material, making identification of small variants as ''Kavik\" points problematical. On the question of the possible presence of microblades, it is clear that Connolly (1991:73-92) has made a serious and laudable attempt to resolve this question, but we are divided as to his success in providing convincing information that a microblade technology is not present. Probably, only a first hand inspection and comparison with accepted microblade technologies will resolve this issue. At this point, the Standley and Looney sites remain ambiguous in dating and as to whether they include microblades and Kavik points showing similarities with northern Athapaskan components. In an overview of southwestern Oregon and northern California, Connolly (1986:151 ff) reviews the Gunther Pattern, which he sees as '! ... without exception, May 20, 2003-415 associated with speakers of the intrusive Athapaskan or Algic ianguages\", a position also taken by Fredrickson (1984). The Gunther Pattern is thought to first make its appearance at about A.D. 900, at the Gunther Island site, in Humboldt Bay, California (Connolly 1986:160, 1991 :13) and continues through historic components of both Athapaskan and Algic peoples. We report slightly later, though, that this initial date is not very certain. The Gunther Pattern is seen as entirely coastal in adaptation, not as an interior derivation. Turning to the northern Plains, Kehoe (1973) claimed that Athapaskans manufactured Avonlea points at Gull Lake in Saskatchewan, and several others since have promoted this idea (e.g., Wilcox 1981, 1988). Perry (1980) states that Besant points are Athapaskan markers on the northern Plains. In Wyoming Frison (1973) suggested that the Wardell bison trap site may have been constructed and used by Athapaskans, while Reher and Frison (1980) make the same suggestion for late prehistoric layers of the Vore site in eastern Wyoming. Both of these cases are based on the presence of Avonlea points, and Vore is located in the Black Hills, a supposed home of the Kiowa\u00C2\u00ADApache. Wilcox (1981, 1988 but see Wilcox 2001) accepted Kehoe' s hypothesis, mainly, since to him, Avonlea points make their appearance in the right places at the right times (Wilcox 1988: 276) to account for southward movement of Athapaskans. Avonlea is a side-notched point style with narrow notches, low base height, and delicate flint\u00C2\u00ADknapping. Why would these derive from the north, when in the northern Plains the antecedent Pelican Lake and Besant series provide a good technological and cultural developmental basis? Furthermore, Morlan (1988) provides a thorough analysis of radiocarbon dates for Besant, Old Women's and Avonlea phases, demonstrating May 20, 2003-416 considerable overlap in the sequence. Peck and Ives (2001:185) support our view that Avonlea and Besant are not Athapaskan. Wilcox (1981, 1988) has presented a critical review of writings dealing with possible routes of migration for Apacheans, coming to the conclusion that a high plains route is most likely, and that an intermontane route is most unlikely. He aiso accepted Kehoe's assertion that Avonlea equals Athapaskan, and would have them present on the northern Plains 1500 years ago, living in the Black Hills of South Dakota until about A.D. 1450, at which time they began to move south and west to their present locations.Part of the reasoning here is that slightly later dates for Avonlea in Montana of A.D. 550 - 900 lead Wilcox to take this as evidence of a southward movement. More recently,though, Wilcox (2001) has reviewed his ideas and admitted that there is evidence that Avonlea is too old to be Athapaskan but continues to believe that his model is still defensible. Despite Perry's (1980) hypothesis of an Athapaskan Besant antecedent to an Athapaskan Avonlea, the Subarctic archaeological record exhibits no indication of Besant origins in the Yukon, or Northwest Territories, nor of small side-notched point origins in the north. At Lake Athabasca Wright (1975) considers a date of A.D. 690 + /-170 (Lab#??) from the Big Bay site (lgOo 1) to indicate the ''possibility\" that this and later strata represent the \"cultural development of Athapaskan Beaver.\" Workman (1978) considers the beginning of the Aishihik Phase at A.D. 400 to represent Athapaskan arrival at Fisherman Lake in the western Northwest Territories, Gordon (1975) places early Chipewyan culture at A.D. 500, at the beginning of late Thaltheilei, and similarly Clark (1981) equates the Spence River phase with Athapaskans, at A.D. May 20, 2003-417 500. These inferences are consistent with those based on linguistic and archaeologicalinformation previously reviewed for Alaska and the Yukon. For A vonlea itself, Morlan (1988) shows very clearly that A vonlea starts about A.D. 200 and continues to ca. A.D. 1200, across the northern Plains. Note that Kehoe(1973) in the Gull Lake report presented no material antecedents to A vonlea in the north. His argument appears to be based on the assumption that bison drives originated from caribou drive complexes in relatively recent times. Perry (1980) takes Kehoe' s argument a bit further, and uses examples of what he considers to be relatively insecure caribou-based subsistence strategies in mountainous environments to illustrate Athapaskan abilities to shift prey. This may be at first glance applicable only to far northern groups, until it is recognized that relict caribou herds still exist in more southerly areas, for example Jasper National Park (the heart of Beaver territory) and in Mount Revelstoke National Park in southeastern B.C., and that the Chilcotin hunted caribou in central British Columbia. Indeed, Perry points out that caribou extended as far south as Elk City, Idaho in the 20th century. Generally, Perry (1980) makes a good case for Apachean mountain adaptations and, as discussed below, this may be germane. But with respect to the bison drive-caribou drive angle, it is now well known that bison drives were in existence for thousands of years before Avonlea, and, in fact, go back to the Folsom complex (Bamforth 1988). In southern Alberta, the Cayley series of points (Peck and Ives 2001) is a quite variable small side-notched point variety, that could include points that equate to our Athapaskan side-notched style (see Peck and Ives 2001, Figure 9) and Kehoe makes reference to a point type not found at the Gull Lake site, the Buffalo Gap Single-Spur May 20, 2003-418 Variety of Plains Side-Notched, which may date to ca. A.O. 1600 that is also very similar. A possible explanation for this high degree of projectile point variability is the routine presence of other peoples among a Blackfoot majority (Peck and Ives 2001 :185), which seems feasible given the amount of movement that appears to have been taking piace on the Plains. As they note, such presences could have included Tsuu T'ina or Apachean peoples, among others. It is worth remembering at this stage that microblade cores are present in Beaver\u00C2\u00ADChipewyan Territory (Figure Vll-3), in northern Alberta, at Ft. Vermilion (Pyszczyk 1991), Peace Point (Stevenson 1986) and the Ft. McMurray area (LeBlanc and Ives 1986; Ronaghan (personal communication 2000). Microblade technology also occurs in a prehistoric and what appears to be a protohistoric context at the Peace Point site. At this superb and rare multiple strata site in northern Alberta (Figure VII-3), Stevenson (1986) found microblades and 2 microblade cores in Level 1, dated to ca. 2200 BP, and another microblade core in Level 16, in association with historic artifacts (1986: 84-87). Although Stevenson, like others, prefers caution in assigning microblade technology to anything in a late context, it is clear from the British Columbia and Alaska examples noted above that this is not an usual situation. Magne does not agree (Ives 1990) that there is an Avonlea point from Peace Point. There are two small side-notched points illustrated; however, one is broken and the illustration shows a reconstructed outline (Stevenson 1986: 83, 32) that may approach Avonlea. So although there is little if any substantial evidence that Avonlea assen-,blages relate directly to A thapaskan speakers, there is also little evidence that some were not made by Athapaskans. We do not have sufficient familiarity with northern plains pottery traditions to May 20, 2003-419 examine possible Athapaskan ceramic origins within them, but Avonlea phase assemblages often include pottery (Davis 1988). Perhaps because of the complexity, the obvious direct comparison of cernmics from Avonlea, Dismal River and other early ceramic cultures of the northern and central plains has not yet been undertaken. Nonetheless, the Tsuu Tina made pottery about which little is known. Jenness (1938:14) records:\" ... Sarcee .... abandoned their clay pots about the time they obtained horses.\" The Blackfoot were still making pots in 1772. lf this is the case, how did the Tsuu Tina and Blackfoot obtain their ceramic traditions? From Missouri horticulturists? From Puebloan interactions? Byrne (1973) presents a strong case for separate traditions of Algonquian and Siouan ceramics, given the single-component Cluny site east of Calgary, with its One Gun Phase ceramics, which are likely Hidatsa (Byrne 1973: 544). Cluny is a fortified village site probably occupied by Crow about A.O. 1700 who split from the Hidatsa to occupy the upper Yellowstone (Forbis 1977). Supporting this, Lewis Henry Morgan (1877) recorded a legend of a Hidatsa group that split to occupy areas of the South Saskatchewan R. For Athapaskans on the Canadian Plains, however, there is no archaeologically recognized ceramic tradition. Although little is known about Tsuu Tina archaeology, some work has been done on their reserve immediately adjacent to Calgary's western City limits. Helmer (1982) reports finding on the east side of the reserve, a possible pithouse site with 8 circular depressions, two large rectangular depressions and other features. Generally, though, archaeological evidence that can be clearly ascribed to Athapaskans is slim to nonexistent on the east side of the Canadian Rockies. The bits of evidence we have presented are only tantalizing. It may be that there is evidence that May 20, 2003-420 we just are not seeing for what it is, and maybe the Kavik points and microblade associations will stimulate reexamination of some existing collections, as well as a more sophisticated analysis of side-notched points, as suggested above. There are vast areas of Alberta, including the mountains themselves, that simply have not been subject to any survey. Certainly, the northwestern portion of the province, the home of the Beaver, has seen very little archaeological research, which is also true of the northeastern part of British Columbia. The most often mentioned complex of Plains sites identified as Athapaskan belongs to the Dismal River Aspect investigated by Gunnerson (1960) and Gunnerson and Gunnerson (1971). The Dismal River Aspect occurs in western Nebraska and northwestern Kansas, and is claimed by Gunnerson to represent Kiowa Apache, although others would claim the sites represent Lipan Apache, and others yet (e.g., Schlesier 1972) would have it be a later phase of the Fremont Culture of the Great Basin as represented by Comanche, a branch of the Numic Shoshone. As we discuss below, this last idea is almost certainly incorrect. Schlesier (1972) also considers the northern aspect of Dismal River Aspect to equal Nebraska Sand Hills Athapaskans or Dismal River proper, while the southern aspect of Plains Apache Tradition in western Kansas, is an Athapaskan response to Pueblo farming. Schlesier links Kiowa and Kiowa-Apache (whom he calls Gattaka as per one of Mooney's appellations) movement onto the Plains to the Fremont movement (withdrawal) from western Colorado on to the Plains. Baugh and Eddy (1987) claim that Athapaskan ceramics were influenced by neighbouring groups, observing that Dismal River Gray Ware has many similarities with Plains Caddoan ceramics. On the other hand, Wedel (1986:134-151) reports on abundant May 20, 2003-421 evidence of contact with the Pueblos in the Dismal River Aspect, including a close resemblance between the pottery ware with Rio Grande utility pottery, although he does admit some of the most eastern material does show some surface finishing traits similar to historic Plains villages (1986:144). Wedel (1986) points out that the Dismal River Aspect has a relatively late date (around A.O. 1700) after at least 150 years of Pueblo contact and probably represents the ancestors of the Kiowa-Apache. Thus the Dismal River Aspect can not represent the Apacheans on the way south. An examination of the Dismal River Aspect by Gulley (2000) has come up with rather different inferences. She examined the historic records, the archaeological field records for several sites, and the ceramics from one of the Dismal River Aspect type sites, the Lovitt site, concluding that the Dismal River Aspect's Athapaskan attribution is highly questionable. On the other hand, the more recent assignment of the \"Tierra Blanca complex\" to both the Querechos and Apacheans (Habicht-Mauche 1992) appears to be very convincing, and is much closer in time to the arrival of the Apachean in the Southwest. In sum, the archaeological situation in the middle areas of Athapaskan occupations includes decent evidence for material culture continuities from central British Columbia to the Yukon and Alaska. On the eastern side of the Rockies, the picture is far less clear, with few continuities evident in archaeological materials from the Subarctic to the Plains, and with increasing confusion about Athapaskan archaeology until one gets to the southern Plains circa A.O. 1450. Furthermore, much of the identified material between the two areas (e.g., Dismal River Aspect) is not closely connected to the inferred Apachean migration. May 20, 2003-422 Pacific Ocean N 1 0 200 400 0 200 \"'-,,,, --' \ \u00E2\u0080\u00A2. -- \u00E2\u0080\u00A2 -J,--P-u_e_b_lo_l_l!_-_-_-__Q\" ,.,,,,- \u00E2\u0080\u00A2'\u00E2\u0080\u00A2 ; C ,\u00C2\u00B7 \u00E2\u0080\u00A2 olorado Plateau _,,-\u00C2\u00B7 \ , Occupation ........ ______ ,,,/---;,,,-!.-______ \u00EF\u00BF\u00BD i/' -.-. \u00E2\u0080\u00A2 Pueblo IVOccupation !-'Basin and Range \ . - .._ _ \ ',. . \ \u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\ 400ml , Great : : ! \u00EF\u00BF\u00BD i!!Salt Lake I ; ; .. l .. :\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD\u00EF\u00BF\u00BD:..\u00EF\u00BF\u00BD-----------: ... . :: ... .. : . . . . . . . . . .-\u00C2\u00B7- -:_ / / ' .\u00E2\u0080\u00A2 \u00E2\u0080\u00A2 \u00EF\u00BF\u00BD11AiVt1r r: \u00EF\u00BF\u00BD! I\u00E2\u0080\u00A2 I .I . I .\ . \ . \ .\ \u00E2\u0080\u00A2I -------\u00EF\u00BF\u00BD\ _____ , __ _l;). Las Vegas \ Et Paso \u00E2\u0080\u00A2- --'. .... . ' . ' . \ \ \u00C2\u00B7--The Situation On The Receiving Ends We now turn to the conditions on the receiving locations which must have allowed the migrating Athapaskans to settle there. The conditions are both better known and more obviously amenable for the Apacheans than the Pacific Athapaskans. The failure of agriculture (Renfrew 2000) of the northern Anasazi (Ancestral Pueblo peoples) area and western plains facilitated the Athapaskan migration into the Southwest. Southwestern archaeology has long puzzled over the abandonment of the San Juan river region (Figure VII-7) in Pueblo Ill times (i.e., AD. 1250-1300). Beginning more than 2000 years ago (Matson 1991, 2002) maize agriculturaiists belonging to the Anasazi tradition occupied the area north of the San Juan River, including some of the densest and best known Anasazi sites (e.g., Mesa Verde). Yet by A.O. 1300 this area was completely abandoned; the Anasazi in Pueblo IV times (A.D. 1300-1450) not only disappeared from the area near the San Juan, but also concentrated into a few areas that had large towns (Figure Vll-7). The result was that most of the previously occupied area was no longer used for agriculture. lt was these areas vacated by the Anasazi, that the Apacheans west of the Rockies re-occupied. Similarly, the Fremont horticultural culture, located to the north and west of the Anasazi, collapsed and was replaced by Numa speakers (a Uta-Aztecan language group, which includes Shoshone) in the eastern Great Basin and neighboring portions of the Colorado Plateau (Lipe 1995). FIGURE May 20, 2003-423 Figure VII-7. Changes in Anasazi Occupation in Pill and PIV Periods. In the volume edited by Cameron (1995), Lipe (1995) argues convincingly that at least 10,000 people were found north of the San Juan alone in A.D. 1250 compared with estimates of 30,000 to 60,000 for the totai Pueblo population at the time of Spanish contact. Lipe (1995) presents compelling evidence that at about A.D. 1250 this northern Anasazi region began to be abandoned and was completely deserted by 1281. In the succeeding 150 years the area occupied by the Pueblo Indians shrank to that seen today (Figure VII-7). Deteriorating environmental conditions were clearly an important factor (Ahlstrom, Van West, and Dean 1995) in this abandonment. In this way space was made available which was later occupied by the Athapaskan speaking Navajo and Apache peoples. Although there is disagreement for the earliest date of Athapaskans in the Southwest (Schaafsma 1996, 2002; Sesler et aL 2000; Towner 1996, 2000), all substantiated claims are post A.D. 1400. We discussed earlier the identification of the Tierra Blanca Complex (A.O. 1450-1600) on the Southern Plains to the east of the Rockies as Apachean. For the specifically ancestral Navajo, the earliest date currently under discussion is an unpublished structural tree-ring date of A.D. 1451 in northwestern New Mexico near the San Juan river (Towner 2000; Wilcox 2001) which is also apparently reported as \"16th century\" (Sesler et aL 2000:162). This as-yet\u00C2\u00ADunpublished date is not in question, but its identification as \"Athapaskan\" may be (e.g. Schaafsma 1996) and it is clear that full publication is needed before it can be fully evaluated. There are numerous tree-ring dates demonstrating that the Navajo were in this area well before A.D. 1700 (Sesler et aL 2000), the same area abandoned by the May 20, 2003-424 Pacific Ocean Anasazi by A.O. 1300. Questions have been often raised about the conditions that the would allow the Athapaskans to travel the distances involved between Subarctic and the Southwest within a short period. These can be answered by the environmental consistency along the east flank of the central and southern Rocky Mountains and the collapse of the Upper Republican tradition. To large mammal hunters the eastern slopes of the Rockies are very much alike from the Peace River to Las Vegas, New Mexico ( Figure VII-8). An adaptation at one end works at the other, as bison, along with elk and deer are present all the way down, with caribou present as far south as Idaho. We assume (as we present below) that the proto-Apacheans had such a hunting adaptation on the northern/ central Plains. FIGURE Figure VII-8. Rocky Mountain route for Athapaskans. How and why did the Proto-Apacheans move south? Around A.D. 1300 the collapse of the Upper Republican culture left the eastern slopes of the southern Rocky Mountains uninhabited. The Upper Republican culture was a pithouse, maize using culture located in the river valieys flowing east from the Rocky Mountains. Slightly before A.D. 1300, the range of this culture contracted, probably moving downstream, towards the east ( Wedel 1986, 2001; Wood 1967; Wood 1998), presumably because of the same environmental changes that resulted in the abandonment of the northern Southwest by the Anasazi. Although the Upper Republican is not as well dated as May 20, 2003-425 Pueblo III, Wedel (2001) dates it to A.D. 1100-1250, the same period. The Upper Republican adaptation, like other plains agriculturalists, also relied on large game hunting (Wood 1967). In fact, the western variant (in eastern Colorado) has been claimed to be totally hunting and gathering (Gunnerson 2001 ). It dates to the same time, and collapsed prior to A.D. 1300 (Wood 1967; Gunnerson 2001) probably indicating a n ecessary symbiotic relationship with the Upper Republican sites in Nebraska and Kansas. Prior to the introduction of the horse in the late 1600s, the hunting area of the Plains agriculturalists must have been limited. Thus the A.D. 1300 contraction left the eastern flanks of the southern Rockies un-used which encouraged the proto-Apacheans to travel down this corridor, entering the Southwest (Figure Vll-8). Thus we have a three stage migration. First, moving onto the northern or north\u00C2\u00ADcentral piains and becoming adapted to that situation. Then positioning on the northern portion of the Central Plains so when the late thirteenth century agricultural contraction occurred the Proto-Apacheans were adjacent to the vacated area and entered it. Then at the south end of this area some of this group moved west and became agricultural, while others remained on the plains -- at least for a time (Querechos) and some (Kiowa-Apache) stayed there. The Numic expansion into the Great Basin, briefly referred to above, (Figure Vll-8), probably resulted from the collapse of the agricultural Fremont cultures at the same time, for the same reasons (Lipe 1995). Schlesier (1972) as reviewed above, presents an argument for identifying the Fremont as ancestors of the Kiowa-Apache. Hill (2002) has also presented an argument for the Numa having been present in the Great Basin for May 20, 2003-426 thousands of years. The perishables, in particular, the \"utter dissimilarity\" (Adovasio and Fedler 1994:121) of these between the Numa and the Fremont, show this latter can not have been the case for the eastern Fremont (which must be the Fremont variant Schesier is referring to). Recent Mitochrondrial DNA evidence (O'Rourke et al. 2000) also demonstrates the discontinuity between Fremont and Numic. Furthermore, there is good evidence for the eastern Fremont having been present since A.O. 500 (Geib 1996), which is very difficult to reconcile with the similarity of the Kiowa-Apache with the other Apacheans and northern Athapaskans. The Numic groups are all very similar in terms of language, and it is an open question whether Navajo and Apache have differentiated enough to be considered separate languages. The Numic and Apachean language distributions have the exact pattern expected for a recent agricultural driven migration-- a large group of people with similar language(s) and similar cultures spread through a large area which resulted from an agricultural collapse and ended up with a migration by hunter\u00C2\u00ADgatherers replacing agriculturalists. An obvious question that arrives from this perspective is why were the Navajo able to adapt agriculture successfully in the Southwest in areas abandoned by the Anasazi? We see two reasons (Matson 2002). First, arable land (which in the Southwest is defined by deep soils and either the presence of enough rainfall or floodwater) is very discontinuously distributed. Second, during Pueblo IV (A.O. 1281-1450) the population was aggregated into a few large communities. Locations that would support these larger communities were few and far between further isolating them. LeBlanc (1998, 1999) has demonstrated that conflict had an important role in establishing these larger May 20, 2003-427 communities. Once communities were aggregated for defence, it was not possible for the Anasazi to go back to the eariier pattern of smaller, dispersed communities. It was this empty, abandoned niche that was colonized by the Apacheans. Turning to the situation in southern Oregon and northern California, the archaeology is less well developed and our expertise is both much more limited (no first hand fieldwork) and dated, so our view is both less clear and less reliable. In general terms, the critical time of Athapaskan arrival is about the same as for the Apacheans, between 500 and 1000 years ago. The archaeology is best known for the Northwest Coast portion of Pacific Athapaskan territory (Matson and Coupland 1995; Gould 1966; Fredrickson 1984) although not really well known. Matson and Coupland (1995:259) suggest that \"it may be that the fully Developed Northwest Coast Pattern was spreading down the Oregon coast when it was truncated by extensive contact with Europeans\" based on the presence of iarge-multifamily houses occurring only on the northern Oregon coast. Here we review the evidence for the dating of the small house adaptation to the south and how this adaptation may have allowed the Athapaskan penetration. This adaptation, which may have included wood working and salmon storage, has been connected to the ''Gunther Pattern'', already introduced above (Fredrickson 1984:483). This culture is identified by the Gunther Island barbed point, based on excavations on Gunther Island, which is located in Humboidt Bay, a Wiyot (Algic language) village site. ln a reconstruction developed by Whistler (1979, cited in Fredrickson 1984 and Tasa 1997:30), the Gunther Island pattern is brought in by the Algic speaking Wiyot, at about A.D. 900, followed by the Yurok (also Algic speakers) at May 20, 2003-428 about A.D. 1100. This is followed by the various Athapaskan groups moving to the coast at A.D. 1300. This reconstruction would put the Athapaskans in the area soon after the first development of the stored salmon economy, which allows for a much increased population density, up to six times more than previous adaptations (Croes and Hackenberger 1988). Such a population increase is observed when the stored saimon economy developed in the Gulf of Georgia (Matson and Coupland 1995:156). Where such a transformation is occurring, the effective carrying capacity is higher than the population density (at least to the users of the new adaptation) which would allow for the insertion of newcomers such as the Athapaskans. Furthermore, this new adaptation also allows larger local settlements given greater potential political power than neighbors with both a lower population density and a more dispersed settlement pattern. If such an argument is to be made, it is important that the coastal winter village pattern associated with the stored salmon economy be shown not to be much older than the purported arrival of the Pacific Athapaskans at the California/Oregon border. A good summary of dated coastal rectangular houses is in Lyman (1991 :127 and elsewhere). Although the sample is small, all well-dated houses appear to be less than 500 years old,. There is, however, one Lone Ranch Creek site (Berreman 1944) house which Lyman lists as possibly 1000 years old, but it is not directly dated. A close look at the Lone Ranch site, though, shows no indication of substantial age, as it is almost entirely a Gunther Pattern site, although the house in question, a rectangular house, was found at a greater depth than the other four which are given more recent dates. Interestingly enough, the Lone Ranch Creek site is weil within historic Chetco territory, May 20, 2003-429 one of the Pacific Athapaskans (Berreman 1944). As no other house of this sort in Oregon or northern California (Fredrickson 1984) is known to be oider than 500 years, there is not sufficient evidence to indicate a substantially earlier date. Part of the expectation of a significantly earlier date is based on its relatively great depth from surface (7 ft) but other excavations of Gunther Pattern sites in northern California show that greater depths have deveioped in much less than 1000 years (Fredrickson 1984:488). Perhaps the oldest radiocarbon date at the time of writing for a house is the approximately A.D. 1600 corrected and calibrated date of Beta 93965 for House 2 at the Pistol River site in Oregon (Erlandson et al. 1997; Matson and Coupland 1995:255-6; Tveskov 2000:169). This date is from a clam shell which means it is subject to more than the usual questions. Erlandson et al. (1997) present an argument that the five other dates (when corrected and calibrated) from this site all have the highest probability intercepts in the A.D. 1600s indicating that this site was predominantly occupied after A.D.1600.In summary, the stored salmon economy-coastal winter village pattern does not have good dates of more than 500 or so years old in either northern California or southern Oregon. Given that it is unlikely with a small range of dated houses/sites that the first occurrence of this adaptation has been dated, it is likely at least a little older than these dates indicate. The arrival of the Athapaskans and the beginnings of this adaptation are at approximately the same time. Although we have cited others crediting this adaptation to the Algic speakers, slightly before the arrival of the Athapaskans, we are not clear why this should be so. Connolly (1986) also questions this association. Given the uncertainties in dating, would not the Athapaskans, passing May 20, 2003-430 through areas where winter settlements based on stored salmon had been present for thousands of years (Both the P.P.T. in the Southern B.C. interior and the P.P.T. variants in interior Washington and northern Oregon) be likely contributors to this new adaptation? This scheme proposes an interior migration; there is an alternative idea that Athapaskans came down the coast (Connolly 1986; Tasa 1997). We see this as less likely, as the northern Oregon coast would have been thickly populated at this time with the stored salmon adaptation. Furthermore, Tasa (1997) reports that there is evidence supporting an Athapaskan spread north along the coast in recent times, something hard to visualize if they had just recently come down the coast, but this fits with the above model of an interior route with a descent to the coast south of the existing winter village coastal pattern. Additional evidence, though, is needed to discriminate between these alternatives. A Revised Model Of Athapaskan Migrations What we propose is that an early southward movement of Athapaskans took place along the east side of the Coast Range and later along the east side of the central to southern Rocky Mountains. Beginning with the later White River eruption at A.D. 800 covering an area of about 800 km x 800 km (Figure Vll-9), Athapaskan peoples displaced outwards caused ripples resulting in Athapaskan peoples far from the ash to move aswell. Recent investigations involving trees buried in ash has yielded a more precise mean estimate of the younger White River eruption of 1147 calendar years Mav 20, 2003-431 .I before 1950, or A.D. 803 (Clague et al. 1995) dating the beginning of this process. We propose that Athapaskan peoples moved southward reaching south coastal Oregon by A.D. 1200-1400 and the US Southwest by A.D. 1450 (Figure VII-10).FIGURE Figure VIl-9. White River Ash Fall area, circa A.D. 800. Originating in the mountainous areas of southern Yukon, northern British Columbia and western Northwest Territories, Plateau and mid-Plains Athapaskan groups essentially continued their basic adaptations into regions which were ecologically quite different from the northern boreal forest. The general boreal forest environment, though, extends from the northern limit of the tree line down to the northern edges of the P.P.T. in southern B.C. It is not an accident that this environment is occupied exclusively by Athapaskan speakers, and that they form dialect chains, as there are no clear breaks in subsistence practises in this range, despite the presence of mountain ranges and rivers. It was surprising to us to find out how similar the environments of the Chilcotin and Tutchone were, for instance, extending to preferred site locations. Even when the different environments were encountered, on the Plateau or mid\u00C2\u00ADPlains, life in each region could have continued with many aspects of the basal adaptation, as per Perry's (1980) hypothesis of mountain adaptations. Living in the mountains and foothills requires highly mobile and very seasonal settlement patterns as in the boreal forest. In both the Plateau and the Plains, use of reliable major local food May 20, 2003-432 resources was not a problem -- on the Plateau this was salmon, on the Plains, bison -- as both these resources also occurred in the basal Athapaskan area, and so were not new resources. Salmon are widely distributed in the Pacific subarctic drainage and Athapaskans use them when present, but they usually do not regularly occur in large enough numbers to become a mainstay of the diet. Winter whitefish lakes are usually more important, showing that fishing is an important part of the basal Athapaskan adaptation. Bison live in the boreal forest and the mountains as well as on the Plains proper. There is ample evidence of this in the Canadian Rockies and today the largest bison herd is in Wood Buffalo National Park in northern Alberta. Once these two key resources had become the mainstay of the revised adaptations, no large significant changes were needed until the Southwest and the Pacific Coast were reached. Movements through both Plateau and mid-Plains areas would have required social means of forming alliances with neighbouring groups, and, as Ives (1990) has argued, such social means of adaptation is a basic ingredient of Athapaskan society throughout the Subarctic. Ives (1990) developed a model to explain Athapaskan kinship and residence rules, that he describes as \"Dravidian\", offering plausible social mechanisms to explain how Athapaskans in various areas adapt so readily to new social and environmental situations. Using in part the kinds of analyses that he has developed, it would seem profitable to model generational periods of residential movement and sedentary residence establishment, in reiation to major subsistence resource patterns and co-requisite logistical needs at particular locations, in tune with particular resources. What we see in the protohistoric period in central subarctic through early historic times, are additional movements, tied directly to west-flowing fur Mav 20, 2003-433 _, trade pressures, but which mirror the earlier eastward movements and which likely used similar social processes. One of the difficulties here is trying to reason why the Pacific Coast groups appear to have been in place for a longer time, and in a more secure fashion, than the southern Plains groups. It may simply have been a matter of group size. We can get some idea of the size of the group moving south by looking at estimates of the precontact group sizes of Northern Athapaskans. Looking at the first twelve estimates given for various Athapaskan groups in Volume 6, Subarctic of the Handbook of North American Indians (Helm, 1981) we find populations from 200 to 3600 with the two largest (Carrier and Chipewyan) being also the ones with the largest territories. The next largest are several with estimates of 1000. This is probably a good upper limit, as it is hard to imagine the population of a very large territory such as Carrier or Chipewyan all moving at once. lt is less clear what a likely lower limit would be, but a group too small would either get absorbed or not have enough poiitical power to maintain itself for the distances and time involved; perhaps 300 or 400 is an appropriate lower limit. It may be that the size of the group that reached the Southwest was larger, and, since at that time the Pueblo Indians were concentrated in a few large towns, that not a lot of continuous contact occurred. In contrast a smaller group may have made it to the Pacific, and had to insert itself among an already densely populated area, and so was exposed to a higher amount of acculturation. On the other hand, as indicated earlier, we find the evidence for the Pacific Athapaskans having migrated earlier than the Apacheans not all that definitive, given the differences in the settings. To review, the average of 3-6 fewer shared cognates (out Mav 20, 2003-434 ., of 100) is not very impressive, given the different settings, with more intensive and continuous contact along the coast. And the differences are: 1) a possibly larger, less divided group, moving into a sparsely populated area along the western Plains and northern Southwest, versus, 2) one squeezing into an area with a much higher population density and more sedentary life style in the Pacific Northwest. The idea that these two migrations occurred at the same time still appears feasible. Another more direct reason for the migration may have to do with environmental episodes. It has recently been argued that the Medieval Warm Period, from A.D. 800 to A.D. 1200, was a world-wide event (Broecker 2001 ). This, in its early phases, could have magnified the effects of the White River Ash eruption of A.D. 800. Then, with the environmentally related Pueblo Ill/Upper Republican collapse in A.D. 1250-1300, Athapaskans found additional niches to occupy. Spielmann (1983) notes a sudden increase in Plains - Pueblo exchange at ca. A.D. 1450, indicating the development of Plains/Pueblo interdependence. Spielmann's model of Plains hunter\u00C2\u00ADgatherers acting as go-betweens Plains horticulturists and Pueblo villages, may fit the Kiowa-Apache case, who may have been in contact with the Upper Missouri Mandan and Siouan horticulturists of the southern Canadian plains. Furthermore, we have already indicated that the two variants of the Upper Republican Phase appeared to have this sort of relationship earlier. Perry (1980) also argued that climate warming at about A.D. 1000 caused southward migrations of major game animals, leading in turn to southward movements of some Athapaskans. Other writers have commented on an Athapaskan propensity to take advantage of a vacancy. Moratto (1984:570) notes for Pacific groups: \"Especially in the north, the May 20, 2003-435 Athapaskans seem to have occupied lands that were under utilized previously, and at least some of them appear to have advanced by settiing areas only sparsely inhabited.\" To paraphrase Anthony (1990), migration means never going where you haven't been before. In the realm of group movements, migrations or colonizations, then, this may be taken as meaning that a group as a whole has ways of knowing what lies ahead, that no mass movement or large-scale movements take place unless scouting, warfare, inter-marriage or exchange has provided contact with neighbouring groups. Thus, the Kiowa movements to Tsuu T'ina territory and Tsuu T'ina contacts southward make sense. The seeming paradox of apparent southern and Plains Athapaskan propensity for raiding, warfare and other forms of hostility, in the face of a widespread dispersal, has been addressed with Alaskan Unit groups. Burch and Correa! (1972) demonstrate that movements from one region to another could only occur in northern Alaska if groups had already established alliances in the area to which they were about to move. Alliance took the form of trade and intermarriage. They show that kinship alliances often existed hand-in-hand with \"intense mutual hostility\". Having found in independent studies that alliances and conflicts vary directly, not inversely as might be expected, they state (Burch and Correll 1972) that \"groups that fight one another stick together.\" We need to find ways to recognize sites that the Athapaskans left on their way south, particularly on the Plains-Rocky Mountain border, and Ives (1990: 322- 328) has provided a start on thinking about what archaeological sites might look like under conditions of group growth versus group alliance. He has begun to apply this approach to projectile point styles (Peck and Ives 2001 ). May 20, 2003-436 Given t hat cooperative efforts at bison hunting may be more necessary for group survival than cooperative salmon fishing, it is perhaps not surprising that Apachean divergence lagged behind that of the Pacific Coast. The Athapaskans east and west of the Rockies were adept at adopting their neighbour's habits in many respects other than language. The Pacific Coast Athapaskans that were on the coast were coastal in all respects, and certainly the ethnographic evidence is that Tsuu Tina in most aspects were indistinguishable from Blackfoot, and the same holds for Kiowa-Apache and Kiowa. The more northern Chilcotin group at Anahim Lake interacted greatly with the Salishan Bella Coola and other Chilcotin to the east interacted intensively with Canyon Shuswap. To summarize this revised model, which shares some aspects with Wilmeth's (1979), we are proposing that the most likely scenario is for a close relative of the Carrier and the Chipewyan to have spread across the northernmost Rocky Mountains, at the ends of ripples radiating from the White River Ash fall population displacement (Figure VII-10.2, 10.3). Under this pressure we see two offshoots heading south through thePlateau, with the Pacific one possibly, but not necessarily, leaving slightly earlier (Figure VII-10.4). The Pacific one picks up, if it did not already have, the salmon storage economy, and moves down the east side of the coastal mountains. At this time (circa 1000 years ago) the stored salmon, winter village coastal adaptation was aiready established along the northern Oregon coast (Matson and Coupland 1995) but not along the southern Oregon and northern California coast. When the Pacific Athapaskans reached this area (Figure VII-10.5), they also turned to the coast, as well as staying on the east side of the coast range, and participated in the establishn1ent of the stored May 20, 2003-437 salmon, winter village coastai adaptation, as well as the more dispersed interior stored salmon, acorn village economy. The adoption of this subsistence pattern increased the carrying capacity and so created the empty niche for the Pacific Athapaskans. FIGURE Figure VIl-10. Athapaskan migrations routes. 1) Circa 500 B.C.; 2) Circa 2000 BP, North Lobe White River Ash; 3) A.D. 800, Initial Migration; 4) A.O. 1000-1200 B.C. Plateau Migration. 5) A.O. 1400 Initial Apachaen and Pacific Athapaskan arrivals; 6) Ethnohistoric distributions. The ancestor of the Apacheans then continued the reaction to the ripple by spreading from the Plateau along the east side of the southern Rocky Mountains (Figure VII-10.4, 10.5), probably initially maintaining contact with their iinguistic relatives andconcentrating on bison hunting. They likely extended from the upper Fraser River through the Okanagan region to northern Montana, at about A.O. 1250. At this time the collapse of the Upper Republican Phase (Figure Yll-8) meant the niche to the south was empty and the Apacheans moved down to central New Mexico, at the end of the Rockies. They then picked up parts of the Anasazi-Pueblo adaptation, and entered empty parts of the northern Southwest, perhaps at the time of the collapse of the Galisteo basin pueblos. The Apacheans then split into three different adaptations: with the Kiowa-Apacheans continuing the plains adaptation (Figure VII-6), soon to be greatly changed with the introduction of the horse; the Navajo, a more sedentary agricultural and, later, pastoral adaptation in the northern Southwest; and the Apaches, a less sedentary one to the south. Overall we believe we have demonstrated that there are clear prehistoric links of May 20, 2003-438 Athapaskans to small contracting stemmed points, rectangular houses, and possibly microblades and on the Plateau and coastal Oregon and Caiifornia. These linkages to the north are stronger than any connected to A vonlea, and while there are clear ethnohistoric links among northern Plains Athapaskans, these appear to be quite late in time. Environmental conditions may have augmented migration processes, both at the early and late stages of the migration. Counter to Greaves (1998: 662) and others, there are sites with microblades that may be Athapaskan. Just as microblades are part of, but are not always, associated with Sanger's Nesikep Tradition, we propose that Kavik points, rectangular, gabled structures, and possibly microblades are part of, but not always, associated with the western Athapaskans. The historic period moves east of the Rockies are a reflection of what transpired earlier in that people had moved eastward and were now moving west and southwest. Fur trade pressures from the east in the subarctic in the 1700s reinforced long-range contact strategies that had been in place for some time. In the far western Plateau area, the areas of the Ulgatcho and protohistoric Chilcotin, the dates of European contact were quite late (ca. A.O. 1850), and by that time the Plateau - Coastal Athapaskan contacts were gone as were those of the Plateau-southern Rockies. The fur trade activities on the Plateau caused secondary southward and eastward movements in the early 1800s. On the northwestern Plains long-range movements were taking place among many people during the post-contact period, so that a relatively new relationship became possible between the Subarctic Canadian Athapaskan and northern US Plains Algonquian and Kiowa groups. In our view, the Tsuu Tina's location is not indicative of the Apacheans' route to the Southwest; it is rather an artifact of the historic May 20, 2003-439 fur trade. The principal Athapaskan group of interest to us, the Chilcotin, were first part of the interior Plateau movement (including the Southern Carrier and Nicola) as the tail of the same wave that earlier sent the Pacific Athapaskans and Apacheans southward. By this time, movement into marginal or abandoned areas had become a part of these Athapaskans' culture, as demonstrated by the way the Apacheans and Pacific Athapaskans were able to infiltrate into their current territories. The historic Chilcotin abandonment of the Anahim Lake area and shift toward the Fraser River are just the latest such moves. Although the Chilcotin have occupied the Eagle Lake area for about 400 years, during that time they have expanded east and south, while the boundaries of their territory to the north and south have retracted. It is clear (Matson 1988) that the P.P.T. territory also expanded further north and northwest in the past. The motivations behind these migrations remain unclear. Although the timing of the two White River eruptions does coincide with both the earlier Athapaskan movement towards the northeast, and then the later one to the east and south, other movements, such as the Chilcotin, occurred without anv known such event. The White _, River eruptions remain the leading hypothetical explanations partly because of the lack of plausible alternatives. In order to better document these moven1ents, however caused, we require appropriately dated archaeological assemblages in southern British Columbia near the US border of Washington State, and we need to identify Athapaskan material culture in May 20, 2003-440 Alberta and northern Montana. There is a need to model more explicitly how the Athapaskan migration could have taken place, in terms of processes involved, perhaps much as Duff (1998) has attempted for the US Southwest. By outlining what those processual aspects are, in terms of the scale of the migration, how the migration decision is actually made, how the destination is selected, how the migrating groups are structured, and how identity is maintained by those groups, we will understand more what to look for in material culture. May 20, 2003-441 Chapter VIII CONCLUSIONS. Introduction The primary goal of the Eagle Lake Project was to identify the entrance of the Chilcotin into the southern Chilcotin region. The Eagle Lake locality was chosen after a reconnaissance to identify an area within the pre-1850 Chilcotin territory that had a similar environment to that of the previous P.P.T. investigations at the Mouth of the Chilcotin to the east. The Eagle Lake area had the advantages of being adjacent to the important salmon stream, the Chilko River, of having relatively open areas which are easy to survey, and of being not-too-far from Anahim Lake where Wilmeth (1978) was conducting his research. We planned to identify the Athapaskan entrance using the Parallel Direct Historic Approach, by comparing two sequences; the one from the MOC within P.P.T. territory where the Athapaskan migration did not occur, and Eagle Lake, where the migration could be recognized by the divergence from the MOC sequence. Because of the presence of numerous concentrated pithouse villages with Kam loops side-notched points-- which together signified a recent P.P.T. occupation-- in the Eagle Lake area and ethnographic inferences (Lane 1954) we expected that the Athapaskan migration occurred relatively recently. Thus we were interested in the n-,ost recent parts of the regional sequences. In parallel with developing and comparing the sequences we gathered May 20, 2003-442 information from ethnographic and archaeological sources about potential discriminating traits between P.P.T. and Athapaskan traditions. These included house structures, settlement patterns, and artifact types. In addition we developed procedures that would allow us to use other aspects of the material culture to identify ethnicity, primarily by gaining access to collections of known P.P.T. and Athapaskan collections so that a standard set of analytical procedures could be used to discover more subtle material culture differences, both of single artifact types and complete assemblages. Field Investigations As part of the parallel approach, the initial 1979 quad rat survey followed the procedures used at the Mouth of the Chilcotin. Related surveys using similar procedures were later carried out at the adjacent alpine areas of Potato Mountain and further afield at Taseko Lakes. Because we used similar survey methods, we were able to make detailed comparisons with both the MOC and Upper Hat Creek (Pokotylo 1978, 1981) projects. Our quadrat survey resulted in an unbiased sample of the archaeological remains in the area, for settlement pattern analyses, and for evaluation of possible Athapaskan sites. A secondary survey along the Chilko River was initiated primarily to locate possible pre-P.P.T. sites to help resolve long standing cultural historical issues. This located slightly more than 100 new sites, but no obvious pre\u00C2\u00ADP.P.T. candidates, although the site distribution confirmed quadrat-based patterns and did include several possible Athapaskan sites (Brittany Creek, CR 73) which were also investigated. In 1983 additional quadrats were surveyed in the \"Grassland\" zone in order to May 20, 2003-443 increase our knowledge of that zone primarily to see how different it was from similar environments at Upper Hat Creek. In addition an ethnoarchaeological study was begun in 1983, which continued with different funding in 1984. A dendrochronology project was also carried out by Marion Parker, who was successful in developing a local sequence which was used to date the historic Lulua Phase component at the Bear Lake site. The main effort, though, was in excavating the best Athapaskan candidate, the Bear Lake site, located in Quadrat 32, and two P.P.T. sites, the Boyd and Shields sites. The Bear Lake site had the following Athapaskan characteristics according to mapping and surface collection; a Kavik point, a shallow rectangular depression, and a location near a fishing lake (Eagle Lake) but not adjacent to the shore. The Boyd site, a tight cluster of five pithouses immediately adjacent to Eagle Lake, was an obvious P.P.T. site, but the presence of microblades and some large projectile points and an absence of Kamloops points in excavation, led to the inference that this was a mixed deposit and a decision was made to move our P.P.T. excavations efforts to the similar Shields site less than a kilometer away. Analytical Methods. In the broadest sense, our approach was comparative, anchored by assemblages of known P.P.T. and Athapaskan ethnicity. In addition to examining ethnographic and archaeological reports, we had access to a number of P.P.T. assemblages as well as to those from the MOC and the important Chinlac Athapaskan collection. Also, we obtained loans of side-notched points from a number of different sites for our more detailed analysis of this artifact type in our search for ethnic variation. May 20, 2003-444 The comparative approach thus included everything from ethnographic summaries of winter structures to complex muitivariate analyses of assemblage data that had been coded by the same analysts, using methods appropriate to the data. In some cases these were applications of techniques already in hand -- as in the projectile point analysis, Ritter and Matson (1972) -- in other cases, such as the debitage assemblages, they included approaches developed during the project (Magne 1983, 1985). The interpretation of the results of these analyses were anchored by the twin presences of both assemblages of known ethnicities and of inferred P.P.T. and Athapaskan components at Eagle Lake reducing the chance that other factors such as geographical factors were being mistaken for ethnicity. In these analyses the greater amount of information for the P.P.T. made this part of the comparisons more reliable. For archaeological settlement patterns this was particularly one-sided. The more involved comparisons validated the more intuitive comparisons of Eagle Lake components. Results We discovered that the published ethnographic and archaeological information was broadly both correct and useful. By its use we were able to locate likely P.P.T. and Athapaskan sites through mapping and surface collections. These putatively identified sites were then excavated and the resulting collections were dated and compared in detail with assemblages from the MOC project and elsewhere. In terms of settlement pattern, the vast majority of the material obtained appears Mav 20, 2003-445 _, h 0 y bb aa to be very similar to that found at the Mouth of the Chiicotin and the extensions for root procurement and processing discovered at Upper Hat Creek and those later identified for the Alpine zone on Potato Mountain. Since the ethnographic descriptions of P.P.T. and Athapaskan settlement patterns show a large amount of overlap, it may be that we have missed some subtie differences in this case. The concentration of pit features near salmon streams and Eagle Lake was demonstrated to a rigorous degree, although this association was not original to us. The primary goal, though, was identification of Athapaskan assemblages, and narrowing the time of arrival. Three clear examples of P.P.T. and Athapaskan components were identified, the Bear Lake site (Figure Vlll-i ), the Boyd site (Figure VIII-2), and the Shields site (Figure VIIl-3). The Bear Lake site turned out to have twoAthapaskan components, both with inferred rectangular winter lodges. The historic component, the Lulua Phase, has tree-ring dates of the 1870s, and the prehistoric component, the Eagie Lake Phase, dated to A.O. 1645-1660, according to averaged and calibrated radiocarbon dates. This site not only included the best Kavik point found at Eagle Lake, but the two complete side-notched points recovered were also classified as Athapaskan by our analytical techniques. FIGURE Figure VIII-1 Bear Lake Site Eagle Lake Phase Lithic Component. The Bear Lake site fits the criteria expected of an Athapaskan winter site in terms May 20, 2003-446 a C h r s t u aa w X y 2 3 of location, domestic structure description, \"diagnostic\" projectile point type, side-notch point variant, and dating. With these factors in mind, it is not surprising that in the assemblage comparison, (which included only one of the above) the Bear Lake site was closely associated with known Athapaskan collections. It was more surprising that the much-delayed debitage analysis showed the Bear Lake excavated material was almost as distinctively Athapaskan as the full assemblage analysis. The faunal remains were also distinctive, showing larger amounts of both ungulates and fur bearing mammals than found in the two P.P.T. sites analyzed in the Eagle Lake area. These characteristics are similar to those reported by Stewart (1978) for the inferred Athapaskan fauna} remains found at Anahim Lake. The increase in fur bearers and fish remains between the Eagle Lake and Lulua phase is also a trend reported by Stewart (1978). FIGURE Figure VIIJ-2 Boyd Site P.P.T. Component FIGURE Figure VIII-3. Shields Site P.P.T. Component The Boyd and Shields sites, on the other hand, were clearly associated with other P.P.T. pithouse sites, although only one multi-notched side-notch point was found on May 20, 2003-447 them. In settlement iocation, tight grouping of housepits, and dating, these sites fit what expected of P.P.T. winter viii age sites. In both the assemblage analyses, and the debitage analyses these sites were similar to other P.P.T. assemblages, particular the ones from MOC. As Figure Vl-14 demonstrates, they show remarkable similarities in their artifact profiles and clear differences from the Eagle Lake Component of the Bear Lake site. There is little doubt that these two are good members of the Plateau Pithouse Tradition. The latest date for the Eagle Lake P.P.T. occupation (calibrated radiocarbon date, with estimated correction for structural wood) is A.D. 1475, a Ii ttle less than 200 years prior to the averaged and calibrated date for the Eagle Lake Component. The clear ethnic associations--using multiple means-- and good dating of these three sites are the core results of the Eagle Lake Project. Three other components are worth briefly reviewing in this regard, though. First, Q 19-1, the riverside fishing site, had eight side-notched points all clearly assigned to the P.P.T. variety. There appears little doubt that this is a P.P.T. riverside fishing site, although root-roasting also occurred on it. On looking at Figure Vl-14, though, besides the projectile points, there is little in the assemblage that makes it look very P.P.T. The debitage analysis, though, did put it in the P.P.T. surface collection class, supporting this assignment. Brittany Creek (aka CR 92, EkSa 33) was inferred to be functionally the same as Q 19-1, but with a mixture of both P.P.T. material (as supported by the radiocarbon date860+ I - 80 BP; SFU 14) and later A thapaskan material as indicated by the projectile point analysis which indicated the presence of both ethnicities and by the recent Chilcotin use. Examining Figure Vl-14, though, shows very little that can be said to be significantly different from Q 19-1. Our judgement is that this site does include both May 20, 2003-448 a b e g k ethnicities, but whether the assemblages are not really very different at the level we are analyzing them, or whether the Athapaskan assemblage is only a minor portion, is unclear. FIGURE Figure VIIJ.-4. The Chilko River 73 (EkSa 35) Component Finally there is CR 73 (aka EkSa 35), by far the most anomalous assemblage recovered. Figure Vlll-4 illustrates almost the complete tool component recovered from this pithouse. With a radiocarbon date of 360+ /- 80 BP (SFU 15) this site could be either an early Athapaskan or a recent P.P.T. component. Since the P.P. Tradition is much better known than the Athapaskan tradition, we are 111ore certain that it would be a very unusual P.P.T. site. It is thus more iikely a variant of the lesser known Athapaskan tradition, simply because with less information on this tradition, it is not definitely so anomalous. This assignment is supported by the strange projectile point, and the less strange side-notched one which was classified as \"Athapaskan\" by our multivariate analyses. In sum, we have demonstrated that many of the existing criteria for ethnic differences relying on artifact types are useful. In addition we have discovered that subtle variations within the side-notched point type are correlated with ethnicity and can be used to determine ethnicity. The use of complete assemblages can also distinguish between ethnicities by using common lithic classification schemes, as long May 20, 2003-449 Large Bifaces Deer Fish A.D 800 - 1475KamloopsPPT 50 60 70 80 % Gravers Sinuous Ret.Flk. Deer A.O. 1645. 1810 Eagle Lake Phase A.D.1810 \u00E2\u0080\u00A2 1890Lulua Phase Historical Artifacts \u00EF\u00BF\u00BD@ @ .] -\u00EF\u00BF\u00BD\u00C2\u00B7.,::.-\u00EF\u00BF\u00BD ...... \u00EF\u00BF\u00BD-\u00EF\u00BF\u00BD\u00C2\u00B7 .... 0 4050% O 10 20 3 as site function is held constant. We have also found that some common flake tools also vary according to ethnicity, at least at Eagle Lake (Figure Vl-14), and debitage patterns also vary according to ethnicity. Microblades, although not a factor at Eagle Lake, often appear in recent Athapaskan assemblages and they remain an embarrassment to archaeology. Are they a valid association in this situation, or were they even actually used in the last 2000 years? FIGURE Figure VIIl-5. Summary of Ethnicity and Parallel Sequences Our idealized summary is shown in Figure Vlll-5 which illustrates the two parallel regional sequences, Eagle Lake and Mouth of the Chilcotin. Note that the actual nature of the Lillooet Phenomenon -- whether it is exclusive or not -- remains to be demonstrated. Also, some of the ethnic differences among flake tool types have been demonstrated only for Eagle Lake. Future Research The issue of migration, once a popular topic in archaeology, then avoided, is now again favored. As we have argued, this assumes being able to make ''ethnic identifications\". These two topics need more research--under what conditions do migrations occur? How do ethnic differences in material culture occur? ls it important that material differences be recognized by the ethnic groups themselves? Useful May 20, 2003-450 generalizations of these theoretical concerns are absent and badly needed. In recent years there has been rapid progress on migration, but in contrast ethnicity is relatively under studied. We believe the wide range of methods that we have used shows that there are a wide range of tools that are useful in examining these questions. Although one may question the technical choices made in certain situations, there are a variety of alternative techniques that give similar results. It remains to be dete rmined, though, which are most useful, which should be used first, and which should orientate the research. Obviously, we did not think that the debitage approach would be useful, and we still question the scope of its utility. We do believe that \"hands on\" coding of collections of known ethnicity, along with the unknown, is essentiaL We have demonstrated how a variety of approaches can be used with success in an area where lithic tools dominate the archaeology, but which ones are most useful can only be determined through further investigation. Microblades remain to be confirmed as valid parts of late Athapaskan tool kits -- or even to being used in the last 2000 years. Turning to geographical distribution, we focus first on the Northern Athapaskans. Clearly, our Eagle Lake results should be confirmed elsewhere in B.C. Wiimeth's (1978, 1979) collections from Anahim Lake (which were loaned to us in the 1980s but too late to fully analyze) are an obvious subject for re-analysis. In general, further excavation, dating, and analysis of known Athapaskan sites (and regions) are required further north within British Columbia. Similarly, identification of Athapaskan material in northwestern Alberta is needed, along with a detailed comparison with a more detailed understanding of B.C. Athapaskans. According to the model presented Mav 20, 2003-451 .I in the last chapter, the last 1000 years should look very similar in both areas. These areas should be further investigated so that known Athapaskan material is more fully described and the general modei presented in the last chapter can be filled out and evaluated. There are also areas further north where traces of transitory Athapaskans should be sought. Southern Alberta is an area that did have Athapaskans, even if it is not central to the model presented in the last chapter. A more serious evaluation of existing archaeology needs to be made along with a greater awareness of what Athapaskan assemblages look like. Specifically, there is a report of a possible housepit site on the Tsuu Tina Reserve near Calgary, that should be investigated. Western Montana is an area that also requires a similar awareness and investigation. In addition, re-analysis of Wyatt's (1972) Nicola valley assemblages in B.C. for traces of the Nicola may well be fruitful, now that identifying Athapaskan assemblages is better grounded. Turning to the Pacific Coast Athapaskans, the same awareness and focussed research should to be undertaken in southern Oregon and northwestern California. The parallel direct historic approach appears to be appropriate for that area, in contrast to the plains where the recent history is of dramatic movements, making starting points difficult to identify. In the PCA area Algic and Athapaskan appear to be linked together, yet have very different origins, indicating an obvious subject of investigation. As we have indicated, the Apachean archaeology is much better developed than that in the North, but it does get vague as one goes back in time with radically varying views and identifications (e.g., Towner 1996). There seems to be very little awareness of Northern Athapaskan archaeology in the Southwest. This leads to our last geographical May 20, 2003-452 issue, convening Apacheans, Northern Athapaskans and PCA and those interested in these issues to set up lines of communication and gain first hand knowledge of the material culture and problems involved. Awareness and useful understanding will only be resolved in this manner, as we are spread over six culture areas (Subarctic, Plateau, Northwest Coast, California, Plains, Southwest) and at least seven Handbook Volumes! Migration as an explanation and as a focus of interest has waxed and waned (and waxes again) in archaeology. If we are ever to understand the human condition, how we came to be, we need to be able to identify and explain past migrations. The Athapaskan language family certainly has a distribution that can only be understood as the result of multiple migrations. Archaeology has developed (or adopted) procedures by which migrations can be identified and dated, even of Archaic Stage peoples. We have demonstrated this ability here, identifying a specific migration, that of the Chilcotin, and we have placed it in the spectrum of Athapaskan migrations of the last 1500 years. We see no reason why similar successful investigations can not be used on other putative migrations. The methods illustrated here should also be capable of further documenting Athapaskan history. May 20, 2003-453 REFERENCES CITED Adams, William Y. 1968 Invasion, Diffusion, Evolution? Antiquity 42:194-215. Adams, William Y., Dennis P. Van Gerven, and Richard Levy 1978 Retreat from Migrationism. In Annual Review of Anthropology, Vol. 7, pp. 483-532, Annual Reviews, Palo Alto. Adovasio, James M. and David R. Pedler 1994 A tisket, a tasket: Looking at theNumic speakers through the \"lens\" of a basket. Pp 114-123 in Across the West: Human Population movement and the expansion of theNuma.,_ ed. by David Madsen and David Rhode. University of Utah Press, Salt LakeCity.Ahlstrom, Richard, Carla Van West, and Jeffrey Dean 1995 Environmental and Chronological Factors in Mesa Verde-No11hern Rio Grande Migration. Journal of Anthropological Archaeology 14:125-142 (Special Issue, ed. by. C. Cameron, Migration and the Movement of Southwestern Peoples)Albright, S.L. 1982 An ethnoarchaeological study of Tahltan subsistence and settlement patterns . Unpublished Masters thesis, Simon Fraser University, Burnaby. Aiexander, Diana 1992 A Reconstruction of Prehistoric Land Use in the Mid-Fraser River Area based on Ethnographic Data. Pp. 99-176 in A Complex Culture of the British Columbia Plateau: Traditional Stl' atl'imx Resourse Use,ed. by. Brian Hayden, University of British Columbia Press, Vancouver. Alexander, Diana, Robert Tyhurst, R.G. Matson. and Linda Burnard 1985 A Preliminary Ethnoarchaeological Investigation of the Potato Mountain Range and the Eagle Lake Area. Report prepared for the Heritage Conservation Branch of B.C. (Permit No. 1984-14 ), Victoria. Alexander, Diana and R.G. Matson 1987 Report on the Potato Mountain Archaeological Project (1985). Repo11 prepared for the Heritage Conservation Branch of B.C. (Penn it No. 1985-1 I), Victoria. Annas, R.M. and R. Coupe (Editors) 1979 Bioclimatic Zones and Subzones of the Cari boo Forest Region. Province of B1itish Columbia, Ministry of Forests, Victoria. Anonymous n.d. Fort Chilcotin. Typescript, British Columbia Archives MMC43. Victoria. May 20, 2003--454 Anthony, David W. 1990 Migration in Aschaeology: The Baby and the Bathwater, American Anthropologist 92: 895-914.Anthony, David W. 1997 Prehistoric Migration as Social Process. In Migrations and Invasions in Archaeological Explanation, ed. by. John Chapman and Helena Hamerow, pp. 21-32, BAR International Series 665, Archaeopress, Oxford. Apland, B. 1979 Reconnaissance survey in the Rainbow Mountains region of west-central British Columbia. Annual Report for the Year 1976. He1itage Conservation Branch of British Columbia, Victoria. Aro, K.V. and M.P. Shepherd 1967 Pacific salmon in Canada. In Salmon of the North Pacific Ocean, Pa11 IV, Spawning populations of N01th Pacific Salmon. International North Pacific Fisheries Commission, Bulletin 23:225 - 327. Baker, James 1975 Report on Archaeological Research Done in the Lytton Region, May to August, i974. Report to Archaeological Sites Advisory Board, Victoria. Bakewell, Edward F. and Anthony Irving 1994 Volcanic Classification in the Pacific Northwest: Petrographic and Geochemical Analysis of Northwest Coast Artifacts. No,thwest Anthropological Research Notes 28(1):29-37 Bamforth, Douglas, 1988 Ecology and Human Organization on the Great Plains. Plenum Press, New York and London. Banfield, A.W.F. 1974 The Mammals of Canada. National Museum of Natural Science, Ottawa, and University of Toronto Press, Toronto. Baumhoff, Martin A. 1958 California Athabaskan Groups. Anthropological Records I 6:5. University of California Press, Berkeley. Baugh, Timothy G. and Frank W. Eddy 1987 Rethinking Apachean Ceramics: The 1985 Southern AthapaskanCeramics Conference. American Antiquity 52(4): 793-798. May 20, 2003--455 Bellwood, Peter, Colin Renfrew and K.V. Boyle (Editors) 2002 Examining theFarming/Language Dispersal Hypothesis. McDonald Institute for Archaeological Research, Cambridge. Berreman, Joel V. 1944 Chetco Archaeology. General Series in Anthropology, No. 1 L Menasha, Wisconsin. Berry, Michael S. 1982 Time, Space, and Transition in Anasazi Prehistory. University of Utah Press, Salt Lake City. Bettinger, R. 1975 The surface archaeology of Owens Valley, Eastern California. Unpublished Ph.D. dissertation, University of California, Riverside. Binford, Lewis R. 1963 A proposed attribute list for the description and classification of projectile points. Miscellaneous Studies in Typology and Classification, ed. by A.M. White, LR. Binford, and M.L. Papworth, pp. 193-221, University of Michigan Museum of Anthropology, Anthropology Papers no. l 9, Ann Arbor. 1964 A consideration of archaeological research design. American Antiquity 29: 425-44L 1968 Smudge pits and hide-smoking: the use of analogy in archaeological reasoning. American Antiquity 32: l - 12. 1978a Dimensional analysis of behavior and site structure: learning from an Eskimo hunting stand. American Antiquity 43: 330 - 361. 1978b Nunamiut Ethnoarchaeology. Academic Press. New York 1979 Organization and formation processes: looking at curated technologies. Journal of Anthropological Research 3: 255 - 273. 1980 Willow smoke and dog's tails: hunter-gatherer settlement systems and archaeological site formation. American Antiquity 45: 4 - 20. Borden, Charles 1952 Results of Archaeological Investigations in Central BritishColumbia. Anthropology in British Columbia 3: 31 - 43. 1979 Peopling and Early Cultures of the Pacific Northwest. Science 203:963-971. May 20, 2003--456 Bouchard, Randy and Dorothy Kennedy i984 Indian History and Knowledge of the Lower Similkameen River -Palmer Lake Area. Unpublished ms. Draft Report to the US Army Corp ofEngineers, BC indian Language Project, Victoria. Boudreau, Norman J., editor 1974 The Athapaskans: Strangers of the North, National Museum of Man, Ottawa. Brant, Charles S. 1949 The C ultural Position of the Kiowa-Apache. Southwestern Journal of Anthropology 5: 56-61.1953 Kiowa Apache Culture History: Some Fm1her Observations. Southwestern Journal of Anthropology 9: 195-202 Broderick, M. 1979 Ascending paper chromatographic technique in archaeology. In Lithic use-wear analysis, edited by B. Hayden, pp. 375 - 383. Academic Press. New York. Broecker, Wallace 2001 Was the Medieval Warm Period Global? Science 29! (5508): 1497 Brugge, David M. 1983 Navajo Prehistory and History to 1850. In Handbook of No11h Arnerican Indians Vol. 10, Southwest, edited by Alfonso Ortiz, pp. 489 -SOL Smithsonian Institution. Washington. Burch, Ernest S. Jr. and Thomas CoJTell 1972 Alliance and Conflict: Inter-Regional Relations in North Alaska. In Alliance in Eskimo Society: Proceedings of the American Ethnological Society, 1971,. Supplement Ed. by D.L. Guemble, pp. 17-39, University of Washington Press, Seattle. Burmeister, Stefen 2000 Archaeology and Migration. Cunent Anthropology 41(4):539-567. Burnard Linda 1987 Metes and Bounds (A Search for archaeological indicators of hunter-gatherer territoriality). Unpublished Masters thesis, University of British Columbia, Vancouver. Bussey, J. 1983 Alexis Creek archaeological investigations. Points West Consuitants Ltd. report to Heritage Conservation Branch of British Columbia, Victoria. May 20, 2003--457 Byrne, William J. 1973 The Archaeology and Prehistory of Southern Albe1ta As Reflected By Ceramics . National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 14. Ottawa.Cameron, Catherine (Editor) 1995 Migration and the Movement of Southwestern Peoples. Journal of Anthropological Archaeology 14:99-250 (Special Issue). Campbell, J.M. 1968 The Kavik site of Anaktuvuk Pass, arctic Alaska.Anthropological Papers of the University of Alaska 14: 32 - 42. Campbell, Lyle 1997 American Indian Languages: The Historical Linguistics of Native America. Oxford University Press, New Yorlc. 1998 Historical Linguistics: An Introduction. Edinburgh University Press, Edinburgh. Campbell, Lyle and Marianne Mithun, Editors. 1979 The Languages of Native America: historical and comparative Assessment. University of Texas Press, Austin. Campbell, R.C. 1967 Statistics for Biologists .. Cambridge University Press, Cambridge. Canada Land Inventory 1970 Land capability for wildlife-waterfowl, Mount Waddington Map Sheet 92U. Department of Regional Economic Expansion. Carlson, Roy L. 2000 Archaeological Excavations at Hallett Lake. Permit 1966-155. Report on File, Archaeology Branch, Victoria. Clague, John, S.G. Evans, V.N. Rampton, and G.J. Woodswo1th 1995 Improved age estimates for the White River and B1idge River tephras, western Canada. Canadian Journal of Earth Sciences 32: 1172-1179. Clark, A. McFadyen 1996 Who Lived in this House? A study of Koyukuk River semisubte1i-anean Houses. Canadian Museum of Civilization Mercury Series, Archaeological Survey of Canada Paper 153, Hull. Clark, A. McFadyen and Donald W. Clark 1974 Koyukon Athapaskan Houses as seen through Oral Tradition and through Archaeology. Arctic Anthropology XI (Supplement): 29-38. May 20, 2003--458 Clark, Donald W. 1975 Archaeological Reconnaissance in No1thern Interior District of Mackenzie: 1969, 1970, and 1972, National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 27, Ottawa. 1981 Prehistory of the Western Subarctic. In Handbook of No1thAme1ican Indians, Vol. 6; Subarctic, edited by June Helm, pp. 107 - 129, Smithsonian Institution, Washington .. 1982 The Vanishing Edge of Today in the No1thern Dist1ict of Mackenzie: A view from field Archaeology. Canadian Journal of Archaeology 2(2):107-128. Connolly, Thomas J. 1986 Cultural Stability and Change in the Prehistory of Southwest Oregon and Northern California. Unpublished Ph.D. Dissertation. University of Oregon, Eugene. 1991 The Standley Site (35DO182): Investigations Into the Prehistory of Camas Valley, Southwest Oregon. Anthropological Papers No.43. University of Oregon, Eugene. Connolly ,William 1825 Visit to the Chilcotin, December 1825. Hudson's Bay Company Archives, B.188/a/4. Winnipeg. Conover, W.J. 1971 Practical Nonparametric Statistics. John Wiley &Sons Inc .. New York. Cook, John P 1969 The Early Prehistory of Healy Lake, Alaska. Unpublished Ph.D.Dissertation, University of Wisconsin, Madison. Cook, S.F. 1956 The Aboriginal Population of the North Coast of California. Anthropological Records Vol.16, No. 3 (81-129), Berkeley. Cooley,W.W. and P.R. Lohnes 1971 Multivariate data analysis. Wiley and Sons. New York. Copp, Stanley A. 1997 A Plateau Microblade Tradition Site: Upper Similkameem Valley B.C. Paper Presented at the N01thwest Anthropology Conference, El!ensburg, Washington. Coupland, Gary 1996 The Early Prehistoic Occupation of Kitselas Canyon. In Early Human Occupation in British Columbia, ed. by R.L. Carlson and L. Dalla Bona, pp. 158-166, University of British Columbia Press, Vancouver. May 20, 2003--459 Cowan, Ian McTaggart and Charles Guiguet n.d. The Mammals of British Columbia. British Columbia Provinciai Museum Handbook No. 11, Victoria. (1965 revision according to Acknowledgements). Croes Dale R. and S. Hackenberger 1988 Hoko River archaeological complex: Modelling prehistoric N011hwest Coast Economic evolution. In Research in Economic Anthropology Supplement 3: Prehist01ic Economies of the Pacific Northwest Coast, ed. by B.L. Isaac, JAi Press, Greenwich, CT. Pp. 19-85. Crowiey, Terry 1997 An Introduction to Historical Linguistics, 3rd ed. Oxford University Press, Auckland. Davis, C. 1975 Some notes on Piateau Athapaskan. National Museum of Man, Canadian Ethnology Service Paper, Mercury Series No. 27. Proceedings of the Northern Athapaskan Conference 1971, Volume 2: pp. 619 - 629. Davis, Leslie B. (Editor) 1988 Avonlea Yesterday and Today: Archaeology and Prehistory. Saskatchewan Archaeological Society, Saskatoon. Dawson, G.M. 1891 Notes on the Shuswap people of British Columbia. Royal Society of Canada, Transactions. Section II, pp. 2 - 44. Dawson Brothers. Denniston, Glenda 1981 Sekani. In Handbook of Nmth American Indians, Volume 6; Subarctic, ed. by June Helm, pp. 433-441, Smithsonian Institution, Washington. Derry, David E. 1975 Later Athapaskan Prehistory: A Migration Hypothesis. Western Canadian Journal of Anthropology 5(3-4): 134-147. Derry, David E. and Douglas R. Hudson (Editors) 1975 Athapaskan Archaeology (Special Issue). Western Canadian Journal of Anthropology 5(3-4): 1-Dinwoodie, David W. 1998 Authorizing Voices: Going Public in an Indigenous Language. Cultural Anthropology 13(2): 193-223. 2002 Reserve Memories: The Power of the Past in a Chilcotin Community. University of Nebraska Press, Lincoln. May 20, 2003--460 Donahue, Paul F. 1972 Preliminary Repo11 of Excavations at Tezli (FgSd l), B.C. Rep011 on file at Archaeological Sites Advisory Board, Victoria. (Cited by Helmer 1977). 1973 Ulkatcho: An Archaeological Outline. Syesis 6: 154-178. 1975 Concerning Athapaskan Prehistory in B.C. Western Canadian Journal of Anthropology 5(3-4):21-64. 1977a 4500 Years of Cultural Continuity on the Centrai Interior Plateau of British Columbia . Unpublished Ph.D. Dissertation. University of Wisconsin. Madison. 1977b Prehistoric Relationships between the Plains, Boreal Forest and Cordilleran Regions. In Problems in the Prehistory of the N011h American Subarctic: The Athapaskan Question, ed. by James Helmer, S. Van Dyke, and F. J. Kense, pp. 83-89, Archaeological Association, University of Calgary, Calgary. Duff, Andrew 1998 The Process of Migration in the Late Prehistoric Southwest.In: Spielmann, Katherine A. (Editor), Migration and Reorganization: The Pueblo IV Period in the AmericanSouthwest. Anthropological Research Papers No. 51, pp. 31-52. Arizona StateUniversity. Tempe.Duff, Wilson 1981 Tsetsaut. In Handbook of North American Indians. Vol. 6. Subarctic, ed. by J. Helm, pp. 454-457. Smithsonian Institution, Washington. Dumond, D 1978 Alaska and the Northwest Coast. In Ancient Native Americans. Edited by J. Jennings, pp. 43 - 93. W.H. Freeman and Sons. San Francisco. Dyen, Isadore and David F. Aberle 1974 Lexical Reconstruction: the Case of the Proto-Athapaskan Kinship System. Cambridge University Press. Cambridge. Eldridge, A. 1982 Cambium resources of the Pacific n011hwest: an ethnographic and archaeological study. Unpublished ms. Department of Archaeology. Simon Fraser University, Burnaby. Eldridge, M. 1974 Survey sampling on the Fraser Plateau. Unpublished ms. on file, Heritage Conservation Branch of British Columbia. Eldridge, M. and A. Eldridge 1980 An evaluation of the heritage resource potential of the Dean River Valley. Final report on the Cariboo planning study. Unpublished report to the Heritage Conservation Branch of British Columbia. May 20, 2003--461 Erlandson, Jon, Mark Tveskov, and Madonna Moss 1997 Return to Chetlessenten: The Antiquity and Architecture of an Athapaskan Village on the Southern Northwest Coast. Journal of California and Great Basin Anthropology 19(2):226-240. Fafard, Melanie 2001 Dechyoo Njik (MlVm-4) and the Traditional Land Use Patterns of the Southwestern Portion of the Old Crow Flats, Yukon Ten-itory. Occasional Papers in Archaeology No. 8. Hude Hodan Series. HeritageBranch. Government of the Yukon, Whitehorse. Farley, AL. 1979 Atlas of British Columbia: people, environment and resource use . University of British Columbia Press, VancouveL Farrand, L. 1898 The Chilcotin. British Association for the Advancement of Science 68: 645-648. 1900 Traditions of the Chilcotin Indians. American Museum of Natural History Memoir 4, Anthropology Vol. 3, Pait I, Jesup North Pacific Expedition Vol. 2, Pait 7.New York. 1910 Tsilkotin. Handbook of American Indians North of Mexico, Vol. 2:826, ed. by Frederick W.Hodge, Bureau of American Ethnology Bulletin 30, Washington. Fladmark, K.R. 1974 Punchaw Village: A Prelinmary report on the Archaeology of a Prehistoric Settlement. Report to the Archaeological Sites Advisory Board,Victoria. 1976 Punchaw village: preliminary report, archaeology of a prehistoric settlement. In Cun-ent Research Reports. ed. by R. Carlson, pp. 19 - 32, Depa1tment of Archaeology Publication No. 3, Simon Fraser University, Burnaby. 1982 An introduction to the prehistory of British Columbia. Canadian Journal of Archaeology 6: 95-156. Forbis, Richard 1977 Cluny, An Ancient Fortified Village in Alberta.Depa1tment of Archaeology, Occasional Paper No. 4. University of Calgary.Calgary. Foster, Michael K. 1996 Language and Culture History of North America. In: Handbook of North American Indians Vol. 17, Languages, ed. by Ives Goddard, pp. 64-110. Smithsonian Institution. Washington. May 20, 2003--462 Fraser, Simon 1966 The Letters and Journals of Simon Fraser l 806-1808. Edited by W .K. Lamb. Macmillan and Co. Toronto. Fredrickson, David A. 1984 The North Coastal Region, In California Archaeology by Michael J. Moratto, pp. 471-527, Academic Press Inc., Orlando. Fried, Morton 1967 The Evolution of Political Society; An essay in political anthropology. Random House, New York. 1975 The Notion of Tribe. Cummings, Menlo Park. Frison, George 1973 The Wardell Buffalo Trap 48SU301: Communal Procurement in the Upper Green RiverBasin, Wyoming. Anthropologicai Papers of the Museum of Anthropology, no. 48. University of Michigan, Ann Arbor. Gauch, Hugh G. Jr. 1982 Multivariate Analysis in Community Ecology. Cambridge University Press, Cambridge. Geib, Phil R. 1996 Glen Canyon Revisited. University of Utah Anthropological Papers No. 119, University of Utah Press, Salt Lake City. Germann, C. 1979 Cariboo regional archaeological impact assessment, 1979. Unpublishd ms., Heritage Conservation Branch of British Columbia, Victoria. Goddard, Ives 1996 The Classification of the Native Languages of North America .. In: Handbook of North Ame1ican Indians Vol. 17, Languages, ed. by Ives Goddard, pp. 290 -323. Smithsonian Institution. Washington. Godfrey-Smith, D. and J.M. D' Auria 1984 New data on obsidian exchange routes in western Canada and Ontario. Paper presented to the 17th annual meeting, Canadian Archaeological Association, Victoria. Gordon, Bryan H. C. 1975 Of Men and Herds in BaJTenland Prehistory. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 28.Ottawa. May 20, 2003--463 Gould, Richard 1966 Archaeology of the Point St. George Site and Tolowa Prehistory. University of California Publications in Anthropology No.4, Berkeley. 1978 Tolowa. In Handbook of North American Indians Voi. 8: California. ed. by R.F. Heizer (W.C. Stmtevant Gen. ed.) Smithsonian Institution, Washington. Pp. 128-136. Greaves, Sheila 1982 Upon the point: a preliminary investigation of ethnicity as a source of metric variation in lithic projectile points. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 109, Ottawa. 1991 The Organization of Microcore Technology in the Canadian Southern Interior Plateau . Unpublished Ph.D. Dissertation. University of British Columbia, Vancouver. 1998 Plateau Microblade Tradition. In: Archaeology of Prehistoric Native America: An Encyclopedia, ed. by Guy Gibbon, pp. 661- 662. Garland Publishing, Inc. New York. Greer, Sheila 1985 Kusawa Lake Archaeology Project Field Report. Report prepared for Archaeological Survey of Canada, National Museum of Man and Parks Planning, Dept. of Renewable Resources, Yukon, Whitehorse. 1986 Kusawa Lake Archaeology: Management, Re.search and Interpretation. Report prepared for ArchaeologicalSurvey of Canada, National Museum of Man and Parks Planning, Dept of Renewable Re.sources, Yukon, Whitehorse. Greiser, Sally T. 1988 Lost Terrace Avonlea Mate1ial Culture. In: Avonlea Yesterday and Today: Archaeology and Prehistory, ed. by Leslie B.Davis, pp. l l 9 - 128. Saskatchewan Archaeological Society, Saskatoon. Gulley, Cara C. 2000 A Reanalysis of Dismal River Archaeology and CerarnicTypology . Unpublished Masters thesis, University of Colorado, Boulder. Gunnerson, James H. 1960 An Introduction to Plains Apache Archaeology - The Dismal River Aspect. Bureau of American Ethnology Bulletin i 73, Anthropological Papers No. 58 . Smithsonian Institution. Washington. 1969 Apache Archaeology in Northeastern New Mexico. American Antiquity 34:23-39. May 20, 2003--464 1979 Southern Apachean Archaeology. In Handbook of No11h Ametican Indians, Vol. 9, Southwest, ed. by Alfonso Ortiz, pp. 162-169, Smithsonian Institution, Washington. Gunnerson, James H. and Dolores A. Gunnerson 1971 Apachean Culture: A Study in Unity and Diversity. In: Apachean Culture History and Ethnology, edited by H. Basso and Monis E. Opler, pp. 7 - 27. Anthropological Papers of the University of Arizona No. 2 L University. of A1izona Press, Tucson. Habicht-Mauche, Judith. 1992 Coronado's Querechos and Teyas in the Archaeological Record of the Texas Panhandle. Plains Anthropologist 37:247-259. Haig-Brown, A. 1983 British Columbia Indian ianguages: a crisis of silence. B.C. Studies No. 57: 57 - 67. Hair, L.F., R.F. Anderson, R.L. Tatham, and B.J. Grabiowsky 1979 Multivariate data analysis PPC Books, Tulsa. Hale, Horatio 1885 Report on the Blackfoot Tribes. Repo11 of the BritishAssociation for the Advancement of Science, 1885: 698 - 699. London. Ham, Leonard C. 1975 Shuswap Settlement Patterns. Unpublished Masters thesis, Simon Fraser University, Burnaby. Hayden, Brian (Editor) 1979 Lithic use-wear analysis Academic Press, Inc. New York. Hayden, Brian 1997 The Pithouses of Keatley Creek. Harcourt Brace, Fo11 Worth. (Editor) 2000 The Ancient Past of Keatley Creek. Volume I: Taphonomy. Archaeology Press, Simon Fraser University, Burnaby. (Editor) 2001 The Ancient Past of Keatley Creek. Volume 1 I: Social Economic Interpretation. Archaeology Press, Simon Fraser University, Burnaby Hayden, Brian and June M. Ryder 1991 Prehistoric Cultural Collapse in the Lillooet Area. American Antiquity 56(1):50-65. Hayden, Brian and Rick Schulting 1997 The Plateau Interaction Sphere and Late Prehstoric Cultural Complexity. American Antiquity 62(1):51-85. May 20, 2003--465 Heitzmann, Rod 2001 Historic Artifacts from Bear Lake site, EkSa 36. Report on File, Laboratory of Archaeology, University of B1itish Columbia. Helm, June (Editor) 1981 Handbook of Nonh American Indians, Vol. 6, Subarctic. Smithsonian Institution, Washington. Helmer, James W 1977a Site dispersal in the Blackwater Valley: theoretical perspectives and preliminary synthesis. In Annual report for the year 1975 of the Archaeological Sites Advisory Board and Selected Research Rep011s, edited by B.Simonsen, pp. 46 - 107. 1977b Points, People, and Prehistory: A Preliminary Synthesis of Culture History in Nor th Central B1itish Columbia. In Problems in the Prehistory of the N01th American Subarctic: The Athapaskan Question, ed. by James Helmer, S. Yan Dyke, and F. J. Kense, pp. 90-96 Archaeological Association, University of Calga1y, Calgary. 1982 Report of the 1982 Archaeological Survey of the Tsuu T ina Indian Reserve #145. Unpublished ms. Dept. of Archaeology. University of Calgary, Calgary. Helmer, James W. and Donald T. Hanna 1996 Archaeological Investigations at the Old Agency Site and Clubhouse Site on the Proposed Buffalo Run Golf Course, Tsuu Tina Nation.Unpublished ms, Dept. of Archaeology, University of Calgary, Calgary. Helmer, James W., S. VanDyke, and F.J. Kense (Editors) 1977 Problems in the Prehistory of the North American Subarctic: The Athapaskan Question, Archaeological Association, University of Calgary, Calgary. Hester, J.J. 1962 Early Navajo migrations and acculturations in the Southwest. Museum of New Mexico. Papers in Anthropology No. 6 . Santa Fe. Hewlett, E.S. 1973 The Chilcotin uprising of 1864. B.C. Studies 47: 30 - 72. Hill, Jane H. 2001 Proto-Uto-Aztecan: a community of cultivators in central America'? A11erican Anthropologist 103(4):913-34. 2002 Proto-Uto-Aztecan cultivation and the Northern Devolution; Linguistic Approaches. In Examining the Fam1ing/Language Hypothesis, ed. by P. Bellwood, C. Renfrew and KV.Boyle, McDonald Institute for Archaeological Research, Cambridge. May 20, 2003--466 Hobler, Philip and Beth Bedard 1992 Qwliutl and Snxlhh : Prehistory to History in the Bella Coola Valley. Report on file, Archaeology Branch, Victoria. Hoijer, Harry 1956 Chronology of the Athapaskan Languages. International Journal of American Linguistics 22: 219-233. 1963 The Athapaskan Languages. In: Studies in the Athapaskan Languages, by Harry Hoijer and Others, pp. 1 - 29.University of California Publications in Linguistics . Vol. 29. University of CaliforniaPress. Berkeley. 1971 The Position of the Apachean Languages in the Athapaskan Stock. In: Apachean Culture History and Ethnology, ed. by Keith H. Basso and MoITis E. Opler, pp. 3 -6. Anthropological Papers of the University of Arizona No. 21. University of ArizonaPress, Tucson.Holland, S. 1964 Landforms of British Columbia, a physiographic outline. British Columbia Depa1tment of Mines and Petroleum Resources Bulletin No. 48. Hymes, D. H. 1957 A Note on Athapaskan Glottochronology. International Journal of American Linguistics 23: 291-297. 1960 Lexicostatistics So Far. CuITent Anthropology I (I) 3-44. Ives, John W. 1990 A Theory of Northern Athapaskan Prehistory. Westview Press, Boulder. Jenness, Diamond 1938 The Sarcee Indians of Albe1ta. National Museum of Canada Bulletin No. 90. Anthropological Series No. 23, Ottawa. Jongman, R.H., Braak, CJ., and O.F.R. Van Tongeren 1995 Data Analysis in Community and Landscape Ecology. Cambridge University Press, Cambridge. Jorgensen, Joseph S. 1980 Western Indians. W.H. Freeman, San Francisco. Judge, W. James 1973 Paleolndian Occupation of the Central Rio Grande Valley in New Mexico. University of New Mexico Press, Albuquerque. May 20, 2003--467 Judge, W.J., J. Ebert and R. Hitchcock 1975 Sampling in archaeological survey. In Sampling in Archaeology, edited by J. Mueller, pp. 82 - 123. University of A1izona Press, Tucson. Kari, James 1989 Some Linguistic Insights into Dena'ina Prehistory. In Athapaskan Linguistics: Current Perspectives on a Language Family, ed. by Eung-Do Cook and K.D. Rice, pp. 533-574, Mouton de Gruyter, New York. Kay,M. 1975 Social distance among Missouri Hopewell settlements: a first approximation. American Antiquity 40: 64 - 71. Keddie, Grant 1972 The C hilcotin Archaeological and Ecoiogical Survey Project. Report to the Archaeological Site Advisory Board, Victoria. Kehoe, Thomas F. 1973 The Gull Lake Site: A Prehistoric Bison Drive Site in Southwestern Saskatchewan. Publications in Antlhropology and Hi.story no. I. Milwaukee Public Museum, Wisconsin. Kew, Michael I 992 Salmon Availability, Technology, and Cultural Adaptation in the Fraser River Watershed. In A Complex Culture of the British Columbia Plateau, ed. by Brian Hayden, pp. 177-221. University of British Columbia Press, Vancouver. Kidd, K.E. and M.A. Kidd 1970 A classification system for glass beads for the use of field archaeologists. Canadian Histmic Sites Occasional Paper.sin Archaeology and History No. 1: 45 - 89. Kincade, M. Dale and J. V. Powell 1976 Language and the Prehistory of No11h America. World Archaeology 8( I): 83-100. Klassen, Michael 1997 An Archaeological Impact Assessment of the Ministry of Forests-Chilcotin Forest District's Proposed Tatlayoko Lake North Woodlot. Report for Permit 1995-107c, Archaeology Branch, Ministry of Community, Aboriginal & Women's Services, Victoria. Klassen, Michael A. 2002 Chilko Lake Housepit Site (EjSa I I): Archaeological Investigations 200 I (Permit 2001-251). He1itage Resource Centre. Ministry of Community, Aboriginal & Women's Services. Victoria. May 20, 2003--468 Klassen, Michael and Amber Ridington 1998 Chilko River Archaeological Inventory Sample Study: Phase I (1997). Report under Permit 1 997-214, Archaeoiogy Branch, Ministry of Small Business, Tourism and Culture, Victoria. Klecka, W.P. 1975 Discriminant analysis. In SPSS: Statistical package for the social sciences . Edited by N.H. Nie, C.H. Hall, J.G. Jenkins, K. Steinbrenner, and D.H. Bent. pp. 434 - 467. McGraw-Hill, Toronto. Krajina, VJ. 1973 Biogeoclimatic zones of British Columbia. Map published by B.C. Ecological reserves committee. B.C. Department of Land, Forest, and Water Resources. Ottawa. Krauss, Michael E. 1973 Na-Dene. InLinguistics in North Ame1ica, Current Trends in Linguistics, Vol. IO, ed. by Thomas A. Sebeok, pp. 903 - 978. Mouton, The Hague. 1979 Na-Dene and Eskimo-Aleut. Pp. 903-78 irfhe Languages of Native America: historical and comparative Assessment, ed. by L. Campbell and M. Mithun, University of Texas Press, Austin. Krauss, Michael E. and Victor K. Golla 1981 N01them Athapaskan Languages. In Handbook of NorthAmerican Indians, Vol. 6; Subarctic, ed. by June Helm, pp. 67 - 85.Smithsonian Institution, Washington. Kroeber, Alfred L. 1925 Handbook of the Indians of California. Bureau of American Ethnology Bulletin 78., Smithsonian Institution, Washington. Kuijt, Ian 2001 Reconsidering the Cause of Cultural Collapse in the Liliooet Area of British Columbia, Canada: A Geoarchaeological Perspective. American Antiquity 66(4): 692-703. Lamb, W.K. 1960 The letters and journal of Simon Fraser 1806 - i 808. MacMillan,Toronto. Lane,R. 1953 Cultural relations of the Chilcotin Indians of west central British Columbia . Unpublished Ph.D. dissertation, University of Washington, Seattle. 1981 Chilcotin. In Handbook of No1th American Indians. Vol. 6. Subarctic, ed. by J. Helm, pp. 402-412. Smithsonian Institution, Washington. May 20, 2003--469 LeBlanc, Raymond and John W. Ives 1986 The Bezya Site: A Wedge-Shaped Core Assemblage from Northeastern Alberta. Canadian Journal of Archaeology I 0: 59 - 98. LeBlanc, Steven A. 1998 Settlement Consequences of Warfare During the Late Pueblo Illand Pueblo IV Pe1iods. In: Spielmann, Katherine A. (Editor), Migration and Reorganization: The Pueblo IV Period in the American Southwest. Anthropological Research Papers No. 51, pp. I 15 135. Arizona StateUniversity, Tempe.1999 Prehistoric Waifare in the American Southwest. University of Utah Press, Salt Lake City. Lehmann, Winfred P. 1962 Historical Linguistics; An Introduction. Holt.Rinehart and Winston,Inc, New York. Lenert, Michael 2001 Calibrated Radiocarbon Dates and Culture Change: Implications for Socio-Complexity in the Mid-Fraser Region, British Columbia. No,thwest Anthropological Research Notes 35(2):211-228 Lepofsky, D. n.d. Field manual for floatation of archaeological sediments. Unpublished ms. Archaeology Laboratory of Archaeology, University of British Columbia, VancouveL Lerbekmo, J. F., J. A. Westgate, D. G. W. Smith, and G. H. Denton 1975 New Data on the Character and History of the White River Volcanic Eruption, Alaska. In Quaternary Studies Bulletin No. 13, ed. by R. P. Suggate and M. M. Creswell, pp. 203 - 209. Royal Society of New Zealand. Wellington. Lipe, W.D. 1995 The Depopulation of the Northern San Juan: Conditions in the Turbulent 1200s. Journal of Anthropological Archaeology 14: 143-169 (Special Issue, ed. by C. Cameron, Migration and the Movement of Southwestern Peoples) Lipe, W.D. and R.G. Matson 1971 Human Settlement and Resources in the Cedar Mesa Area, S.E. Utah. In The Disuibution of Prehistoric Population Aggregates, ed. by G. Gumerman, Pp. 126-151, Prescott College Anthropological Reports, No. I, Prescott. Lohse, Ernest S. 1985 Rufus Woods Lake Projectile Point Chronology. In Summary of R esults, Chief Joseph Dam Cultural Resources Project. Ed. by S. Campbell, pp. 3 I 7-364. Office of Public Archaeology, Institute for Environmental Studies, University of Washington, Seattle. Mav 20, 2003--470 ., Long, Austin and Bruce Rippeteau 1974 Testing contemporaneity and averaging radiocarbon dates. American Antiquity 39: 205-213 Loy, Thomas and G.R. Poweil 1977 Archaeological Data Recording Guide. British Columbia Provincial Museum Heritage Guide No. 3. Victoria. Ludowicz, Deanna 1980 Features expected in recent Athapaskan Sites. Pp. 221-228 in The Eagle Lake Project: Report on the 1979 Season. R.G. Matson, M. Magne, D. Ludowicz and D.L. Pokotylo. Laboratory of Archaeology, University of British Columbia, Vancouver. 1983 Assemblage Variation Associated With Southwestern Inte1ior Plateau Microblade Assemblages. Unpublished MA Thesis, University ofBritish Columbia, Vancouver. Lyman, R.Lee 1991 Prehistory of the Oregon Coast. Academic Press, Inc. San Diego. MacDonald, George 1968 Debert: A Paleo-Indian site in Central Nova Scotia. Anthropological Papers No. 16. National Museums of Canada, Ottawa. MacDonald, Richard 2001 Structural Strategies for Pithouses at the Keatley Creek Site. In The Ancient Past of Keatley Creek, Volume II: Socioeconomic Interpretation, ed. by Brian Hayden, pp. 239-252, Archaeology Press, Simon Fraser University, Burnaby.Magne, M. 1983 Lithics and livelihood: stone tool technologies of central and southern interior B.C. Unpublished Ph.D. dissertation, University of British Columbia, Vancouver. 1984 Taseko Lakes prehistory project: repo11 on a preliminary survey. Report to B.C. Heritage Trust and the Nemiah Valley Indian Band Council. Laboratory of Archaeology, University of British Columbia. Vancouver. 1985a Lithics and livelihood: Stone tool technologies of Central and Southern Interior B.C. National Museum of Man Mercury Series, Archaeological Survey of Canada, No. 133, Ottawa 1985b Taskeo Lakes Prehistory Project. Phase II: Preliminary Excavations. Report to B.C. Heritage Trust and the Nemiah Valley Indian Band Council. Laboratory of Archaeology, University of British Columbia. Vancouver. May 20, 2003--471 2001 Plateau and Plains Athapaskan Movements in Late Prehistoric and Early Historic Times: A view from the Middle. Paper delivered at the \"Migration: Then and Now\" symposium of the Society for Ame1ican Archaeology Annual Meetings, April, 2001, New Orleans. Magne, M. and D. Fedje 2002 Evidence of Relationships Between Microblades and Athapaskans. Paper presented at the International Conference \"Blade and Microblade Industries in Asia and America\", Vladivostok. Magne, M. and D. Ludowicz 1984 Features Expected in Recent Athapaskan Sites. Pp. 86-94 In Athapaskan and Earlier Archaeology and Big Eagle Lake,B1itish Columbia. Magne, M.and R.G. Matson Laboratory of Archaeology, University of B1itish Columbia,Vancouver. Magne, M. and R.G. Matson 1982 Identification of \"Salish\" and \"Athapaskan\" side-notched projectile points from the Interior Plateau of British Columbia. In Approaches to Algonquian Archaeology, ed. by M. Hanna and B. Kooyman, pp. 57-79. University of Calgary Archaeological Association. Magne, Martin P.R. and R. G.Matson 1984 Athapaskan and Earlier Archaeology and Big Eagle Lake,British Columbia. Laboratory of Archaeology, University of British Columbia,Vancouver. 1987 Projectile Point and Lithic Assemblage Ethnicity in InteriorBritish Columbia. In Ethnicity and Culture, ed. by ReginaldAuger, Margaret Glass, Scott MacEachern, and Peter McCattney, pp. 227 -242. The University of Calgary Archaeological Association, Calgary. Magne, Martin and Robe1t Tyhurst 1984 Chilcotin Ethnography. In Athapaskan and Earlier Archaeology at Big Eagle Lake, British Columbia., Magne and Matson, pp. 27-44. Matson, R.G. 1971 Adaptation and environment in the Cerbat Mountains. Arizona. Unpublished Ph.D. Dissertation, University of California, Davis .. 1974 The Determination of Archaeological Structure: an example from the Cerbat Mountains, Arizona. Plateau 47:26-40. 1976 The Glenrose Cannery Site, National Museum of Man Mercury Series, Archaeological Survey of Canada, No. 52, Ottawa. 1978 The Eagle Lake Archaeological Project. Social Sciences and Humanities Research Council of Canada research proposal. Submitted OctobeL 1978. May 20, 2003--472 1980a General considerations. In The Eagle Lake Project: reprn1 on the 1979 season, by R.G. Matson, M. Magne, D. Ludowicz and D. L. Pokotylo, pp. 3 - 10. Laboratory of Archaeology, University of B1itish Columbia, VancouveL 1980b Settlement pattern analysis. In The Eagle Lake Project: report on the 1979 season, by R.G. Matson, M. Magne, D. Ludowicz and D.L. Pokotylo, pp. 114 - 133. Laboratory of Archaeology, University of B1itish Columbia,VancouveL 1982 The Parallel Direct Historical Approach: Ethnic Identification at Eagle Lake, B.C. In Approaches to Algonquian Anthropology ed. by M. Hanna and B. Kooyman, pp. 233- 241. University of Calgary Archaeological Association. Calgary.1988 The Origins of the Plateau Pithouse Tradition -- or the Selah Springs Pattern 20 years later. Paper presented at the 21st Canadian Archaeological Association Meetings, Whistler, B.C. 1991 The O1igins of Southwestern Agriculture. University of Arizona Press, Tucson. 2002 The Spread of Maize to the U.S. Southwest. In Examining the Farming/Language Hypothesis, ed. by P. Bellwood, C. Renfrew and K.V.Boyle, McDonald Institute for Archaeological Research, Cambridge. Matson, R.G. and D. Alexander 1990 Potato Mountain: The Archaeology of Alpine root Resource Procurement. Paper presented at the Canadian Archaeological Association Meetings, Whitehorse. Matson, R.G. and Gary Coupland 1995 The Prehistory of the No11hwest Coast. Academic Press, Inc. San Diego. Matson, R.G., L. Ham and D. Bunyan 1984 Prehistoric settlement patterns at the Mouth of the Chilcotin River, B.C. Repmt to Heritage Conservation Branch, Victoria. Laboratory of Archaeology, University of British Columbia, Vancouver. Matson, R.G. and W. D. Lipe 1975 Regional sampling in Cedar Mesa southeastern Utah: a case .study. In Sampling in Archaeology, ed. by J. Mueller, pp. 124 - 143. University of Arizona Press, Tucson. Matson, R.G., W.D. Lipe, and W. Haase 1988 Adaptational Continuities and Occupational Discontinuites: The C edar Mesa Anasazi. Journal of Field Archaeology 15:245-264. Matson, R.G. and Martin P.R. Magne (Organisers) 2001 Migration: Then and Now. Symposium at the Society for Ame1ican Archaeology Annual Meetings, April 2001, New Orleans. May 20, 2003--473 Matson, R. G., Martin P. R. Magne, Deanna Ludowicz and D. L. Pokotylo 1980 The Eagle Lake Project: Report on the 1979 Season. Report to Heritage Conservation Branch, Victoria,. Laboratory of Archaeology, University of British Columbia. Vancouver. Matson, R.G., Martin Magne, and Marion Parker. 1988 Archaeological Identification of Ethnic Migration at Eagle Lake, British Columbia. MS in possession of Sr. Author. Matson, R.G. and David Pokotylo 1998 Chilcotin Plateau. In Archaeology of Prehistoric NorthAmerica: An Encyclopedia. Ed. by Guy Gibbone, pp. 146-148, Garland Publishing, New York. Matson, R.G. and D.L. True 1974 Site relationships at Quebrada Tarapaca, Chile: a comparison of clustering and scaling techniques. American Antiquity 39: 51 - 74. Mitchell, Donald H. 1970a Archaeological Investigations on the Chilcotin Plateau, 1968. Syesis 3: 45 - 66. 1970b Excavations on the Chilcotin Plateau: Three sites, three phases. Northwest Anthropological Notes 4: 99-1 16. Mohs, Gordon 1973 EkRo 48, A Preliminary Site Report. MS on file. Reading Room, Laboratory of Archaeology, University of British Columbia. Vancouver. Montgomery, Pamela. 1978 Stone artifacts from the Punchaw Lake site (Area C): a late prehistoric occupation in central British Columbia. Unpublished Masters thesis, Simon Fraser University, Burnaby. Moodie, D. Wayne, A. J. W. Catchpole, and KeJTy Abel 1992 Northern Athapaskan Oral Traditions and the White RiverVolcano. Ethnohistory 39:2: 148 - 171 Mooney, James 1898 Calendar History of the Kiowa Indians. Bureau of AmericanEthnology, Seventeenth Annual Repo1t 1895-96, ed.by J. W. Powell, pp. 129-440. Moratto, Michael J. 1984 California Archaeology. Academic Press. New York. Morgan, Lewis Henry 1877 Indan Migrations. In The Indian Miscellany. W.W. Beach (Editor), pp. 158-257, J. Munsell, Albany. May 20, 2003--474 Morice, A. 1890 The Western Denes: Their Manners and Customs. In: Proceedings of the Canadian Institute for 1888 - l 88\u00EF\u00BF\u00BD Series 3, Vol. 7 pp. 109 - 17 4, Toronto. 1893 Are the Carrier Sociology and Mythology Indigenous or Exotic? In Proceedings and Transactions of the Royal Society of Canada for the Year 1892, Seri es 1, Section 2, Vol. 10, pp. 109-126, Ottawa. 1895 Notes Archaeological, Industrial and Sociological on the Western Denes with an Ethnographical Sketch of the Same. Transactions of the Canadian Institute 4: 1 - 222. Toronto. 1906 The history of the northern interior of British Coi umbia 1660 - I 880. Ye Galleon Press, Washington. Morlan, Richard E. 1973 The Later Prehistory of the Middle Porcupine Drainag, Northern Yukon TeITitory . National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 11, Ottawa. Morlan, Richard E. 1988 Avonlea and Radiocarbon Dating. In: Avonlea Yesterday and Today: Archaeology and Prehistory, ed. by Leslie B. Davis, pp.291-309. Saskatchewan Archaeological Society. Saskatoon. Nash, Stephen E. (Editor) 2000 It's About Time: a history of archaeological dating in North America, University of Utah Press, Salt Lake City. Nelson, D.E. and G. Will 1976 Obsidian sources in the Anahim Peak area. In Current Research Reports, ed. by R. L. Carlson, pp. 151 - 154. Department of Archaeology Publication No. 3, SimonFraser University, Burnaby.Nelson, D.E., J.M. D' Auria and R.B. Bennett 1975 Characterization of Pacific Northwest Coast obsidian by X-ray fluorescence analysis. Archaeometry 17: 85 - 97. Nichols, Johanna 1997 Modeling Ancient Population Structures and Movement in Linguistics. Annual Reviews of Anthropology 26:359-384. (Ed. by W. H. Durham) Osgood, Cornelius B. 1932 The Ethnography of the Great Bear Lake Indians. National Museum of Canada Bulletin 70:31-98, Ottawa. May 20, 2003--475 O'Rourke, D.H., M.G. Hayes, and S.W. Carlyle 2000 Spatial and temporal stabiiity of mtDNA haplogroup frequencies in Native North Americans. Human Biology 72: 15-34. Paull, William 1984 Organic Residue Analysis of Stone and Bone Tools. In Athapaskan and Earlier Archaeology at Big Eagle Lake, British Columbia .\u00E2\u0080\u00A2 Magne and Matson, pp. 239-266. Peacock, Sandra L. 1998 Putting Down Root-;: The Emergence of Wild Plant Food Production on the Canadian Plateau. Unpublished Ph.D. dissertation, University of Victoria, Victoria. Peck, Trevor R. and John Ives 2001 Late Side-Notched Projectile Points in the Northern Plains. Plains Anthropologist 46: 163-193. Perry, Richard J. 1980 The Apachean Transition From the Subarctic to the Southwest. P!ain.s Anthropologist 25-90: 279 -296.Petitot, Emile 1886 Les Litteratures Populaires De Toutes Les Nations. Traditions Indiennes Du Canada Nord-Quest Vol. 23. G.P. Maisonneuve etLarose. Paris. Pettigrew, Richard M. 1980 Archaeological Investigations at the Looney Site (35DO13 ), Douglas County; Oregon. University of Oregon Anthropological Papers No. 18, Eugene. Pettigrew, Richard M. and Clayton G. Lebow 1987 Data Recovery at Sites 35JA27, 35JA59, and 35JA 100, Elk CreekLake Project, Jackson County, Oregon. Report to U.S. Army Corps otEngineers, Portland District. Portland. Plog, S. 1976 Relative efficiencies of sampling techniques for archaeological surveys. In The Early Mesoamerican Village, ed. by K. Flannery, pp. l 36 - 160. Academic Press. New York. Pokotylo, David L. 1978 Lithic technology and settlement patterns in Upper Hat Creek Valley, B.C. Unpublished Ph.D. dissertation, University of British Columbia. Vancouver. 1981 Towards an understanding of prehisto1ic upland settlement behaviour in the British Columbia Interior Plateau. In Networks of the past: Regional interaction in archaeology, ed. by P.O. Francis, F.J. Kense, and P.G. Duke, pp. 379-396. Archaeological Association of the University of Calgary. Calgary. May 20, 2003--476 Pokotylo, D.L. and P. Froese 1983 The archaeological evidence for root gathering in prehistoric plateau settlement -subsistence patterns: a case study from the Upper Hat Creek Valley, B.C. Canadian Journal of Archaeology 7: 127-58. Pokotylo, David and Donald Mitchell 1998 Prehistory of the Northern Plateau. In Handbook of North American Indians. Vol. 12, Plateau, ed. by Deward E. Walker, Jr., pp.81-102. Smithsonian Institution. Washington. Prince, Paul 2001 Dating and Interpreting Pine cambium coilection scars from two pans of the Nechako River draingage, B1itish Columbia. Journal of Archaeological Science 28(3): 253-263. Pyszczyk, Heinz 1991 A Wedge-Shaped Microblade Core, Fo11 Vermilion, Alberta. In:Archaeology in Albe11a 1988 and 1989, edited by Martin Magne, pp. 199- 203. Archaeological Survey, Provincial Museum of Alberta, Occasional Paper No. 33. Edmonton. Ray, V.F. 1939 Cultural Relationships in the Plateau of No11h-Western America Publications of the Frederick Webb Hodge Anniversary Publication Fund, Vol 3, Southwestern Museum, Los Angeles. 1942 Culture element distributions XXII: Plateau. University of California Anthropological Records 8 (2). Berkeley. Reher, Charles A. and George C. Frison 1980 The Vore Site, 48CK302: A Stratified Buffalo Jump in the Wyoming Black Hills. Plains Anthropological Memoir 16. Renfrew, Colin 1988(1987] Archaeology and Language: The Puzzle of lndo-European O1\u00C2\u00B7igins. University of Cambridge Press, New York. Renfrew, Colin 2000 At the Edge of Knowability: Towards a Prehistory of Languages. Cambridge Archaeology Journal 10:7-34. Richards, Thomas H. and Michael K. Rousseau 1987 Late Prehisto1ic Cultural Horizons on the Canadian Plateau. Dept of Archaeology Publication No. 16, Simon Fraser University. Burnaby. May 20, 2003--477 Ridington, Amber and Michael Klassen 1998 Archaeological Site EjSa-11, Chilko River, B.C. Archaeological Impact Assessment. Report on Permit 1997-266, Archaeoiogy Branch, Ministry of Community, Aboriginal & Women's Services, Victoria. Ritter, Eric and R.G. Matson 1972 Form categories, cluster analysis and multidimensional scaling: a case of projectile points. Southwestern Lore 34: 102 -116. Rostlund, Erhard 1952 Freshwater Fish and Fishing in Native North America, University of California Publications in Geography, Volume 7. University of California Press, Berkeley. Rouse, Irving 1986 Migrations in Prehistory: InfeJTing population movement from cultural remains. Yale University Press, New Haven. Rousseau, Michael K. and Thomas H. Richards 1985 A culture-histoical sequence for the South Thompson River-Western Shuswap Lakes Region of British Columbia: the last 4000 years. Northwest Anthropological Research Notes 19(1): 1-32. Rousseau, Michael K., R. Muir, D. Alexander, J. Breffit, S. Woods, K. Berry and T. Van Gallen 1991 Results of the 1989 Archaeological investigations conducted in the Oregon Jack locality, Thompson River region, south-central British Columbia. Permit 1989-76 report. Archaeology Branch, Victoria. Ruggles, Richard I. 1991 A Country So Interesting. The Hudson s Bay Company and TwoCenturies of Mapping, 1670 - 1870 McGill-Queen s University Press.Montreal. Sanger, David 1967 Prehistory of the Pacific N011hwest Plateau as seen from the Interior of British Columbia. American Antiquity 32: 186-197. 1968 The High River Microblade Industry. Plains Anthropologist 13(41 ): 190-208. 1970 The Archaeology of the Lochnore - Nesikep Locality. Syesi.s. Yol.3, Supplement 1. Schaafsma, Curtis 1996 Ethic Identity and Protohisoric Archaeological Sites in Northwestern New Mexico: Implications for Reconstructions of Navajo and Ute History. Pp. 19-47 The Archaeology of Navajo Origins, ed. by Ronald Towner, University of Utah Press, Salt Lake City. 2002 Apaches de Navajo: Seventeenth-century Navajos in the Chama Valley of New Mexico. University of Utah Press, Salt Lake City. May 20, 2003--478 Schlesier, Karl H. 1972 Rethinking the Dismal River Aspect and the Plains Athapaskans, A.O. 1692- 1768. Plains Anthropologist 17: 101-133. Schulting, Rick 1995 Mortuary Variability and Status Differentiation on the Columbia-Fraser Plateau. Archaeology Press, Simon Fraser University, Burnaby. Sesler, Leslie, Timothy Hovezak, and Richard Wilshusen 2000 Cultural Landscape of Dinetah: The Navajo Occupation of Frances Mesa. Pp 159-253 in Frances Mesa Altenative Treatment Project, compiled by Richard Wilshusen, Timothy Hovezak and Leslie Sesler, La Plata Archaeological Consultants, Research Papers No. 3, Volume 1, Dolores, Colorado. Shinkwin, Anne D. 1979 Dakah de nin s Village and the Dixthada Site: A Contribution to No11hern Athapaskan Prehistory. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 91, Ottawa. Smith, Gerald R. and Ralph F. Stearley 1989 The Classification and Scientific Names of Rainbow and Cutthroat Trouts. Fisheries 14(1):4-10. Smith, H.I. 1900 Archaeology of the Thompson River region. British Columbia. American Museum of Natural History, Jesup No11h Pacific Expedition Memoirs Vol. 2, Part 6: 401 - 442. Sneath, P.H.A. and R.R. Sokal 1973 Numerical Taxonomy W.H. Freeman, San Francisco. Sneed, Paul G. 1971 Of salmon and men: An investigation of ecological determinants and aboriginal man in the Canadian Plateau. In Abo1iginal man and environments on the plateau of N011hwest America, ed. by Arnoud Stryd and Rachel Smith, pp. 229-242. University of Calgary Archaeology Association, Calgary. Spielmann, Katherine A 1983 Late Prehistoric Exchange Between the Southwest and Southern Plains. Plains Anthropologist 28 (102), Part l: 257 - 272 Spurling, Brian E. 1978 The Site C he1itage resource inventory and assessment. Depal1ment of Archaeology, Simon Fraser University. Rep011 to B.C. Hydro and Power Authority. May 20, 2003--479 1980 The Site C Heritage Resource Inventory and Asse.s.sment, Final Repo11. Report on file, B.C. Heritage Conservation Branch, Victoria.Stevenson, Marc 1986 Window On the Past: Archaeological Assessment of the Peace Point Site, Wood Buffalo National Park, Alberta. Studies in Archaeology,Architecture and History. Parks Canada. Ottawa. Steward, Julian 1977 The direct historical approach to archaeology. In Evolution and Eco logy, by J. Steward and ed. by J.C.S. Steward and R.F. Murphy, pp. 201 - 207. University of Illinois Press. Urbana. (Originally published in American Antiquity 7:337-343, 1942.) Stewart, Frances L. 1978 Ve1tebrate Fauna! Remains from the Potlatch Site (FcSi-2) in South Central British Columbia. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 82, Ottawa. Stryd, Arnoud 1971 An Archaeological Research Design for the Investigation of Pithouse Archaeology. In Aboriginal Man and Environments on the Plateau of Northwest America., pp. 36-43, ed. by A. Stryd and R. Smith, University of Calgary Archaeological Association, Calgary. 1973 The Later Prehistory of the Lillooet Area, British Columbia. Unpublished PhD Disse1tation, Depa1tment of Archaeology, University of Calgary, Calgary. 1974 Lillooet Archaeological Project: 1974 Field Sea.son. Cariboo College Papers in Archaeology no. 1. Kamloops. 1995 Archaeological Evidence for Aboriginal use of Scheidam Flats near Karnloops, B.C. Report preparedfor Legal Services Branch, Ministry of Attorney General, Victoria. Stryd, Arnoud and Stephen Lawhead 1978 Reports of the Lillooet Archaeological Project Number I. Introduction and .setting . National Museum of Man Mercury Series, Archaeological Survey of Canada 73, Ottawa. 1983 Bethlehem Copper Corporation Lake Zone Development Heritage Mitigation Study. Report on Pe1mit No. 82-19. British Columbia Archaeology Branch, Victoria. Stryd, Arnoud and Michael Rousseau 1996 The Early Prehistory of the Mid Fraser-Thompson River of British Columbia. In E arly Human Prehistory in B.C. ed. by R. Carlson and Luke Dalla Bona, pp. 177-204 University of British Columbia Press, Vancouver. May 20, 2003--480 Stuiver, M. 1978 Radiocarbon time.scale tested against magnetic and other dating methods. Nature 273: 271 - 274. Stuiver, M., Reimer, P.J., Bard, E., Beck, J.W., Bun-, G.S .\u00E2\u0080\u00A2 Hughen, K.A., Kromer, B., McCormac, F.G., v.d. Plicht, J., and Spruk, M. 1998 INTCAL98 Radiocarbon Age Calibration, Radiocarbon 40(3): 1041- 108 L Tabachnick, Barbara and Linda Fidell 2001 Using Multivariate Statistics, 4th ed. Allyn and Bacon, Boston. Tasa, Guy L. 1997 Skeletal and Dental Variation of Pacific Coast Athapaskans: Implications for Oregon Prehistory and Peopling of the New World. Unpublished PhD dissertation, University of Oregon, Eugene. Teit, James A. 1900 The Thompson Indians of British Columbia. Memoirs of the American Museum of Natural History, No. 2, Pa11 4: 163-392. 1906 The Lillooet Indians. American Museum of Natural Hi.story, Jesup North Pacific Expedition, Memoirs Vol. 2, Part 5: 193 - 300. 1909a The Shu.swap. Memoirs of the American Museum of Natural Histo1\u00C2\u00B7y, Jesup N011h Pacific Expedition, Memoirs, Vol. 2, Part 7: 443 - 758. New York. 1909b Notes on the Chilcotin Indians. In \"The Shu.swap''. American Museum of Natural Histmy, Jesup North Pacific Expedition, Memoirs Vol. 2, Patt 7: 759 - 789. New York. 1928 The Salishan tribes of the western Plateaus. Extract from the 45th Bureau of American Ethnology repo11 1927-28. Facsimile reproduction 1973. The Shorey Book Store,Seattle. 1979 Notes on the Early History of the Nicola Valley. Abhandlungun der Volkerkundlichen Arbeitsgemeinshaft. (Bureau of AmericanEthnology manuscript No. 4581). N011orf. Thomas, D.H. 1969 Regional sampling in archaeology: a pilot Great Basin research design. In Archaeological Survey Annual Repo11 11: 87 - 100. University of California, Los Angeles, Los Angeles. Thompson, Raymond H. (Editor) 1958 Migrations in New World Culture History. University of Arizona Social Science Bulletin no. 27, Tucson. May 20, 2003--481 Tipper, H.W. 1971 Glacial geomorphology and Pleistocene history of central British Columbia. Geological Survey of Canada Bulletin No. l 96. Torgerson, W. 1958 Theory and methods of scaling. John Wiley, Toronto. Towner, Ronald H. (Editor) 1996 The Archaeology of Navajo O1igins. University of Utah Press, Salt Lake City. Towner, Ronald H. 2000 Dendrochronology and Hist01ical Records: Concordance and Conflict in Navajo Archaeology. Pp. 168-185 inlt's About Time: a history of archaeological dating in North America, ed. by Stephen Nash, University of Utah Press, Salt Lake City. Trigger, Bruce 1968 Beyond History: The methods of prehistory. Holt.Rinehard and Winston, New York. Turner, Nancy J. 1978 Food Plants of British Columbia Indians. Part 2,'Interior Peoples'. Brtish Columbia Provincial Museum Hanbook No. 38. Victoria. Tveskov, Mark A. 2000 The Coos and Coquille: A Northwest Coast Historical Anthropology. Unpublished Ph.D. Dissertation, University of Oregon. Eugene. Tyhurst, Robert 1984 An economic history of the Chilcotin Indians of British Columbia. Unpublished Ph.D. Dissertation Draft. MS in possesion of author. Tyhurst, Robe1t 1996 (1993) Fish Lake Heritage Resources Study: Report on the 1993 Archaeological Survey of the Fi.sh Lake Mine Project and Access Corridor inSouth Central British Columbia. Penn it 1993 - 95. Archaeology Branch,Ministry of Tourism and Culture, Vict01ia. U garenko, Leonard G. 1979 The Beaver Indians and the Fur Trade on the Peace River 1700-185(1 Unpublished Masters thesis, York University, Toronto. Vanags, Anthony 2000 An Archaeological Perspective on Alpine/Sub-Alpine Land Use in the Clear Range and Pavilion Mountains, South-Central British Columbia. Unpublished Masters thesis, University of British Columbia, Vancouver. May 20, 2003--482 Van de Geer, J.P. 1971 Introduction to multivariate analysis for the social sciences. W.H. Freeman and Co. San Francisco. Walker, Deward E., Jc (Editor) 1998 Handbook of North American Indians, Vol. 12, Plateau, Smithsonian Institution. Washington. Wedel, W. 1986 Central Plains Prehistory: Holocene Environments and Cuiture Change in the Republican River Ba.sin. Univeri.sty of Nebraska Press, Lincoln. 2001 Plains Village Tradition:Central. In Handbook of North American Inidans, Vol.13, Plains, ed. by R.J. DeMallie, pp. I 73-i 85. Smithsonian Institute, Washington. West, Frederick Hadleigh 1967 The Donnelly Ridge Site and the Definition of an Early Coreand Blade Complex of Central Alaska. American Antiquity 32: 360 - 382. Whistler, Kenneth W. 1979 Linguistic prehistory in N011hwest California Culture Area. In A Study of Cultural Resources in Redwood National Park, by P.M. Bickel and Others, pp. 11-26, National Park Service (as cited in Connolly 1986). Whitlam, R. 1976 Archaeology in the Williams Lake area, British Columbia . Occasional Papers of the Archaeological Sites Advisory Board of British Columbia No. I. Victoria. Wilcox, David 1981 The Entry of Athapaskans into the American Southwest: The Problem Today. In The Protohistoric Period in the North Amerian Southwest, A.O. 1450-1700. Ed. by D. Wilcox and W.B. Masse, pp. 213-256. Arizona State University Anthropological Research Papers No. 24, Tempe. 1988 Avonlea and Southern Athapaskan Migrations. In Avonlea Yesterday and Today, edited by Leslie B. Davis, pp. 273-280. Saskatchewan Archaeological Society, Saskatoon. 2001 Avonlea as Early Apacheans: A Re-evaluation. Paper presented at symposium Migration Then and Now, 66th Annual Meeting of the Society for American Archaeology, New Orleans. April, 2001. Wilshusen, Richard, Timothy Hovezak, and Leslie Sesler, Compilers 2000 Frances Mesa Altenative Treatment Project, La Plata Archaeological Consultants, Research Papers No. 3, Volume I. Dolores, Colorado. May 20, 2003--483 Willey, G. and P. Phillips 1958 Method and theory in American archaeology. University of Chicago Press, Chicago Willey, G. (Editor) 1956 Prehistoric settlement patterns in the New World. Viking Fund Publications in Anthropology No. 23. New York. Wilmeth, Roscoe 1969 The Potlatch site, British Columbia: a preliminary report. Kansas Anthropological Association Newsletter 14(7):2-9. 1971 Historic Chilcotin Archaeology at Anahim Lake, British Columbia. In Aboriginal man and environments on the plateau of Northwest America, ed. by Arnoud Stryd and Rachel Smith, pp. 56-59. University of Calgary Archaeology Association, Calgary. 1975 The Protohistoric and Historic Athapaskan Occupation of British Columbia: The Archaeological Evidence. Western Canadian Journal of Anthropology 5(3-4):4-20. 1977 Pit-house construction and the disturbance of stratified sites. Canadian Journal of Archaeology I: 135 - 150. 1978 Anahim Lake Archaeology and the Early Histo,ic Chilcotin Indians . National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 82. Ottawa. 1979 An Athapaskan Hypothesis. Canadian Journal of Archaeology 3:33- 40. Wilson, Ian R. 1978 Archaeological Investigations at the Atigun Site, CentralBrooks Range, Alaska. National Museum of Man Mercury Series, Archaeologicai Survey of Canada Paper No. 78. Ottawa.Wilson, E.F. 1888 Report on the Sarcee Indians. In Repm1 on the No1th-Western Tribes of Canada. pp. 10 - 23, with notes by H. Hale.British Association for the Advancement of Science. London Wissler, F. Clark 1917 [1938] The Relation of Nature to man in Aboriginal America. McMu,tie, New York. Wood, John J. 1967 Archaeological Investigations in NortheasternColorado. Unpublished Ph.D. dissertation, University of Colorado, Boulder. Wood, W. Raymond (Editor) 1998 Archaeology on the Great Plains. University Press of Kansas, Lawrence. May 20, 2003--484 W orlanan, William B. 1978 Prehistory of the Aishihik-Kluane Area, Southwest YukonTen-itory. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 74. Ottawa. 1979 The significance of volcanism in the prehistory of subarctic northwest North America. In Volcanic activity and human ecology, ed. by P.O. Sheets and D.K. Grayson, pp. 339 - 37L Academic Press, NewYork. Wright, James V. 1975 The Prehistory of Lake Athabasca: An Initial Statement. National Museum of Man Mercury Series, Archaeological Survey of Canada Paper No. 29. Ottawa. Wyatt, David. 1971 A preliminary outline of Nicola Valley prehistory. In Aboriginal man and environments of the Plateau of Northwest America. Ed. by A. H. Stryd and R.A. Smith, pp. 60 -72. University of Calgary Archaeological Association.1972 The Indian history of the Nicola Valley. Unpublished Ph.D. disse11ation, Brown University, Providence. 1998 Nicola. In Handbook of North American Indians, Vol. 12, Plateau, ed. by Deward E. Walker, Jr. pp. 220 - 223. Smithsonian Institution, Washington.May 20, 2003--485 "@en . "Book"@en . "Preprint"@en . "10.14288/1.0363461"@en . "eng"@en . "Unreviewed"@en . "Vancouver : University of British Columbia Library"@en . "Attribution-NonCommercial-NoDerivatives 4.0 International"@* . "http://creativecommons.org/licenses/by-nc-nd/4.0/"@* . "Faculty"@en . "Athapaskans and Migrations : The Archaeology of Eagle Lake, British Columbia"@en . "Text"@en . "http://hdl.handle.net/2429/64600"@en .