Dy ben i ii at Ny we _ SMITHSONIAN INSTITUTION Ke Me "BUREAU OF AMERICAN ETHNOLOGY A a ‘BULLETIN 189 ( BN APs" lA) AS 4 F | & RIVER BASIN SURVEYS PAPERS : "Frank H. HL Ronen, Jr., Editor. ) Tote Ageney Archeological lease ErOgrem NUMBERS 33-38 ; AEE INCRE Siac AS ace Ato 9 ony SMITHSONIAN INSTITUTION BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 RIVER BASIN SURVEYS PAPERS Frank H. H. Roserts, Jr., Editor Inter-Agency Archeological Salvage Program NUMBERS 33-38 U.S. GOVERNMENT PRINTING OFFICE WASHINGTON : 1964 ee ee ee For sale by the Superintendent of Documents, U.S. Government Printing Office Washington D.C., 20402 + Price $3.75 cloth i OED ENED a a LRG t LETTER OF TRANSMITTAL SMITHSONIAN INSTITUTION, Bureau or AMERICAN ETHNOLOGY, Washington, D.C., September 28, 1962. Sir: I have the honor to transmit the accompanying manuscripts, entitled “The Paul Brave site (82SI4), Oahe Reservoir area, North Dakota,” by W. Raymond Wood and Alan R. Woolworth; “The Demery site (839CO1), Oahe Reservoir area, South Dakota,” by Alan R. Woolworth and W. Raymond Wood; “Archeological investigations at the Hosterman site (89PO7), Oahe Reservoir area, Potter County, South Dakota, 1956,” by Carl F. Miller; “Archeological in- vestigations at the Hickey Brothers site (39LM4), Big Bend Reservoir, Lyman County, South Dakota,” by Warren W. Caldwell, Lee G. Madison, and Bernard Golden; “The Good Soldier site (89L.M238), Big Bend Reservoir, Lyman County, South Dakota,” by Robert W. Neuman; “Archeological investigations in the Toronto Reservoir area, Kansas,” by James H. Howard, and to recommend that they be pub- lished as a bulletin of the Bureau of American Ethnology. Very respectfully yours, Frank H. H. Ropzerrs, Jr., Director. Dr. Leonard CARMICHAEL, Secretary, Smithsonian Institution. EXPLANATION OF THE INTER-AGENCY ARCHEOLOGICAL SALVAGE PROGRAM The Inter-Agency Archeological Salvage Program is a cooperative plan of the Smithsonian Institution; the National Park Service and the Bureau of Reclamation, Department of the Interior; and the Corps of Engineers, Department of the Army. It was formulated, through a series of interbureau agreements, for the purpose of recov- ering archeological and paleontological remains that would other- wise be lost as a result of the numerous projects for flood control, irrigation, hydroelectric power, and navigation improvements in the river basins of the United States. Various State and local agencies have assisted in the work. To carry out its part of the joint under- taking, the Smithsonian Institution organized the River Basin Sur- veys as a unit of the Bureau of American Ethnology. The National Park Service has served as liaison between the various agencies and has provided the Smithsonian Institution with all of the necessary information pertaining to the location of proposed dams and other construction and their priorities. It has also had responsibility for budgeting costs of the program, funds for which are provided in the annual appropriations of the Department of the Interior. The opera- tions of the River Basin Surveys, Smithsonian Institution, have been supported by funds transferred to it from the National Park Service. Through agreements with the National Park Service, money has also been made available to State and local agencies to supplement their own resources and aid them in their contributions to the program. The River Basin Surveys Papers, of which this is the ninth bulletin, are issued under the scientific editorship of Frank H. H. Roberts, Jr., director of the Bureau of American Ethnology. PUBLISHER’S NOTE A separate edition is published of each paper in the series entitled “River Basin Surveys Papers.” Available copies of Papers 1-38 can be had upon re- quest to the Publications Office, Smithsonian Institution, Washington, D.C., 20560. No. No. RiIveR BAsIn SuRVEYS PAPERS PUBLISHED PREVIOUSLY 1. Prehistory and the Missouri Valley Development Program: Summary Report on the Missouri River Basin Archeological Survey in 1948, by Waldo R. Wedel. Bull. 154, pp. xv—xviii, 1-59, pls. 1-12, fig. 1. 1953. . 2. Prehistory and the Missouri Valley Development Program: Summary Report on the Missouri River Basin Archeological Survey in 1949, by Waldo R. Wedel. Bull. 154, pp. 61-101, pls. 18-15. 1953. . 8. The Woodruff Ossuary, a prehistoric burial site in Phillips County, Kansas, by Marvin F. Kivett. Bull. 154, pp. 103-141, pls. 16-28, figs. 2-3. 1953. . 4. The Addicks Dam site: I. An archeological survey of the Addicks Dam Basin, Southeast Texas, by Joe Ben Wheat. Bull. 154, pp. 143-252, pls. 29-47, figs. 4-23. 1953. II. Indian skeletal remains from the Doering and Kobs sites, Addicks Reservoir, Texas, by Marshall T. Newman. Bull. 154, pp. 253- 266, figs. 24-28. 19538. . 5. The Hodges site: I. Two rock shelters near Tucumari, New Mexico, by Herbert W. Dick. Bull. 154, pp. 267-284, pls. 48-54, figs 29-30. 1953. II. Geology of the Hodges site, Quay County, New Mexico, by Sheldon Judson. Bull. 154, pp. 285-802, figs. 31-85. 1953. . 6. The Rembert Mounds, Elbert County, Georgia, by Joseph R. Caldwell. Bull. 154, pp. 303-820, pls. 55-56, figs. 36-40. 1953. . 7%. Archeological investigations in the Oahe Dam area, South Dakota, 1950-51, by Donald J. Lehmer. Bull. 158, 190 pp., 22 pls., 56 figs., 6maps. 1954. . 8. Excavations in the McNary Reservoir Basin near Umatilla, Oregon, by Douglas Osborne. With appendixes by Marshall T. Newman, Arthur Woodward, W. J. Kroll, and B. H. McLeod. Bull. 166, 250 pp., 40 pls., 6 figs.,19 maps. 1957. 9. Archeological investigations in the Heart Butte Reservior area, North Dakota, by Paul L. Cooper. Bull. 169, pp. 1-40, pls. 1-12, figs 1-2. 1958. . 10. Archeological investigations at the Tuttle Creek Dam, Kansas, by Rob- ert B. Cumming, Jr. Bull. 169, pp. 41-78, pls. 18-24. 1958. . 11. The Spain site (89LM301), a winter village in Fort Randall Reservoir, South Dakota, by Carlyle S. Smith and Roger T. Grange, Jr. Bull. 169, pp. 79-128, pls. 25-36, figs 3-4. 1958. Iv No. No. . 12. abe . 14. . 15. . 16. . 18. ra i 5 OP » 20. . 25. . 26. er eiie 28. 29. PUBLISHER’S NOTE Vv The Wilbanks site (9CK-—5), Georgia, by William H. Sears. Bull. 169, pls. 37-45, figs 5-9. 1958. Historic sites in and around the Jim Woodruff Reservoir area, Florida- Georgia, by Mark F. Boyd. Bull. 169, pp. 195-314, pls. 46-55, figs. 10- 11. 1958. Six sites near the Chattahoochee River in the Jim Woodruff Reservoir area, Florida, by Ripley P. Bullen. Bull. 169, pp. 315-357, pls. 56-73, figs. 12-13. 1958. Historie sites archeology on the Upper Missouri, by Merrill J. Mattes. Bull. 179. pp. 1-24. Historie sites archeology in the Fort Randall Reservoir, South Dakota, by John E. Mills. Bull. 176, pp. 25-48, pls. 1-9, figs. 1-2, map 1. 1960. . 17. The excavation and investigation of Fort Lookout Trading Post II (39LM57) in the Fort Randall Reservoir, South Dakota, by Carl F. Miller. Bull. 176, pp. 49-82, pls. 10-18, figs. 3-14, map 2. 1960. Fort Pierre II (39ST217), a historic trading post in the Oahe Dam area, South Dakota, by G. Hubert Smith. Bull. 176, pp. 83-158, pls. 19-30, maps 3-4. 1960. Archeological investigations at the site of Fort Stevenson (32ML1), Garrison Reservoir, North Dakota, by G. Hubert Smith. (Introduc- tion by Robert L. Stephenson and an appendix by Carlyle S. Smith.) Bull. 176, pp. 159-238, pls. 31-54, figs. 15-20, maps 5-6. 1960. . The archeology of a small trading post (82MN1) in the Garrison Reser- voir (Kipp’s Post), South Dakota, by Alan R. Woolworth and W. Raymond Wood. Bull. 176, pp. 239-305, pls. 55-65, figs. 21-25, map 7. 1960. . Exeavations at Texarkana Reservoir, Sulphur River, Texas, by Edward B. Jelks. Bull. 179, pp. xiii-78, pls. 1-17, figs. 1-9. 1961. Archeological investigations at the Coralville Reservoir, Iowa, by War- ren W. Caldwell. Bull. 179, pp. 81-148, pls. 18-29, figs. 10-24. 1961. The McNary Reservoir: A study in plateau archeology, by Joel L. Shiner. Bull. 179, pp. 151-266, pls. 30-46, figs. 25-40, maps 1-7. 1961. . The Sheep Island site and the Mid-Columbia Valley, by Douglas Osborne, Alan Bryan, and Robert H. Crabtree. Bull. 179, pp. 269-321, pls. 47-56, figs. 41-43. 1961. Archeology of the John H. Kerr Reservoir Basin, Roanoke River, Virginia-North Carolina, by Carl F. Miller. (Appendix by Lucile E. Hoyme and William M. Bass.) Bull. 182, 447 pp., pls. 1-110, figs. 1-65, maps, 1-8. 1962. Small sites on and about Fort Berthold Indian Reservation, Garrison Reservoir, North Dakota, by George Metcalf. Bull. 185, pp. 1-56, pls. 1-11, figs. 1-5, 1 map. 1962. Star Village: A fortified historic Arikara site in Mercer County, North Dakota, by George Metcalf. Bull. 185, pp. 57-122, pls. 12-17, figs. 6-16, 8 maps. 1962. The dance hall of the Santee Bottoms on the Fort Berthold Reservation, Garrison Reservoir, North Dakota, by Donald D. Hartle. Bull. 185, pp. 123-132, pl. 18, figs. 17-18. 1962. Crow-Flies-High (32MZ1), a historic Hidatsa village in the Garrison Reservoir area, North Dakota, by Carling Malouf. Bull. 185, pp. 133- 166, pls. 19-26, figs. 19-25. 1962. VI No. PUBLISHER’S NOTE . 80. The Stutsman Focus: An aboriginal eulture complex in the Jamestown Reservoir area, North Dakota, by R. P. Wheeler. Bull. 185, pp. 167— 238, pls. 27-36, figs. 26-38. 1962. . 81. Archeological manifestations in the Toole County section of the Tiber Reservoir Basin, Montana, by Carl F. Miller. Bull. 185, pp. 235-255, pls. 37-45, figs. 39-40, 1 map. 1962. No. 32. Archeological salvage investigations in the Lovewell Reservoir area, Kansas, by Robert W. Neuman. figs. 41-48. 1962. Bull. 185, pp. 257-306, pls. 46-57, CONTENTS No. 33. The Paul Brave site (32514), Oahe Reservoir area, North Dakota, by W. Raymond Wood and Alan R. Woolworth_----...--.---------- No. 34. The Demery site (89CO1), Oahe Reservoir area, South Dakota, by Alan R. Woolworth and W. Raymond Wood. -.-_.......----------. No. 35. Archeological investigations at the Hosterman site (89PQ7), Oahe Reservoir area, Potter County, South Dakota, 1956, by Carl F. No. 36. Archeological investigations at the Hickey Brothers site (39LM4), Big Bend Reservoir, Lyman County, South Dakota, by Warren W. Cald- well, ibee G. Madison, and Bernard Golden..-.2---..--..2.2sc-222es< No. 37. The Good Soldier site (39LM238), Big Bend Reservoir, Lyman County, South Dakota, by Robert W. Neuman-_-_-__--_--__---_------ No. 38. Archeological investigations in the Toronto Reservoir area, Kansas, ivan euln CS MEO Wie Ceca = a etches ere eV rate ah eye 28 oy oe ape es List of reports, articles, and notes relating to the salvage programs pub- HIESSTEVI SPV Sin. ty FLAPS M7 ci ales SN ge RI ae eer nee ae SR SR BT ANA Ne PAGE ) inate aevctioveieoe® ofiG (9eet) ‘Win teeta = zs th eh. ele. coor WE ral. er hoo’! brome : totnG. dinek gre 164 nih 4 at te (tO U2) athe cet of ig I ; P es Oy 2) e dives SL OW bas doco fooW uf sath (FOURS) wie mpanmmdeol, wht fe pines Set iexigolonionk BE 4 hot vd OOP wicuall dives lave wtioT wow tiovmeshfi i § “as HET y hs SiG ae ee eal oats ; : CEM. LYTO é tint solid edd te 2 o BS > hie) WF need pel ofedath: ha WY ied pees Tag - 4. ee ; f I , 2 hy a, pitt gay © 4 a ros Bow Riot seetwo = Lae .W i AS) FS . , OF tod ahs ert a4 a od he : Pa | : H : t fant nipais: sod DA tys nae ; 3 h : i is = bee ah Pa J) fo ae sine p ae Pr. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 189 River Basin Surveys Papers, No. 33 The Paul Brave Site (32SI4), Oahe Reservoir Area, North Dakota By W. RAYMOND WOOD and ALAN R. WOOLWORTH Ix iy ae ee ST aN aS vt. MOUWUTTeK) “ATHORIY IME tyolonita nastemA Yo poondl ; —. ext mitefiot aes: ee && oi 21sqs4 eqorme, tien z97il atk Niovioesii silnO Resta 28) tif svar tox oft giaisG dyn. HINO WIGOW 3 VATA bas GOOW GVOMYART .W a CONTENTS PAGE Prefacess& O82 580s helps em Peeled SEN Bris Ieee ee eee XIII Introduction. =< oo ae en dee ae ek aes by 98 1 Archeology of the site BRPCIMCACIVO (HOE ee otic; ie ean ee ha ee ees ae en 2 Hixcavation methods = 2_- Saeeas rey ie. Mr anne Se ee aes 2 eApUTESEee seee ee coe tees OSes 2 PR SU. ee Mobs Bae 4 Excavation unit 1 (house 1) 4 (i Excavation unit 2 (house 2) Hxcavation: Unit.o (MOUSe: 6) 2-2-2 52-20 pee 9 xeavation. Unit 4. 5-500 2500 seteeel Syren hen ines Se 11 Excavation Unit. b. -:.--- | ssaeVeie anate SMe Bo he a 11 Hxcavation unit G22. Ie es seek ei ee ee ee ed aly ee AT) BxcaVveuoncuunit: (2 s22= 4. Oe hy Stee Tae ee 11 Bxcavation unit 8 (house 4)_..........2. 2.2 pd deol: 12 AINISUIU DUTTA seen ee oe eh Ls es og soe Ce Sei 12 AEE SUC US Se eco eae ee ee ates ees ee erage eee 14 Pottery types. 2... 22. 222.2 beat thas has ese Se ctest 14 Rpes ware. 028. 29. kee 2 a eS Ge 16 Hone VatCsawale ste eee Bee Lee ee 20 Wmelassified awaresi = 40) sneer ee 21 Miniature vessels... 2 =... 2 en _ Bae aa eva ie 22 Body sherds= ==." = 50.2. 228 eiiieehers Selarons pers 23 Miscellaneous objects of baked clay________------------------ 24 WorkJntatone?s eens Shy ee 1) 2 aa en. 8 gents 25 Worksinsbones 222-228 2. ee ee naa e 2 ere 35 Wiorkanfantler‘andsteeth=—-2 22 98. eae ee eee eee 45 Workanishells 22 s-2 252 eee Ess Gis oe hd Sn BRT eed, TOES 48 Wie esod rr crete osname na nye eee, Bone Sa eee eee ees 50 Unmoditvedtbone andishelll fas. 2 Uses Sees ie 50 WROTE own cek Ties nd. Sete ne bo fe 8 ge pepe 28 ee pe 51 PSU GHEE OS ee ee per oy eR 1a te oo ge ees po tee BI POLIS G OMNES x. cpanel te eek ys FO gh raed 52 CEN OTS SGC ere he ge ee ae re re ete Ee een ee te GY Ropert Zalai site (ozo). oe. Sere ee Se Ee ee 57 Havens: site (och MEE) oes 2 ye) Fe oS a oe ee 2 60 MERTENS orn cy bec PAE yt ole ks Ee ies BOE 2 61 WIRE AIEROR ECE Sc 22 Ses ee ea 8 2 ek Ne ik ee en a eee ee se 63 TABLES 1; MesGrisienien Ten lUren. oW2=) 22S. Ao ee eh Oe ee oe 13 2. Seven species of fossil shells examined______-------------- Crh. gin 50 gun opeciesnuentined at: Panl Braves 2-628. 32 aie) She pee des 51 4. Pottery frequencies at Thomas Riggs and Paul Brave sites___-___--- 55 5. Comparison of traits at Paul Brave and Thomas Riggs sites---__ ---- 58 xI Oar ON or ON CONTENTS ILLUSTRATIONS PLATES (All plates follow page 66) . a, Aerial view of the Paul Brave site and environs. 6, House 2. . a, Feature 70 in House 3. 6, Burned timbers in House 3. Pottery rim sherds. . Crosshatched rim sherds. Pottery examples A-D, and cord-impressed rim sherds. . Pottery disks, body sherds, and vessels. TEXT FIGURES ,. Pottery design elements 5-42-22. 2- 22602 26 ee ee ee ,, Pottery desion: dlemente_-- === 2522-22585" oe ee ee ee ) Projectile pointe; .= ==>. =-- 2222-2 2-2=55 202 ont oe End scrapers and knives! <=22==- = 2=-2252229= 22 eee eee ee Chipped and ground stone artifacts ---.-----=----------------.- . Baked clay effigies and scapula hoes__-_...----------------------- » Worked bones. 2+ S25 -2ae-< 55s 22222 peee sees es eee eee ee MWOrkeG Ones ac 2 22> cee. te sea eee eee Soe een ee erent ReROnewawisses= see oe eee eee eae See ee eae ae eee eee eee Mo WorkedbOn@s* 22552 s===-ase. ses oea esas ieee es = aes ee ee eee . Objects of antler-and animal teeth_=—-.. -- =====2-==22=25+-_==-2=4- JASHei benOs SHOCGISKS2on == Snes =stecet eco eases hoes eee oem aee ; Paul Brave site and environs. .2.-.--.--=._--.. -S2essc Sseeeer—e- | Site map, showing @xcavations....-....--_-- 2 = S5 see eee » Houses, -excavation 1c. +--2. 286.268 26. Soe Se eee Stee p OUse 2 VeXCAValION 2. 202504 eo eo nea eee ee ee eee pHipuse a. -6xCaVatlouiss =~. 22.52.2550 -ooS-cene eee een PREFACE In 1947 an archeological field party, sponsored by the University of North Dakota and the State Historical Society of North Dakota, carried out excavations in the upper limits of the Oahe Reservoir, in North Dakota. Test excavations were made at the Paul Brave site (32S14), also known as the Fort Yates site. The elevation of this prehistoric village is between 1,600 and 1,610 feet. The site will be flooded by the Oahe Reservoir when backwater reaches the maximum pool level of 1,620 feet. The work in 1947 was directed by Dr. Gor- don W. Hewes, then with the University of North Dakota. The limited excavations in 1947 indicated the desirability of further and more intensive work at the site, and in 1955 the State Historical Society of North Dakota sponsored a second party for full-scale excavation. Funds were made available for this archeological salvage work through a cooperative agreement with the National Park Service. Between July 6, 1955, and August 30, 1955, excavation was carried out under our supervision when we were both staff archeologists with the State Historical Society. The assistance of Oriol Pi-Sunyer, then a graduate student at Harvard, is gratefully acknowledged. The genial and competent crew members contributed further to the sum- mer’s accomplishments. These were Stephen W. Robinson and Robert P. Barr, of Grand Forks, N. Dak., and Russell B. Lawrence and Robert F. Gipp. of Fort Yates, N. Dak. The overburden at the site necessitated the use of a bulldozer, which was rented from the Standing Rock Tribal Council, and operated by Jack McLaughlin of Shields, N. Dak. The use of this machine implemented the removal of the nearly 3 feet of sterile overburden from the house floors. The River Basin Surveys, Smithsonian Insti- tution, provided the field party with cameras and other equipment necessary to document features found in the excavations. We took aerial photographs on a flight early in September 1955. We wish to thank Robert L. Stephenson, Charles H. McNutt, and Warren Caldwell for constructive criticism of this report. Joseph P. EK. Morrison identified the shell material from the site. Russell Reid, superintendent of the State Historical Society of North Dakota, assisted in the identification of the baked clay animal figurines from the site and aided the investigators in many other ways. Photographic plates were prepared with the assistance of Bernard Weinreich, Bis- xiit XIV PREFACE marck photographer. The maps and line drawings were prepared by Wood. Permission to excavate on the Standing Rock Indian Reservation was generously granted by the Standing Rock Tribal Council. Mem- bers of the Paul Brave estate gave their permission to excavate site 32S14. Superintendent J. W. Wellington extended every effort to make the summer comfortable and successful, and Tribal Chairman David Black Hoop aided the excavators in many ways. This assist- ance made the summer of 1955 a profitable and enjoyable one, and it is gratefully acknowledged. The field notes, maps, and artifacts from the site were returned for study and preservation to the State Historical Society Museum in Bis- marck, and they are on file in that institution. Some of the field ree- ords and collections made in 1947 are also in the collections at that museum, and the rest are at the University of North Dakota, in Grand Forks. AIl of the 1947 material at the Museum was reanalyzed and is incorporated in the present study. W. Raymonp Woop Avan R. Wootworru THE PAUL BRAVE SITE (32814), OAHE RESERVOIR AREA, NORTH DAKOTA’? By W. Raymonp Woop and AnAn R. WooiworTH INTRODUCTION The purpose of this study is to present a detailed descriptive state- ment of the archeology of one of the early village sites in the northern part of the Middle Missouri area. The Middle Missouri area consists of the Missouri River Valley and the lower reaches of its tributaries between Bismarck, N. Dak., and southeastern South Dakota (see Lehmer, 1954, p. 140). The Paul Brave site is significant because it was occupied at the time when the first village or town dwellers were establishing themselves in the northern part of this area. The deri- vation of these early village people and the subsequent settlement patterns of the Northern Plains are topics of interest to the anthro- pologists and historians of the region. Since 1906, when George F. Will and Herbert J. Spinden reported work at the Double Ditch (Bourgois) Mandan site, north of the city of Bismarck, N. Dak., archeological interest in the northern Middle Missouri area has increased. Some of the stages in the history of the area have been blocked in roughly, particularly by George F. Will and Thad. C. Hecker (1944). Details, however, are scant, and the present study provides a base from which a more adequate definition of the early village people may be made. The Jack of any real fund of comparative data precludes any sweeping conclusions, although a few tentative generalizations are justified on the basis of excavations at the Paul Brave site. Preliminary statements of the archeology of Paul Brave have ap- peared in several publications. The first reference to the site is in Will and Hecker (ibid., p. 89), where it is described as an unnamed “Archaic Mandan” site on tribal land north of the Paul Brave estate. Test excavations in 1947 resulted in two brief articles by the exca- vator, Gordon W. Hewes. The first of these (1949 a) describes the excavations in summary form, discusses some of the significant finds and tentatively places the site in time. A second article (1949 b) is 1 Submitted September 1959 ; some revision April 1961. 2 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 a preliminary classification of the pottery from the excavations, with notes on extra-site distributions. In addition, there are references to Paul Brave in the report on the Thomas Riggs site, 39HU1 (Hurt, 1953). In Hurt’s report Paul Brave is referred to as the “Fort Yates site.” In the present study, site 32SI4 is designated as the Paul Brave site to avoid duplication in terms, since the term “Fort Yates” has been used in other contexts. ARCHEOLOGY OF THE SITE SITH DESCRIPTION The Paul Brave site is located in the N14N1,NEY, of sec. 8, T. 129 N., R. 79 W., Sioux County, N. Dak. It is on the west bank of the Missouri River 6 miles downstream from the town of Fort Yates, and about 6 miles upstream from the North and South Dakota boundary. The site is on a rolling terrace that overlooks the Missouri River flood plain, this terrace averaging 20 feet above the flood plain. To the south, the land rises in a series of low hills that abut against the Missouri River bluffs, about one-half of a mile distant (map 1; also pl.1,a). Fire Heart Butte, a high, flat-topped butte, is about 4 miles to the southeast. At the time the site was investigated in 1955, it was found that at least 14 oval house depressions were distributed over an area of about 4 acres. These depressions were as much as 65 feet long and were 2 to 3 feet deep. They were filled with a lush vegetation that contrasted with the short, dry grass on higher points in the village area. The site is said to have been cultivated (Will and Hecker, 1944, p. 89), but activity does not seem to have obscured the surface indications of the larger subsurface features. The village may have been larger originally, because Will and Hecker (1944, p. 89) also noted house floors exposed in the terrace edge facing the river. This terrace was checked in 1955. A number of artifacts were recovered from the sur- face, but there was no indication of house floors. There is no evidence that the site was fortified, but it is partially isolated from the re- mainder of the terrace by a narrow swale south of the village (map 2; also pl. 1, a). EXCAVATION METHODS Preliminary testing determined that the houses were built in shal- low pits. Three houses were chosen for excavation after the depth of the house floors and the house walls had been determined. These houses were evenly distributed over the site, and none of them had been tested previously. The entire site was mantled by a deposit of light-buff aeolian soil. This mantle was as much as 4 feet deep over some of the houses, but high points within the village were buried Riv. Bas. Sur. Pap. No. 33, PAUL BRAVE SITE—WOOD AND WOOLWORTH 3 Winond caTTAIL —= ae Will ES — Uplands Map 1.—Paul Brave site and environs. only about a foot. A bulldozer was used to strip this overburden from the three houses down to house fill, which was excavated by conventional hand methods. House fill was a dark, mottled, and re- latively soft mixed earth that contrasted with the lightly colored sandy native soil. The depth of the bulldozer cut was regulated to approximate the depth of the former village occupational level. Three other bulldozer cuts were made in the eastern part of the site to expose cache pits and other features, and to check for deeper occu- pational levels. No evidence was observed that indicated any occupa- tion prior to that represented by the houses. House floors were care- fully exposed, and all postholes, fireplaces, pits, and other features were cored. The structures were mapped radially by use of a plane table, open-sight alidade, and steel chain after excavation. Photo- graphs were taken of the excavated houses, features, significant finds, and progress of the work. Several aerial photographs were taken to 4 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 show the work accomplished and the general environmental setting of the village after the regular field season had ended. The work done at Paul Brave in 1955 included the testing of seven areas. These tests, or excavation units, are herein termed X units, and are designated as X1 through X7. The excavations of Gordon — W. Hewes in 1947 are hatched on the site map (map 2). Eighty- eight features, designated F10 to F97, were recorded (table 1). FEATURES EXCAVATION UNIT 1 (HOUSE 1) This long-rectangular house was in the western part of the site (map 3). It had a maximum length of 45.5 feet, and was 32.5 feet wide at the end near the entrance and 31 feet wide at the end opposite the entrance. The long axis was oriented northeast and southwest, with the entrance in the southwest end of the house facing away from the river. The floor was 3.6 feet below the present ground level. The house was originally built in a pit 2 feet deep. The house walls and floor were of unfaced native soil. The entrance was marked by three postholes on either side of a small bench that projected into the house floor. This bench was composed of undisturbed native soil. The entrance postholes were 1.0 to 2.3 feet deep. Three other postholes southwest of the house but in line with the entrance suggest that this passage was originally 14 feet long. Three large postholes were in the midline of the floor. One of these, in the end opposite the entrance, was 3.1 feet deep. It was lined with stones, and the hole inclined toward the entrance at a 13-degree angle. Another post near the rear wall was 1.6 feet deep and also inclined toward the entrance. Midway between the end post and the entrance was a post 3.0 feet deep. Another post, not located along the midline of the floor, leaned toward the end post. It may have served as a brace. There were 20 postholes along each of the long walls of the house, from 0.9 to 2.1 feet deep, averaging 1.5 feet deep. There were no posts in the house ends. On either side of the midline of the floor, there was an intermediate row of posts. In each row were five posts, spaced 8 to 10 feet apart (posts I to X, map 8). The fireplace, F19, was centered on the midline of the house, off- set toward the entrance. It was circular and basin shaped, and contained ash, mixed earth, and a burned floor. Another basin- shaped pit, F81, may also have been a fireplace, owing to the fact that its floor was burned, but there was little depth to the burning. Perhaps it was used sparingly or for only a brief time as a fireplace. PAUL BRAVE SITE—WOOD AND WOOLWORTH Riv. Bas. Sur. Pap. No. 33] ~*— 13N3y OL 06s) Ogg, 009) ols) O29; *suoIqeAvoXS SuImoys “deur 91S —"7 dv] Sido suojssaidap asnoH %,, SUOIJDADDXA JHE YA SUOIJDADDX3 GG6I Fz Mu, ON st™ SOC Wn a JOAIY oo0e Q}IS SO}DA 4404 JO aADIg {NDY VISZE ofl N S34¥A Iyoy a [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY al UOI]BAROXSS I osnoyj—*¢ dvi swoeq j001 40 UOI}ISOd B]QDqGoiq ----- auo}s MO3ID $0 UO!}DaJIP ur Buiuoe; ‘ejoyjsog BJOYyjJsod peonig 9}0y}Sog 12 S- Bi @ 4/4 pedoys-uisog ©) 4d pedoys -j10g iC) er) eopjdesi4 v— | asnoy iS Ge Ix e Riv. Bas. Sur. hw Pap. No. 33]. PAUL BRAVE SITE—WOOD AND WOOLWORTH a A number of large stones were along the edge of this pit. At the rear of the house was an area of burned earth (F94), which appeared to be a surface hearth. A small, basin-shaped fireplace (F71) con- tained white ash, and the underlying soil was burned. Fourteen bell-shaped pits of various sizes were in the house floor. They tended to be in a line with the intermediate rows of posts. These pits were small, averaging less than 2 feet in depth. Six basin- shaped pits contained mixed earth and refuse. One of these (F82), may have been dug after the house was abandoned, since the wall postholes observed in the pit floor were not visible within the pit fill. A layer of charcoal blanketed the floor of the house, and many of the postholes contained the charred remains of posts. Apparently the house was destroyed by fire. EXCAVATION UNIT 2 (HOUSE 2) This long-rectangular house was in the north-central part of the site (map 4; pl. 1, 5). It had a maximum length of 37.5 feet, was 24.0 feet wide near the entrance, and 26.0 feet wide at the rear. The long axis of the house was nearly north and south, with the entrance in the south end facing away from the river. The house floor lay 2.8 feet below the surface, and the house was built in a pit 1.4 feet deep. The house floor and walls were of unfaced native soil. The entrance was marked by four postholes, two on each side of a small bench of native soil that projected into the house floor. These postholes were 1.1 to 2.4 feet deep. An entrance was indicated by 11 postholes that outlined a passage 18 feet long and 4 to 5 feet wide. The postholes began at a point about 10 feet from the house wall. Despite a careful search, no postholes could be located between the house wall and that point. A hole 1 foot in diameter and 0.8 of a foot deep was in the midline of the entrance, but it probably was not a part of that feature. Two large postholes were in the midline of the house. The post- hole in the rear wall was 0.9 foot in diameter and 2.1 feet deep and was filled with white ash. Midway between the end posthole and the entrance was a posthole 0.9 foot in diameter and 3.2 feet deep containing the remains of a cedar post that was braced with stones. The long walls of the house were lined with 12 postholes each and, although auxiliary posts occurred, the two rows were mirror images of each other in spacing and placement. These postholes were 0.8 to 2.4 feet deep, averaging 1.5 feet. All house posts observed were cedar, and all were vertical. There were no postholes along the end walls. [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 1395 swoaq 4004 40 UOIISOd B[QDGOIy — «vesseane euojS fo) ejoyujysod paon1g @ B/OYSOg @ ud papis ~jybio.ys 40 padoys -uisog 4, \ ud padoys-ljag ee), \ Al eopjdast4 @) 2 asnoy VISce ————. {z Foe *Z uoIwAvoKe ‘7 OsNOF[—"F AV] Riy. Bas. Sur. baw ap. No. 331 PAUL BRAVE SITE—WOOD AND WOOLWORTH 9 The fireplace (F26) was offset toward the entrance, and was cen- tered on the midline of the house. It was filled with compact white ash and mixed earth and was lined with burned earth. East of the entrance, in a shallow recess in the house wall, a large unmodified stone slab was on the house floor. The material is a conglomerate, and similar stone occurs along the river bluffs south of the site. Five bell-shaped pits and several basin-shaped pits filled with mixed earth and refuse were in the house floor. An irregularly shaped pit (F67) was near the rear wall of the house. A thin layer of charcoal and burned earth marked the floor level, and the tops of most of the posts were charred, a condition indicating that the house was leveled by fire. EXCAVATION UNIT 3 (HOUSE 3) This long-rectangular house was in the eastern part of the site (map 5). It was 46.0 feet long, 32.5 feet wide at the end near the entrance, and 29.0 feet wide at the back end. The long axis of the house was oriented north and south, and the entrance was in the south end, facing away from the river. The house floor was 2.8 feet below the present surface. The structure was originally built in a pit more than 1 foot deep. The floor and walls were of unfaced native soil. The entrance was marked by four postholes, two on each side of a small bench of native soil that projected into the house. These post- holes were 0.9 to 1.8 feet deep. Two postholes south of the house but in line with the entrance suggest that the passage was 15 feet long and 4 to 5 feet wide. There were two large postholes along the midline of the house. A posthole 3.0 feet deep and 1.6 feet wide in the rear wall contained a vertical cedar post 0.9 foot in diameter that was braced with stones. In the house center, midway between the entrance and the end posthole, was a posthole 0.9 foot wide and 1.9 feet deep. Secondary supports were provided by two rows of postholes located between the house midline and the walls. They consisted of two rows of five posts each, spaced from 8 to 10 feet apart (posts I to X, map 5). Several of them were lined with stones. ‘There were no postholes along the end walls. The fireplace was centered on the midline of the house, offset toward the entrance. It contained compact white ash, mixed earth, and was lined with burned earth. Three miniature pottery vessels were found embedded in the ash. Two of these vessels are illustrated (pl. 6, f-g). An oval, basin-shaped pit (F53) and a small rectangular pit (F63) had lightly fired floors and may also have served as fireplaces. Three smaller fireplaces (F64, F65, and F66) contained ash, and were lined with burned earth. [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 10 jaa} i e ol s ) swpeq joor jo WOlpisod ajGoqOsd —aranane QuolS oO sjouysod paonig we BJoyysod ®@ Jaquil} sopag =e ud pays suis B® ae / ‘ jid padoys-)jog 1 ' Wee gli gopjdesl4 ¢ asnoy bIS ce i. Ad *¢"uorjeAvoxe ‘¢ asnop]—"¢ dV] rt oO o> EEE UES —vV Riv. Bas. Sur. Pap. No 33) PAUL BRAVE SITE—WOOD AND WOOLWORTH 11 A unique feature near the rear wall of the house was a long straight trench with rounded ends and a shallow U-shaped cross section. Its long axis was perpendicular to the long axis of the structure. The feature (F'70) contained a few bone fragments and two stones (pl. 2,@). There were a few bison skull fragments near the rear wall in the house, but there was no other evidence of a household shrine. Hight bell-shaped floor pits were along the house walls. Basin- shaped pits, present in the other dwellings, were lacking in this structure. Five charred timbers on the house floor, at right angles to the house walls, were probably fallen wall posts (pl. 2, 6). They were cedar timbers ranging in length up to 4.8 feet. The inner ends of the timbers were about 5 feet from the line of wall posts, suggesting that the wall was at least 5 feet high. The timbers averaged about 0.3 foot in diameter, although they may originally have been larger. The evidence indicates that this house, as well as the other two ex- cavated houses, was destroyed by fire. EXCAVATION UNIT 4 Unit 4 was in the northeastern part of the site, east of House 2. It was a trench oriented north and south, 130 feet long and 14 feet wide, attaining a maximum depth of 5 feet. Near the south end of the unit was a house depression (F12) the floor of which was 4.2 feet below the present surface. Three pits were exposed near the south end of the house. F382 was profiled; this bell-shaped pit could not be associated with the house. F381 wasa small bell-shaped pit in the floor of the house, and F30 was a pit of indeterminate form on the south edge of the house. EXCAVATION UNIT 5 Unit 5 was east of House 2 and northwest of House 3. It was a trench oriented north and south, 85 feet long and 14 feet wide, with a maximum depth of 5 feet. A bell-shaped pit (F15) was in the south end of the unit. EXCAVATION UNIT 6 Unit 6 was east of House 2 and northwest of House 3. It was a trench oriented north and south, 75 feet long and 14 feet wide, attain- ing a maximum depth of 5 feet. A bell-shaped pit (F46) was in the north end of the unit. EXCAVATION UNIT 7 Unit 7 was a test pit approximately in the center of the village area, 5.6 feet north and south and 17 feet east and west. It is an extension of a test pit excavated by Hewes in 1947. The 1947 excavation was a 5-foot square. In 1955 this unit was extended to the west and revealed 1,24 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 a house wall. The wall was of unfaced native soil, sloping up from the floor at an angle of 20 degrees to the former village level. The house pit was 1.8 feet deep, and at the wall the overburden was 2 feet deep. Four postholes along the wall contained the remains of cedar posts or soft mixed earth. The holes varied in diameter from 0.5 to 0.8 foot, and were 1.3 to 2.4 feet deep. No other features of interest were noted on the trench floor. EXCAVATION UNIT 8 (HOUSE 4) Unit 8 revealed the floor of a long-rectangular house. The structure was excavated in 1947, and the following description was paraphrased from Hewes’ original report (1949 a, pp. 22-23). The largest and deepest house depression at the site was cross- trenched, and the greater part of the floor was ultimately cleared. The floor lay at a depth of about 4 feet below the present surface. The upper foot of the overburden was sterile, including only recent de- posits of humus, but the remaining fill included sherds, bone and flint tools, and animal bone. The structure was approximately rectangular, 65 feet long and 35 feet wide, with four rows of postholes. (The floor plan of the house is apparently similar to that of House 1 and House 3.) Three fireplaces, devoid of refuse, were on the floor. The only other significant features were three large piles of bison bone, chiefly cranial parts and horn cores, intermingled with rough stone. Near the floor center, about 50 inches below the surface of the house depression, were two large broken pottery vessels resting mouth down on the floor. These two vessels were illustrated in Hewes’ original report (1949 b, pl. 5, lower right, ¢-w). The house fill above the floor contained such refuse as bone, sherds, and tools, but the floor was free of such detritus. Hewes was inclined to interpret this feature as a ceremonial structure, rather than a household dwelling unit. In size, at least, the house stands apart from the smaller structures excavated in 1955, and Hewes’ interpretation is convincing. HUMAN BURIALS There is a low rise of ground south of the site, and this rise is inter- cepted along its north edge by aroad (map 2). In 1956, George Haiser reported to the excavators that, prior to World War IT, several human burials were gouged from the south bank of the road by road-cutting operations. It is possible that this rise of ground may have served as a cemetery for the village, since there were no human remains in the site. The burial positions and the disposition of the remains from the road cut are not known. Other evidence of activity in the low rises south of the site was encountered. Survey parties in 1955 found circular shell disk beads Riy. Bas. Sur. By, Nor gs; PAUL BRAVE SITE—WO0OD AND WOOLWORTH 13 on a hill about 1,000 yards south of the site. These objects might be grave offerings brought to the surface by animal activity, and they are similar to beads from the excavated houses and features. Unfor- tunately the brief field season did not permit further exploration of the area south of the site. Table 1.—Description of features haa X No. | Depth Length and width Identification 0. Feet Feet 10 7 PERSON OAGH te aes ne nena eee eee Rectangular house. ll 2 1.4 Lene, 37.5 N-S width, 24.0-26.0 | Rectangular house (house 2). 12 4 ee Floor 4.2 below surface-....----------- Rectangular house. 13 1 2.0 | Length, 45.5 N-S width, 31.0-32.5 | Rectangular house (house 1). 14 3 1.0 | Length, 46.0 N-S width, 39.0-32.5 | Rectangular house (house 3). 15 5 3.8 | Orifice diam., 4.0; diam., 6.0._--------- Bell-shaped pit, floor slightly fired. 16 1 1.4 | Orifice diam., 1.8; diam., 2.2-.__.__.._- Bell-shaped pit in House 1. 17 1 .6 | Diameter 1.3 E-W 1.1 N-S_--___----- Basin-shaped pit in House 1. 18 1 1.0 | Orifice diam., 0.9; diam., 1.5-_..-_--._- Bell-shaped pit in House 1. 19 1 BGa Pa tTHOLOn 0 ean eee sane ne Primary fireplace in House 1. 20 1 1.0 | Diameter, 5.2 N-S, 6.8 E-W________--- Basin-shaped pit north of House 1. 21 2 2.6 | Orifice diam., 2.4; diam., 3.2-___..-._- Bell-shaped pit northeast of House 2. 22 2 Ron ED AITIObar. Ae Leeee = Sane eee eee Basin-shaped pit northeast of House 2, 23 2 Tsk |! PBN 4 a1 2) yee ke Ege pe pe ee Basin-shaped pit south of House 2. 24 2 1.7 | Diameter, 9.2 N-S, 7.5 E-W---__-_---- TIrregularly shaped pit south of House 2. 25 2 “(ol | AD AEN Ta has) oy LC) Se A a ee Ae Basin-shaped pit north of House 2. 26 2 atch [le AE Genel SO) eee ky ee Ee Primary fireplace in House 2. 27 2 1.1 | Orifice diam., 1.8; diam., 2.2-_-._.____- Bell-shaped pit in House 2. 28 2 hdl DUET ata) Fee Deg tae el a ed Basin-shaped pit in House 2. 29 2 1.2 | Orifice diam., 1.4; diam., 1.6_._.-_____- Bell-shaped pit in House 2, 30 Be tee oa ideterminateses=s sssere sae ee sk Pit of indeterminate form. 31 1 2.0 | Orifice diam., 2.6; diam., 3.2-_...____.- Bell-shaped pit in F12, 32 4 3.9 | Orifice diam., 2.0; diam., 5.6---.------- Bell-shaped pit. 33 2 2.0 | Orifice diam., 1.8; diam., 3.3-._-__-___- Bell-shaped pit in House 2. 34 2 2.7 | Diameter, 3.1 N-S, 2.5 Ly... ira ae a pit with round bottom in House 2. 35 2 Sai Dinmotenndidecessetee ee emecee ose aea st ye pit with flat bottom in ouse 2. 36 2 1.8 | Diameter, 2.0 N-S, 1.8 E-W--_-.------- Saeed pit with flat bottom in ouse 2. 37 2 2.6 | Orifice diam., 2.5; diam., 3.5-...-.---.. Bell-shaped pit in House 2. 38 2 1.6 | Orifice diam., 2.5; diam., 2.8-_......--- Bell-shaped pit in House 2. 39 2 2.0 | Orifice diam., ca. 1.2; diam., 2.2__.___- Bell-shaped pit east of House 2. 40 2 1.0 | Orifice diam., 2.1; diam., 2.6-.__--_-._. Bell-shaped pit in House 2. 41 2 1.4 | Length 1.7; width 1.2; thickness 0.2-_- ices or cami stone on floor of ouse 2. 42 2 2.7 | Orifice diam. 3.0; diam., 5.0-.---...--- Bell-shaped pit southwest of House 2. 43 2 car | PO iameber al Os as sa eee kee ELS Basin-shaped pit in House 2. 44 2 ae Diameter iG 955 ings ae be Basin-shaped pit in House 2. 45 2 a5D) |! Dismdterss0- Ais. este k Ak Basin-shaped pit southeast of House 2. 46 6 2.6 | Orifice diam., 4.5; diam. 6.5-----..-_--- Bell-shaped pit. 47 Sl eee Length, 4,8; thickness, 0.3_..---------- Charred cedar post in House 3. 48 By | aecsced Length, 3.4; thickness, 0.8_------------ Charred cedar post in House 3. 49 os ok Length, 4.1; thickness, 0.3__-...------- Charred cedar post in House 3. 50 Sys cae Length, 3.4; thickness, 0.3.------------ Charred cedar post in House 3, 51 3D |sdesdsad Length, 1.5; thickness, 0.2__.-.--_----- Charred cedar post in House 3. 52 3 AG alls Diameter. 30.2 ses seen Le Primary fireplace in House 3. 53 3 -6 | Diameter, 4.0 N-S, 3.4 E-W----------- Basin-shaped pit containing stones; bottom slightly fired. 54 3 1.7 | Orifice diam., 2.0; diam., 2.6....___-_-- Bell-shaped pit in House 3. 55 3 1.5 | Orifice diam., 2.0; diam., 2.6_..-._.---- Bell-shaped pit in House 3. 56 3 13 0h]Diamoter, LSE er eee el esr Straight-sided pit with rounded bottom in House 3, filled with fragments of sandstone. 57 3 2.1} Orifice diam., 2.2; diam., 2.8.....-.-.-- Bell-shaped pit in House 3. 58 3 2.0 | Orifice diam., 2.7; diam., 3.0_-_--_----- Bell-shaped pit in House 3. 59 3 2.6 | Orifice diam., 3.0; diam., 3.4___--_----- Bell-shaped pit in House 3. 60 3 2.5 | Orifice diam., 2.5; diam., 3.0_.__.____-- Bell-shaped pit in House 3. 61 3 2.0 | Orifice diam., 1.6; diam., 2.0__.__-___-- Bell-shaped pit in House 3. 62 3 1.8 | Orifice diam., 1.8; diam., 2.4_._.._.___- Bell-shaped pit in House 3. 63 3 .5 | Diameter, 1.9 N-S; 1.7 E-W_____-_---- Basin-shaped pit, rectangular in out- line, with slight firing on bottom. 64 3 -3 | Diameter, 1.0 N-S; 0.6 E-W__-_-_------ Basin-shaped fireplace containing ash, charcoal, burned earth. 65 3 4) || MOA MIGLEIa FO meena ree nasa oe ene Basin-shaped fireplace containing ash, charcoal, burned earth. 14 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLE 1.—Description of featwres—Continued Feature} X No. | Depth Length and width Identification No. Feet Feet 66 3 032) Diameters 0:9. 5 eee ee Basin-shaped fireplace containing ash, charcoal, burned earth. 67 2 1.1 | Length, 7.2 E-W; width, 3.0 N-S__.-. Trregularly shaped pit in House 2. 68 2 1.8 | Diameter, 5.6 N-S; 6.5 E-W___------- Oval, basin-shaped pit in House 2. 69 2 .4 | Diameter, 2.0 N-S; 1.6 E-W_-_...------ Basin-shaped pit in House 2. 70 3 1.0 | Length, 15.4 E-W; width, 2.0...-_---- Long straight trench in House 3. 71 1 SSA POaMoeter, Noe eee aos coe cee eae Basin-shaped fireplace, with ash, burned earth; House 1. 72 1 -8 | Orifice diam., 1.5; diam., 1.7 -.-------- Bell-shaped pit in House 1. 73 1 2.3 | Orifice diam., 2.8; diam., 3.2._.-------- Bell-shaped pit in House 1. 74 1 1.3 | Orifice diam., 1.2; diam., 1.8_..-.------ Bell-shaped pit in House 1. 75 i} 2.1 | Diameter, 2.5 N-S; 3.0 E-W-_-_-_------ Basin-shaped pit with posthole in bottom; House 1. 76 1 1.4 | Orifice diam., 1.8; diam., 2:1-__--_-.---- Bell-shaped pit in House 1. 77 uf 1.2 | Orifice diam., 1.8; diam., 2.0___- -| Bell-shaped pit in House 1. 78 1 1.7 | Orifice diam., 1.8; diam., 2.0. -| Bell-shaped pit in House 1. 79 1 1.8 | Orifice diam., 2.0; diam., 2.4_ .| Bell-shaped pit in House 1. 80 1 1.3 | Orifice diam., 2.4; diam., 2.6_ _| Bell-shaped pit in House 1. 81 1 1.3 | Diameter, 4.1 N-S; 4.9 E-W___._------ Basin-shaped pit containing stones on pit wall; floor is slightly fired; House 82 1 wal MD lameter, 0.0) -2-ssccacetaaee caneaee os Basin-shaped pit with postholes in bottom; House 1. 83 1 10m paMmPLlenn ey Ula oscecwanesnanensscoe Basin-shaped pit in House 1. 84 1 156s), DIAMeLer, AeSs-- 2c) os cee eee Straight-sided pit with round bottom in House 1. 85 1 2.2 | Orifice diam., 2.6; diam., 3.0...-....--. Bell-shaped pit in House 1. 86 1 1.4 | Orifice diam., 2.0; diam., 2.4...-__--___ Bell-shaped pit in House 1. 87 2 150) | #Diamoater.c.0— scons es ecco Basin-shaped pit southeast of House 2. 88 2 2.2 | Orifice diam., 4.3; diam., 6.2...-.---.-- Bell-shaped pit southwest of House 2. 89 2 Sug IGIMO LONG are tone. co won ace acne cee Basin-shaped pit southwest of House 2. 90 2 2.0 | Orifice diam., 3.2; diam., 4.0-_...------ Bell-shaped pit southwest of House 2. 91 1 2.3 | Orifice diam., 4.2; diam., 6.0.-....-.-- Bell-shaped pit north of House 1. 92 2 20lle Diameter, 4:be- aos c oon eeacosncetecans Basin-shaped pit south of House 2. 93 1 1.0 | Orifice diam., 1.0; diam., 1.2._...._.__- Bell-shaped pit in House 1. 94 1 .2 | Diameter, 2.3 N-S; 3.3 E-W--_._-_-_-- Shallow depression, floor of burned earth; House 1. 95 1 SSHiaiNAamoLer- soso os 2! 2 seen oa Bee Basin-shaped pit in House 1. 96 1 SA Ibiame ters 4e oo ° 25s ee Basin-shaped pit in House 1. 97 1 1.1 | Orifice diam., 1.8; diam., 2.2. Bell-shaped pit in House 2. ARTIFACTS POTTERY TYPES Several pottery types from the Paul Brave site are described below. These types differ in certain respects from types previously described by Gordon W. Hewes for some of the same pottery (1949 b, pp. 61-67). This study utilizes the pottery types defined at the Thomas Riggs site (Kleinsasser, 1953), but adds certain types that were lack- ing at Thomas Riggs. This reclassification was deemed necessary for several reasons. Pottery wares and types have been defined for a number of complexes in the Middle Missouri area since Hewes ini- tially defined his types from the Paul Brave site. These classifica- tions seem to have adopted rim form as the primary sorting criterion. Types or subtypes have been established within each ware for vessels or sherds bearing different decorative techniques. It is difficult to compare Hewes’ types with those established by other workers, since Hewes did not use rim form as the primary criterion in his classifica- Riv. Bas. Sr PAUL BRAVE SITE—WOOD AND WOOLWORTH 15 tion of Paul Brave pottery. The classification below uses the con- cepts of types and wares defined by Lehmer (1954, p. 41), which are in current use in the Middle Missouri area. Wares are defined as groups of types which share a majority of basic characteristics in- cluding paste, vessel shape, surface finish, and rim form. Types in- cluded within a ware are groups of vessels or rim sherds decorated or modified in a consistent manner. The paste, surface finish, and form of the pottery from Paul Brave are described below to apply to all ceramics from the site. PASTE: Method of manufacture: Probably lump modeled with paddle and anvil. Temper: The material is uniformly decomposed or calcined granite. The amount and size of tempering varies with the size of the vessel. In the smaller vessels and in miniatures, temper is small and sparse, while in larger vessels it is coarser and more abundant. Particles of quartz, mica, and feldspar are visible in sherd cross sections; particles are 0.5 to 3.0 mm. in diameter. Texture: The surfaces are medium fine to coarse, with the quality of the paste decreasing as vessel size increases. The majority of sherds are rough on the exterior because of irregular smoothing, and many are crackled on the interior. The core is compact to contorted, with evenly distributed temper. Hardness: 3.0 to 3.5, the majority 3.0 (calcite). Color: Buff, through light and dark grays to black, with the majority a dark gray. Source clay: A plastic clay, light gray in color, was used to make the pottery. It is free of sand but contains silt. SURFACE FINISH: Vessel interiors and the upper parts of the vessel exteriors are usually horizontally smoothed. Shoulders were malleated vertically with a grooved paddle, but more frequently the resulting grooves were smoothed over. Rim sherds may be vertically stamped, but ordinarily are smoothed. Only three rims are horizontally stamped. The bases of the vessels are impressed in a random fashion. Some vessels are lightly polished, and their surfaces reflect some light (pl. 6, e, h). ForM : Lip: Rounded, pointed, or flat, depending upon the presence and type of deco- ration. Rim: Two rim forms occur. The Riggs Ware includes those types with ver- tical, straight to outflaring rims. The Fort Yates Ware includes those types with S-shaped rims. Neck: Ordinarily constricted, except on vessels with straight, vertical rims. Shoulder: Rounded and steeply sloping. Body: All vessels appear to be globular, with round bases and wide mouths. In the restorable vessels, vessel height is about equal to maximum di- ameter. Appendages: Loop handles predominate, with strap handles present but rare. Handles are welded to the lip and riveted to the shoulder. Ver- tical, triangular tabs are common on the lip. Usually these elements have a central incision. 16 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 WARES AND COMPONENT TYPES : Riggs Ware: Component types: Riggs Plain Rim Riggs Cross-Hatched Rim Riggs Incised Rim Riggs Pinched Rim Fort Yates Ware: Component types: Fort Yates Cord Impressed Rim Fort Yates Cross-Hatched Rim Unclassified: Example A Example B Example C Example D RIGGS WARE Riggs Plain Rim (Pl. 3, 6, d-h; pl. 6, e) SAMPLE: 619 rim sherds, representing about 381 vessels, and one restored ves- sel. DECORATION : Lip: On 237 sherds there are oblique or vertical tool impressions on the lip or on the outer edge of the lip. Circular and oval punctates and tool- impressed crosshatched lines, as well as fingernail impressions, occur on some lips (pl. 3, e-g). Three rims bear oblique cord-impressed lines, and seven rims have closely spaced castellations (pl. 8, d@). The remain- ing rims have plain lips (pl. 3, h). Rim: Plain, with some horizontal smoothing over grooved paddle impres- sions. A few rims have horizontal paddle impressions unmodified by smoothing. A single decorated rim is included in this type (fig. 1, c). Shoulder: Sherds of 13 vessels are incised with various designs, most of which are rectilinear. Four design motifs are present: (1) Incised chevrons of three or four lines alternate with a triangular “bear foot” element (fig. 1, a-b). The restored vessel bears this design, executed by a stab-and-drag technique (pl. 6, e). (2) Incised “drooping corn stalks” or oblique incisions are superim- posed over horizontally incised lines (fig. 1, f-g). (3) Panels of opposed diagonals occur (fig. 2, b-c). (4) Sherds from two vessels have incised, concentric circles on the upper shoulder (fig. 2, @). Form: Lip: Round, pointed, or flat, depending upon the presence and type of deco- ration. Thickness, 4 to 10 mm., averaging 6 mm. Rim: Straight and vertical to somewhat outflared. Heights are 10 to 70 mm., averaging about 35 mm.; thickness, 5 to 12 mm., averaging 7 mm. Size: The projected orifices of 12 rim sherds indicate diameters between 120 and 346 mm.: 120 240 308 140 243 320 156 264 334 228 286 346 Riv. Bas. Sur. PAUL BRAVE SITE—WOOD AND WOOLWORTH 17 \ VM Ficure 1.—Pottery design elements. Appendages: Sherds from seven vessels have loop handles extending from the lip to the upper shoulder. Oue of them has a vertical, triangular tab with an apical incision. There are two loop handles and two incised tabs on the restored vessel (pl. 6, e). Twenty-three vertical tabs, roughly triangular, are of various sizes. Three of them are apically incised. Three horizontally projecting lugs occur on vessel lips. Reconstructed vessel: A globular vessel with two loop handles and two in- cised tabs has a shoulder design made with the stab-and-drag technique 18 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Ficure 2.—Pottery design elements. (pl. 6, e). When filled to the neck it contains about three-quarters of a pint of water. It has the following dimensions: Heisht. (excluding appendages) <==... 5 eee eee Maximnymvoiameter (at. shoulder)... ee eee Orifice diameter, (inner lip)... => - es Thin pthickness 2s ee a ee ee Rim height__-_-- Thickness of base Mm. Riv. Bas. Sur. pee se Pap. No. 33], PAUL BRAVE SITE—WOOD AND WOOLWORTH 19 Miscellaneous: One straight rim sherd with a plain, flat lip has a cylindrical hole 15 mm. from the lip (pl. 3, b). The hole was punched from the ex- terior while the clay was still moist, and the raised edges on the interior smoothed down. The edges of the hole are not worn. PREVIOUS ILLUSTRATIONS AND DESCRIPTIONS: Sherds of this type are described by Hewes as “Type 9” and as “Type 11” from Paul Brave, and illustrated by several sherds (1949 b, pp. 64-67; pl. 5, lower right, tw; pl. 6, left, i; pl. 6, right, m, 0, 7r). The present type includes those sherds described by Hewes as “Fort Yates Plain.” Riggs Cross-Hatched Rim (PL. 4, a-b, f) SAMPLE: 54 rim sherds, representing 43 vessels. DECORATION : Lip: Oblique tool impressions occur on six sherds, and crosshatched tool impressions appear on two sherds. Rims from one vessel are indented along the lip with oval punctates; the remaining lips are plain. Rim: The rim is filled with crosshatched lines between the lip and the neck. On 11 sherds there is a horizontally incised line encircling the rim below the lip. Shoulder: The four decorated shoulders have horizontally incised lines. Two of them have a “drooping corn stalk” and one of them has an in- cised triangle over the horizontal lines (pl. 4, f). ForM: Lip: Predominantly flat, with some round lips. Thickness, 7 to 9 mm. Rim: Straight to somewhat outflared. Height is between 30 and 43 mm.; thickness, 6 to 8 mm. Neck: The rim joins the shoulder in a smooth curve. Shoulder: Rounded and steeply sloping. Size: The orifices of two vessels, projected from large rim sherds, were 130 and 190 mm. in diameter. Appendages: Small triangular tabs are on the lips of four sherds. The crest of each tab is incised. Sherds of one vessel with an incised tab has, in addition, two plain vertical tabs. PREVIOUS ILLUSTRATIONS AND DESCRIPTIONS: Sherds of this type are included in the pottery described by Hewes as “‘Type 9, Fort Yates Fine Incised,” and one sherd is illustrated (1949 b, pp. 64-65, pl. 6, right, n). CoMMENTS: The type originally defined by Hewes as “Fort Yates Fine Incised’”’ includes rims that were straight and flaring as well as S-shaped rims. ‘The straight or flaring rims bearing crosshatched designs are herein described as “Riggs Cross-Hatched Rim,” and the S-shaped rims with the same decora- tion are herein classed as “Fort Yates Cross-Hatched Rim.” Riggs Incised Rim (PL. 8, c) SAMPLE: 6 rim sherds, representing 6 vessels. DECORATION : Lip: One sherd is obliquely tool impressed. The remaining lips are plain. Rim: Four to seven horizontally incised lines occur, with oblique lines inter- rupting the horizontals (pl. 3, c). 661—932—64——_-3 20 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 ForM : Lip: Round to somewhat flat. Thickness, 4 to 6mm. Rim: Straight to somewhat outflared. Height, 28 to 32 mm.; thickness, 6to 8mm. Riggs Pinched Rim (Pl. 3, 4) SAmpLe: 26 rim sherds, representing 22 vessels. DECORATION: Vessels are plain except for the finger pinching. ForM: Lip: May be round, fiat, or pointed. The lips are wavy, a condition induced by pressing the moist clay with the thumb and forefinger offset, one inside and one outside the vessel. The resulting undulation of the lip, when ob- served from above, is similar to that of Stanley Wavy Rim (Lehmer, 1954: 43-44; pl. 12). Rim: Straight to somewhat outflared. Height, 12 to 52 mm.; thickness, 5 to 7 mm. Neck: Some constriction on outflaring rims. Appendages: One loop handle, originating in the mid-rim, is attached to the upper shoulder. The handle projects straight out from the rim, and defines a right angle before it joins the shoulder. PREVIOUS ILLUSTRATIONS AND DESCRIPTIONS: A sherd of this type is illustrated in Will and Hecker (1944, pl. 18, the left sherd in the fifth row from the top} as “Archaic Mandan.” FORT YATES WARE Fort Yates Cord Impressed Rim (Pls. 3, a; 5, c-j) SAMPLE: 136 rim sherds, representing 106 vessels. DECORATION : Lip: None. Rim: Three to six evenly spaced, horizontally applied cord-impressed lines encircle the vessel rims. The diameter of the cord used varies from 2 to 3 mm., with the majority 2mm. in diameter. Sixty-six sherds have triangu- lar elements. On 7 sherds, the apex of the triangle is angular; on 25 sherds the apex is rounded, forming a curvilinear design. The triangles are formed by either two or three lines. A sherd from one vessel (pl. 5, h) has six horizontal lines of cord-wrapped stick impressions, and two oblique lines. One sherd has a red stain, probably ocher, on both the interior and exterior (pl. 5, e). Form: Lip: Predominantly flat, but may be round. ‘Thickness, 6 to 8 mm. Rim: Outflaring, with a recurving lip, resulting in an S-shaped rim. Height, 26 to 42 mm.; thickness, 7 to 12 mm. Neck: Constricted, with the rim joining the shoulder in a smooth curve. Shoulder: Rounded and steeply sloping. Size: The orifices of three vessels, projected from large rim sherds, were 130, 176, and 260 mm. in diameter. Appendages: A small, triangular tab is present on one rim (pl. 5, f). A single sherd appears to be part of a small strap handle with two vertical cord-impressed lines (pl. 3, @). Riv. Bas. S;)«6 PAUL BRAVE SITE—WOOD AND WOOLWORTH 21 PREVIOUS ILLUSTRATIONS AND DESCRIPTIONS: Sherds of this type were originally defined by Hewes as “Type 10, Fort Yates Cord Impressed,’ and two sherds are illustrated (1949 b, pp. 65-66, pl. 5, lower right, v, pl. 6, right, q). CoMMENTS: The present type description largely duplicates the original type description by Hewes. Pottery of this type is similar to the type “Aldren Cord Impressed” at the Thomas Riggs site (Kleinsasser, 1953, p. 27; fig. 29, 1-3, 6). The sherds designated as “Fort Yates Cord Impressed” from the Thomas Riggs site (Kleinsasser, 1953, p. 27; fig. 29, P/16, 4-5), however, do not fit the type description of Hewes and should be renamed (Wheeler, 1954, p. 8; 1955, p. 398). Fort Yates Cross-Hatched Rim (Pls. 4, c-e; 6, h) SampLr: 30 rim sherds, representing 15 vessels, and one restored vessel. DECORATION : Lip: None. Rim: Between the lip and the neck the rims are incised with crosshatched lines. A horizontal incised line encircles the rim below the lip on sherds from several vessels (pl. 4, e). On six rims there are punctates spaced 10 mm. apart on the rim below the crosshatched lines. Shoulder: The shoulders of two sherds are horizontally incised. ForM : Lip: Predominantly flat, with some round; thickness, 6 to 8 mm. Rim: Outflaring, with a recurving lip, resulting in an S-shaped rim. Height, 21 to 41 mm. ; thickness, 6 to 10 mm. Neck: Constricted, with the rim joining the shoulder in a smooth curve. Shoulder: Rounded and steeply sloping on large sherds. Some smaller sherds suggest a shoulder more nearly flat. Size: A partially restored vessel, 90 mm. high, has a maximum diameter of about 130 mm. When filled to the neck it contains 1 pint of water (pl. 6, h). The orifice of one vessel, projected from a large rim, was 168 mm. in diameter. Appendages: Scars on two rims indicate the presence of tabs on the lip. One sherd retains a tab with an apical incision. PREVIOUS ILLUSTRATIONS AND DESCRIPTION: Several sherds of this type are in- cluded in the pottery described by Hewes as “Type 9, Fort Yates Fine Incised,” and one sherd is illustrated (1949 b, pp. 64-65, pl. 6, left, ic). ComMMENTs: See the statements under the type “Riggs Cross-Hatched Rim” above. UNCLASSIFIED WARES Ezample A (PL. 5, @) SAMPLE: 16 rim sherds, representing 12 vessels. DECORATION : Lip: Oblique tool impressions occur on two rims. Rim: One of the rims with a decorated lip is crosshatched. Another rim has two horizontally applied cord-impressed lines on the interior lip, oblique cord impressions on the outer lip, and five horizontal cord-im- pressed lines on the exterior mid-rim. The remaining rims are plain. 22 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Form: Lip: Round. Thickness, 3 to 5 mm. Rim: Outflaring, with a protruding ridge centered on the rim exterior and channeled on the rim interior. Height, 37 to 55 mm.; thickness, 5 to 9mm. Neck: Constricted, with the rim joining the shoulder in a smooth curve. Appendages: Three rims have small, vertical, plain tabs. PREVIOUS ILLUSTRATIONS AND DESCRIPTIONS: Several sherds of this example were described by Hewes as “Type 6,” and one rim is illustrated (Hewes, 1949 b: pp. 62-63 ; pl. 6, left, 2). Hzample B (Pl. 5, b) SAMPLE: 1 rim sherd. DECORATION : Lip: Oblique indentations. ForM: Lip: Round, with a thickness of 6 mm. Rim: Inflaring near the lip. Appendages: The scar of a triangular, horizontal lug is on the outer rim 12 mm. below the lip. CoMMENTs: This rim form is known only from Paul Brave. EHzample O (P1. 5, c) SAMPLE: 1 rim sherd. DECORATION : Rim: Oval indentations occur on the lower rim. Form: Lip: Round, with a thickness of 6 mm. Rim: S-shaped, with a height of 27 mm. Neck: Constricted. CoMMENTS: This rim form, and the particular design involved, is known only from Paul Brave. Hzample D (Pl. 5, d) SAMPLE: 3 rim sherds from 2 vessels. DECORATION : None. ForM: Lip: Round, with a thickness of 6 to 7 mm. Rim: S-shaped, with heights 25 to 30 mm., and 8 to 9 mm. thick. Neck: Constricted, with the shoulder joining the rim in a smooth curve. Shoulder: Rounded. CoMMENTS: These rims are similar in form to those of Fort Yates Ware, and may be a plain variant of that ware. MINIATURE VESSELS Three minature vessels were in the hard ash of the primary fire- place in House 3. Evidently they had been subjected to an intense secondary firing, for the pots were soft and crumbly. They were treated with a mixture of acetone and ambroid. This treatment gave Riv. Bas. Sur. BAS. No ga] PAUL BRAVE SITE—WOOD AND WOOLWORTH 23 the vessels a light-reflecting quality that they lacked before. Two of them are complete, and both have the following dimensions: height, 68 mm.; maximum diameter, 87 mm.; orifice at inner lip, 75 mm.; neck diameter, 74mm. The size of the incomplete specimen would be much the same. The surfaces are irregularly smoothed, and the shoul- ders are steeply sloping. Two of them have outflaring rims (pl. 6, f) and the other has a constricted neck and a nearly vertical rim (pl. 6, g). Each of the pots contains slightly more than a quarter of a pint of water when filled to the neck. A vessel fragment indicating an orifice of 56 mm., with an outflar- ing rim 8 mm. high, has a steeply sloping and irregularly smoothed shoulder. Another fragmentary miniature is decorated with incised lines (fig. 2, @). This vessel has an estimated orifice diameter of 80 mm., with an outflaring rim 11 mm. high. The surface is irregularly smoothed. Fragments of two other vessels have outflaring rims. One of them shows the scar of a handle that was welded to the lip and riveted to the upper shoulder. Bowls may be indicated by three small sherds, two of which have indented lips. BODY SHERDS The majority of sherds from the site are simple-stamped or smoothed. Some 514 sherds are classed as simple-stamped and bear the characteristic grooves resulting from the malleating of the moist clay with a grooved paddle. The 845 smooth sherds are irregularly smoothed, and only a few of them might be classed as polished. The polished sherds have a low light-reflecting surface. Many of the smoothed sherds show irregularities which suggest that the surface was originally simple-stamped. The 54 decorated body sherds are incised or trailed on the shoulder, which is usually smooth but which may reveal partially obliterated traces of vertical simple-stamps. The decoration may consist of either fine line incising or deep trailing, sometimes in combination with punctates. The width of the lines varies from 0.5 to 6.0 mm., and cross sections are either V-shaped or U-shaped. One of the more common designs consists of alternating “bear tracks” and chevrons (fig. 1, a-b, d@). In only one instance do the “tracks” have more than five “digits” (fig. 1, ¢). A design termed the “drooping corn stalk” is incised on shoulders that are covered with horizontally incised lines and a wavy or zigzag lower border. Oblique incised lines and punc- tates also occur on the same background (fig. 1, f-g). Other fragmen- tary designs consist of oblique, vertical, and horizontal lines in several combinations (fig. 1, 2; fig. 2, b-c). Broad-trailed concentric circles occur on the shoulder of one large rim sherd, and on one body sherd. 24 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 One of these has four punctates in the center (fig.2,@). ‘Two separate designs that are not duplicated in the collections are incised on the shoulder of a small vessel (fig. 2, @). Nineteen body sherds differ from the majority of sherds at the. site in being cord-roughened. These sherds are from vessels that were first malleated with a paddle wrapped with cord, then partially smoothed (pl. 6,¢). Some of the sherds are from the shoulder area, with the roughening extending to the neck. On these sherds the roughening is vertical. The twist of the cord is clearly distinguish- able on three sherds. The twist on two sherds may be duplicated by holding two cords in the left hand and turning the ends to the right with the right hand; on the other sherd the cords were turned to the left with the right hand. The diameter of the cords ranged from 1.0to2.0mm. Thenumber of vessels bearing cord-roughened surfaces may have been greater than the 19 sherds would indicate, but the smoothing and obliteration of the cord marks render the original treatment indistinguishable. The paste of the sherds is the same as that of the Riggs Ware and the Fort Yates Ware and it is likely that they are from vessels of these wares. However, none of them are assignable to specific types. Two sherds are check-stamped. Their exteriors are covered with small, depressed squares bordered by low, partially smoothed ridges (pl.6,d). Thestampsare 3 to5 mm. on aside. Biconical perforations in four body sherds may have served as holes for lacing cracked vessels together. Six loop handles and one strap handle are detached from the vessel rim and cannot be assigned to types. The strap handle is oval and the loop handles circular in cross section. They are all smooth and plain. The handles extend from the lip to the upper shoulder. MISCELLANEOUS OBJECTS OF BAKED CLAY Animal effigies (5 specimens) One of the specimens (fig. 6, @) represents a prairie chicken, or pin- nated grouse (Z’ymfanuchers americanus). 'The general conforma- tion of the body and the structure of the tail conform to this species, although there are some superficial structural similarities to the tur- key. The turkey was not native to this region in early historic times. The head is lacking. The length is 35 mm. and the height 28 mm. A specimen 32 mm. long has a circular body with two front legs (fig. 6,6). The hind part of the body rests on a tail. The modeling is too generalized to permit identification, but it may be a beaver or badger. Riy. Bas. Sur. ine Dap. No 33] PAUL BRAVE SITE—WO0OD AND WOOLWORTH 25 A third specimen may represent the head of a canid. In the illus- tration (fig. 6, c) the object is so oriented that it appears to have two long pointed ears and a sharp nose. These characteristics mark the swift or red fox, both of which were native to the locality in his- toric times. An incomplete specimen suggests a round-bodied, but unidentifiable, animal. There are two short, stubby forelegs but the hind limbs are lacking. It is apparently a heavy-bodied animal (fig. 6, d). The final specimen is an effigy handle from a vessel lip. The profile of the head (fig. 6, ¢), which extends 15 mm. from the lip, suggests that of a bear. Bead (1 specimen) A perforated baked clay bead is oval in longitudinal section and pentagonal in cross section. It is 20 mm. long, with a maximum diam- eter of 18 mm. One intact end is concave. The hole appears to have been made by inserting a grass stem that burned away during the firing (fig. 6, f). Shaped item (1 specimen) An irregular, bowllike object, probably modeled around the finger of a child, has a fingernail impression in the base of a shallow cavity. The base is flat, with a protrusion resulting from a downward pressure of the finger. The maximum diameter is at the midpoint. Height is 12 mm., and the diameter is 19 mm. WORK IN STONE Projectile points (75 specimens) Seventy-five artifacts are classed as projectile points, in that they are small, thin, bifacially worked, approximately symmetrical, and well made. Materials used for the points are: Knife River flint (chal- cedony), 42; gray, brown, and black cherts, 28; white chalcedony, 4; and agate,1. The 55 complete points are classed in 5 categories based on outline and other characteristics. The terms applied to the five projectile-point categories are descriptive (see Davis, 1956, pp. 64-69), but reference is also made to the point outline taxonomy published by Strong (1935, pp. 88-89). Plain lanceolate, convex base (Strong NAb1).—A single specimen from the site has convex blade edges and a convex base. It measures 28X15X3 mm., and has a weight of 1.5 om. (fig. 3, a4). It is un- notched and is evenly and bifacially flaked. 26 BUREAU OF AMERICAN ETHNOLOGY Ficure 3.—Projectile points. [Bull. 189 Riv. Bas. Sur. Pap. No, 33], PAUL BRAVE SITE—WOOD AND WOOLWORTH 27 Pla lanceolate, straight base (Strong NBa).—The 18 examples in this category have gently convex edges and a straight base (fig. 3, b-c). This category is unnotched, with even bifacial flaking. Length: Mean 27.8 mm., range 23 to 42 mm, Width: Mean 15.6 mm., range 11 to 20 mm. Thickness: Mean 3.8 mm., range 2 to 5 mm. Weight: Mean1.5 gm., range 0.9 to 3.0 gm. Plain triangular, concave base (Strong NBb).—The two points in this category have gently convex edges and a concave base. The tips of both specimens are missing (fig. 3, d). The category is unnotched, and is evenly bifacially flaked. Dimensions (estimated when broken) are: 21X15X3mm. and 24X12X2mm. The former point weighs 1.1 gm., and the latter, 1.0 gm. Plain lanceolate, straight base (Strong NBa1).—This is the largest single category, including 22 specimens. They have gently convex edges, a straight base, and two side notches (fig. 8, e-/). They are evenly bifacially flaked, and the blade edges usually have fine second- ary flakes removed. Points are tapering and sharp. Length: Mean 30.2, range 19 to56 mm. Width: Mean 14.1, range 11 to 21 mm. Thickness: Mean 3.8, range 2 to 7 mm. Weight: Mean 1.8 gm., range 0.6 to 6.4 gm. Plain lanceolate, concave base (Strong NBb1).—The 12 specimens in this category have straight to gently concave edges, a convex base, and two side notches (fig. 3, g-h, ¢-/). They are evenly bifacially flaked but the flaking is less delicate on larger examples. Points are sharp and tapering to blunt. Length: Mean 33.4 mm., range 21 to 60 mm. Width: Mean 15.6 mm., range 12 to 24 mm. Thickness: Mean 4.4 mm., range 3 to 10 mm. Weight: Mean1.8 gm., range 0.8 to 2.0 gm. E'nd Scrapers (149 specimens) The majority of these scrapers are made from Knife River flint. Other stone includes a gray chert, petrified wood, and agate. The outlines tend to be triangular to rectangular, although many are ir- regular. The working edges are steeply chipped on the end opposite the bulb of percussion, and the concave undersides are unmodified. Three groups are distinguished. Group I (78 specimens) —Cross sections are triangular to shallowly U-shaped, with flaking over the entire upper surface, and usually with retouching along the sides. Lengths range from 20 to 58 mm., and 28 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Ficure 4.—End scrapers and knives. widths from 15 to 28mm. The working edge is rounded to straight in 43 specimens (fig. 4,a@). In 19 specimens, it is skewed with respect to the long axis, sloping from the lower left up to upper right (fig. 4, 0). In 16 specimens, the blade slopes down from upper left to lower right (fig. 4, ¢). Riv. Bas. Sur. ies Bo. Noo 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH 29 Group 2 (70 specimens) .—Cross sections are triangular, shallowly U-shaped, or irregular, since many of the specimens are made from irregularly shaped flakes. There is little flaking on the upper surface and there is some retouching on the edges. Lengths range from 21 to 51 mm., and widths from 16 to 27 mm. In 47 specimens, the work- ing edge is rounded to straight (fig. 4,e). In 19 specimens, it slopes from the lower left up to the upper right (fig. 4, f) ; im 4 specimens, the blade slopes down from the upper left to lower right. Group 3 (1 specimen) —One large scraper, triangular in cross sec- tion, has a convex, flaked upper surface and an unmodified concave under surface. Larger than any scrapers in the preceding two groups, this implement measures 56 X 34 X 22 mm. (fig. 4, d). Drills (6 specimens) Two slender shafts of Knife River flint and one of reddish-brown chert are 44 mm. long, with expanding bases. A fourth specimen is a base, with the drill shaft absent. They are bifacially flaked, with lenticular to diamond-shaped cross sections (fig. 5, a). One specimen consists of a narrow shaft of Knife River flint, 55 mm. long, with a round base. Near the tip it is bifacially flaked, but at the base only one side is worked. A final example is represented by a unifacially flaked drill shaft tip. Broad Knives (21 specimens) All knives are bifacially flaked and are lenticular in transverse section and longitudinal section. Flaking is random, and the blade edges are retouched. The form outline used here follows that devised by Strong (1935, pp. 88-89). Group 1, NAb1 (1 specimen) —The base is round, sides are convex, and the tip is blunt. Material is a gray chert; it measures 38 X 26 X10 mm. Group 2, NAb2 (4 specimens) —Bases are straight, sides are convex, and the tips are pointed but not sharp. Materials include gray chert and Knife River flint. Lengths range from 40 to 94 mm., widths from 23 to 35 mm., and thicknesses from 5 to 6 mm. (fig. 4, 2). Group 3, NAb2 (2 specimens).—These knives are similar to the Group 2 specimens but they are asymmetrical. The base is straight, but one side is nearly at right angles to the base, while the other side is at about 45 degrees to the base. The sides are gently convex, and the tips are blunt (fig. 4, 7). Materials are gray and black chert. Lengths are 32 and 50 mm., widths 24 and 34 mm., and thickness 8 and 9 mm., respectively. 30 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Group 4, NE (1 specimen).—Oval in outline, this specimen of Knife River flint is 44 mm. long, 28 mm. wide, and 12 mm. thick. The ends and sides are convex. Group 6, base fragments (6 specimens) —These five specimens are . convex base fragments from knives that had gently convex to straight sides. Materials include gray chert, Knife River flint, and petrified wood. These fragments are from knives that were originally more than 62 mm. long. Group 6, blade tips (7 specimens)—These tips are from knives which had pointed but not sharp tips, and had convex sides. Lengths are in excess of 121 mm., and widths are from 22 to35 mm. Materials are Knife River flint, gray chert, and Bijou Hills quartzite. Narrow knives (5 specimens) Long slender knives, bifacially flaked, are made from Knife River flint. Two complete specimens are NAb2 in outline and one is NAa. Two are tips. The sides are gently convex and nearly parallel. Lengths are 66, 72, and 83 mm.; widths, 21, 14, and 18 mm.; thickness, 5,5,and 6 mm., respectively (fig. 4,2). Choppers (34 specimens) Group 1 (12 spectmens).—These tools are oval to nearly circular, and are bifacially flaked, with large flake scars on both faces and retouching on the edges. Battering is evident on some of the ends. The material includes gray chert, quartzite, and quartz. They range in length from 70 to 172 mm.; in width, from 41 to 99 mm.; in thick- ness, from 18 to 28 mm. Group 2 (10 specimens).—Roughly rectangular blocks of quartzite are bifacially flaked along one edge. The faces of the implements are unmodified. There is some use retouching along the blade edge, and the blades extend the length of the tool. Lengths are 90 to 205 mm.; widths, 50 to 90 mm. Group 3 (12 specimens) —Irregular spalls of stone, including gran- ite, quartzite, and gray chert have bifacially flaked edges along one edge of the stone and show some retouching. The specimens attain a maximum diameter of 130 mm. Flake knives (135 specimens) Flakes of Knife River flint, petrified wood, gray chert or quartzite are classed as flake knives when they have a prepared edge on one or more edges. Riv. A Sis Riv. Bas. Sut. PAUL BRAVE SITE—WOOD AND WOOLWORTH 31 Unmodified flakes (203 specimens) A sample of stone was saved from each excavated feature. Many of the pieces retained were unused, but some showed minute flaking along one or more edges, indicating that they had been used as cutting tools. Materials included Knife River flint, petrified wood, gray chert, agatized wood, quartzite, and quartz. Grooved mauls (5 specimens) Each of three granite cobbles, from 87 to 107 mm. long, has a groove pecked around the small diameter of the stone. On two examples the groove encircles the stone; on a third, the groove encircles three- quarters of the stone. Two fragments of much larger mauls are not large enough to permit an estimation of the original extent of the groove. The mauls consist of oval cobbles shaped only by the addi- tion of the pecked groove. Both ends are battered by use, but the stone 1s not otherwise modified. Hammerstones (49 specimens) Cobbles and pebbles of granite, siltstone, quartz, and fine-grained sandstone are classed as hammerstones when one or more edges are battered. Group 1 (45 specimens) —These hammers are made from oval stones. They are either battered on one or both ends or battered over the circumference of the stone. Lengths range from 39 to 221 mm. Group 2 (4 specimens).—In a second group of hammerstones each is nearly circular in outline and has a rectangular cross section. The faces of this group are ground smooth and flat, and they are bat- tered on all edges. One specimen has a small depression pecked into one face. Diameters are from 85 to 110 mm. Ground stone spheres (5 specimens) Four of these objects are made of a buff siltstone and one is of a reddish-brown concretionary sandstone. The entire surface of the sandstone sphere and of one of the siltstone spheres is smoothed. The remaining specimens are partially smoothed and ground, and may be unfinished pieces. Diameters average 70 mm. ys BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Shaft smoothers (6 specimens) These abraders consist of shaped blocks of a coarse, buff sand- — stone with either rounded or squared ends, convex sides, a flat grooved surface, and a flat or convex undersurface. The grooves are straight and are either V-shaped or U-shaped. One complete abrader is 161 mm. long. Another specimen, 64 mm. long, has a second grove on the undersurface. The illustration (fig. 5, /) depicts a fragment with a squared end. Grooved abraders (18 specimens) These objects are composed of a buff or rust-colored, coarse-grained sandstone. There is considerable variation in size, ranging from specimens no more than 75 mm. in diameter to those 175 mm. in diameter. There is no consistent form; the pieces were not altered in form except for the V-shaped or U-shaped grooves worn into their surfaces. The grooves are of various lengths, depending upon the available surface. The grooves are commonly convex in longi- tudinal section. Faceted abraders (59 specimens) Fifty-two pieces of scoria and seven pieces of sandstone, none of which exceeds 105 mm. in diameter, have been used as abraders. These objects consist of irregular pieces of stone that are faceted from rubbing or grinding. In some instances the entire stone is smoothed, and some such stones have a superficial] resemblance in size and form to the bone abraders made from the cancellous tissue of bison humeri. Many of the stones also bear small, shallow V-shaped or U-shaped grooves. Beads (4 specimens) A tubular piece of concretionary sandstone, 40 mm. long and 11 mm. wide, has a natural hole along its long axis. The ends of the hole are smoothed (fig. 5,0). Two circular disks of scoria are biconically pierced. One of them is 8 mm. (fig. 5, ¢) and the other 12 mm. in diameter (fig. 5, @). The latter bead is not complete, since the perforation is not finished. A scoria bead 16 mm. long and 15 mm. wide is split down the long axis, showing that the perforation was drilled from both ends. One end is ground flat (fig. 5,¢). Riv. Bas. Sur. ae Riv. Ras. 31 PAUL BRAVE SITE—WOOD AND WOOLWORTH 33 0 5 fe) eee ——eEeEEeeEeEeEeEeEeEeEeEeEeEeEe cm. Ficure 5.—Chipped and ground stone artifacts. Mealing slab (1 specimen) A large rectangular block of fine-grained sandstone, 460 mm. in length, has one irregularly smoothed surface. Near one edge of this surface is a smoothed and polished depression. Near the edge of the 34 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 depression the stone is pebbled in such a manner that it may have resulted from pecking. The opposite surface of the slab is convex and irregular, and the high points have been ground down. Mano (1 specimen) This object is made of an irregular granite stone measuring 120103386 mm. One smoothly convex surface is stained red, prob- ably by ocher. The stone is pecked smooth and the high points are ground and polished. Fine striations occur along the long axis of the stone, and indicate that it was used with a back-and-forth motion. The sides are battered from use as a hammer. Aw (1 specimen) A block of granite is polished and shaped so that the transverse section is rectangular with rounded edges (fig. 5,g). One end bears a fractured, wedge-shaped bit; the other end is broken. Dimensions are 110 72X32 mm. Celts (18 specimens) There are six complete specimens of celts which range from 122 to 170 mm. in length; diameters are 47 to60mm. ‘Transverse sections are oval to nearly circular; only one example is rectangular. The max- imum width is at the midpoint on five specimens; on one, it is at the cutting edge. Polls are battered from use as hammers. The cutting edge is wedge shaped and blunt. Nearly the entire surface of the stone is pecked and smoothed, and on some specimens there is no trace of the original surface (fig. 5,4). Fragmentary celts suggest that they are from implements exceeding 170 mm. in length. In addition to the 18 recognizable celts, there are 6 large blocks of diorite, all of which show some pecking on high points of the stone. Some of these are pecked in a manner which suggests unfinished celts. Unmodified calcite (10 pieces) These pieces of calcite, none of which exceed 117 mm. in length, are unmodified. One of them, 75 mm. long, has a form spuriously like that of an arrowpoint. Unmodified pebbles (5 specimens) These oval pebbles are 24 to 58 mm. long. They show no signs of use, and the surfaces suggest that they were steam rounded. Riv. Bas. Sur. Bao ONG. 331, PAUL BRAVE SITE—WOOD AND WOOLWORTH 35 Pigments (15 pieces) Fourteen small, irregular fragments of chalk were probably bases for pigment. The colors of these pigments given below approximate those of Maerz and Paul (1930). Three of the chalks are nearly white (10 A 1), and 11 are yellow, approaching in tone Pinard Yellow (9 K 2). Powdered hematite from one feature is Java Brown (8 L 10). The white chalk is fairly soft, approximating in hardness the consis- tency of softer grades of schoolroom chalk. The yellow chalk is more compact, and is about the hardness of talc. WORK IN BONE Scapula hoes (109 specimens) The scapulae of adult bison were used in the manufacture of this implement, although some smaller and more delicate specimens may have been from young bison or from elk. The supra scapular border is beveled on the side that bears the scapular fossae, and these fossae are hacked away so that the surface is nearly level. The edges of the implements are roughened by chopping at a distance of 100 to 150 mm. from the cutting edge, probably to provide a surface for binding a handle to the tool. The cutting edge may be rounded or straight, largely depending on the amount of wear. The articular end is re- tained without modification (fig. 6, 7). Notches occur on the side of nine implements near the articulating end. The notches may occur on either side (fig. 6,2). One hoe has a hole 12 mm. in diameter in the blade 115 mm. from the articulating end (fig. 6, 4). Two hoes have deep, U-shaped indentations in the blade (see Hurt, 1953, fig. 19, 4-5). The edges of these indentations are smoothed, possibly from use as thong stretchers. The range in length of complete hoes is 250 to 410 mm., the shorter specimens show- ing much evidence of use. Serrated fleshers (2 specimens) The one complete specimen, 310 mm. long, was made by cutting the shaft of a bison metatarsal diagonally to produce a chisel edge. The edge is serrated (fig. 7, 2). The implement also includes several of the ankle bones, which were left in place to provide additional lever- age. The shaft of the metatarsal is highly polished near the serrated end, but the rest of the metatarsal and the remaining bones have a natural finish, suggesting that the tool originally retained cartilage and hide over the ankle and heel bones. A fragment of a second serrated specimen was recovered. 661-932—64——4 [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 36 10 ass aves Sa tre Sr EME SESS Ficure 6.—Baked clay effigies and scapula hoes. PAUL BRAVE SITE—WOOD AND WOOLWORTH 37 Riv. Bas. Sur. Pap. No. 33] Ficure 7.—worked bone. 38 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Split metapodial scoops (2 specimens) These implements were made by longitudinally splitting a bison metapodial and sharpening the middle of the shaft to a cutting edge. Lengths are 140 and 250 mm. The articulating ends are cut away so that the center of the tool forms an open, U-shaped trough. It is likely that this trough provided a place for the handle (fig. 7, 7). Bone disks (3 specimens) These specimens are cut from the walls of large long bones, and their outlines are oval. Their edges are rough and show little smooth- ing (fig. 8, a-c). They are 5 mm. thick, with diameters of 12 to 21 mm. The cancellous bone is ground away and the surfaces are lightly polished. A stain, probably red ocher, occurs on one disk. Long bone beads (8 specimens) Short sections of long bones of small mammals or birds were used for beads. The ends are transversely cut, and are imperfectly smoothed or left irregular. They are 14 to 29 mm. long and 5 to 9 mm. in diameter (fig. 8, d-g). Fishhook (1 specimen) The compact outer layer of a bird long bone served as raw material for the hook, which is 27 mm. long. The point is unbarbed, and the attaching end is bulbous, with shallow grooves which served to attach the hook to a line (fig. 8, 2). Spatula tip (1 specimen) A thin piece of bone, 47 mm. long, is smoothed and highly polished. One end is broken, and the other is rounded. The edges are sharp and even (fig. 8,2). Whistle (1 specimen) A section of the wing bone of a large bird is 110 mm. long. One end is transversely cut, polished, and smooth. On the other end is a V-shaped cut that served as the whistle opening. The instrument is highly polished, and the bone protuberances to which the quills were attached are reduced so that the surface is smooth (fig. 8, 7). 39 PAUL BRAVE SITE—WOOD AND WOOLWORTH. Riv. Bas. Sur. Pap. No. 33] Ficure 8.—Worked bone. [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 40 FicureE 9.—Bone awls. PAUL BRAVE SITE—WOOD AND WOOLWORTH 41 33] Riv. Bas. Sur. Pap. No. Ficure 10.—Worked bone, 42 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Tubes (3 specimens) Sections of two bird long bones are transversely cut on both ends. Lengths are 113 and 120 mm. (fig. 8, %). A third specimen, 235 mm. long, is apparently a deer tibia (see fig. 7, a). Chisels (9 specimens) Fragments of large long bones with one end beveled bifacially to form a narrow, sharp cutting edge may have served as chisels or gouges. The tips of two specimens are illustrated (fig. 8, 7-m). Two of the complete tools are 140 and 152 mm. long. The butt end is rounded and polished, providing a convenient grasping surface. The cutting edge may be nearly straight or slightly convex. The can- cellous bone is ground away so that surfaces are smooth, and the blades have a glossy polish. Avwls (100 specimens) The awls ordinarily are sharp and tapered, although some have relatively blunt tips. Each of the tools with tips sharp enough to have served as hide perforators is classed as an awl. The classifica- tion of these implements follows that of Kidder (1932), with modifica- tions as demanded by the material at hand. The awls are divided into four groups, based on the material from which they are made: ia Mammal long-bones22 24222! oe 2 re eee ee eee 62 AeA OLpone JElt intact. 22s sR SS eee al b. Head of bone unworked except by original splitting__._ 18 c Head of- bone partially workedis222285. 2-2 12 da Head, of bone wholly removed]. = 222222 ae LT Cm plinier (aWIS22-2 Fe 2 6s Be ee eee 14 2; Mammal uribs2 = 225 2 2 oe ee eee 33 CS Split TiDS2 2. ee oe ee 28 Oz Splinter” AWS i.e 2 LN eee 5 oS. bird lone DOnes ses lee See 2 2 ee eee oe eee 4 4. Wish> spine sales 2}. Ss 2 aie Rs Se a Be Se eee 1 MaMMAL LONG BONES (62 SPECIMENS) : Head of bone left intact (1 specimen) .—The single awl of this group is made from the ulna of a small mammal, possibly a canid. Length is 88 mm.; the tip is pointed but blunt (fig. 9,@). Head of bone unworked except by original splitting (18 speci- mens).—The specimens in this group are made from the metapodial of deer, or from bones comparable in size and form. The metapodial illustrated (fig. 7, 6) is 210 mm. long and the length of the awls in- dicates that only about half of the length of the bone was used. The metapodials were split by longitudinal sawing in the U-shaped trough on one side of the bone and by wedging apart the proximal end, using the resulting half, third, or quarter of the ends as butts. Riv. Bas. Sur. My Dag. No 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH 43 Two awls made from the proximal end of the bone are 63 mm. long (fig. 9, 6). A third specimen is from an immature animal and the epiphysis is detached. Length is 55mm. (fig. 9,¢). Fifteen awls made from the distal end of the bone are 50 to 121 mm. long. The longer ones are thin and evenly tapered, but shorter specimens are stubby (fig. 9,d). Head of bone partially worked (12 specimens).—One specimen from the distal end of a deer metapodial is 186 mm. long, with a long, tapering shaft. The butt is smooth and rounded, all articulating facets and rough projections having been reduced, leaving a smooth swelling grip. Eleven awls are made from the proximal end of deer metapodials, which are split lengthwise. The ends are smoothed and rounded, with little of the original surface remaining. Length is 60 to 177 mm. (fig. 9, 7). | Head of bone wholly removed (17 specimens).—These tools have rounded and smoothed butts and short, tapering points. Each awl is split from a long bone. Most specimens have a groove down one side (fig. 9, ¢), a remnant of the central cavity of the parent bone, but some are completely smooth, with no evidence of the original surface (fig. 9, g). Lengths are 63 to 125 mm. Splinter awls (14 specimens) .—Fragments of long bones, the shape of which fitted them for use as awls after sharpening one end, are classed as splinter awls. There is no regularity of shape, and only the tip is worked. Lengths are 55 to 97 mm. (fig. 9, h). MamMaAt Rips (33 SPECIMENS) : Split ribs (28 specimens) .—These awls are split from a large mam- mal rib, probably bison, and most of the surface is dressed. The cancellous bone is removed or nearly obliterated. Sides are nearly parallel and butts are squared (fig. 10,5). The awls are homogeneous in form but vary from 61 to 181 mm. in length. Splinter awls (5 specimens).—These specimens are fragments of split ribs that served as awls after one of the sharp ends was ground toa point. Form is not consistent. Presumably the original splinter was chosen for convenience rather than form. Bird LONG BONES (4 SPECIMENS) : Each of these specimens is made from a whole bone, with one end sharpened toa stubby point. Twoawls, made from an ulna (fig. 10, c) and a femur, are 74 mm. long. Two awls are made from wing bones, the longest of which is 175 mm. long. FisH SPINE (1 SPECIMEN) : The spine of a catfish is sharpened to a blunt tip on the end opposite the articulating surfaces. The sharp ridges along one edge are ground away so that the shaft is smooth (fig. 10,2). These spines are some- times erroneously identified as fish mandibles. 44 BUREAU OF AMERICAN ETHNOLOGY [BuLL, 157 made one peculiar error, for a high area that presumably was meant to represent the ball of the foot is immediately adjacent to the heel and a depression was carved out, for what should be the instep, between the toes and the raised area. The length of the object is 4.6 cm., width 2.0 cm., thickness (to the edge of the break) 0.9 cm. _ A few pieces in the collection appear to have been intended to represent animal forms. One interesting specimen is made of ser- pentine in the form of a small fish (pl. 37, a, b, a’, b’). The eyes are formed by moderately deep circles cut with a hollow drill. Sawed lines at the edges of the mouth and around the gills give the object a more realistic appearance although actually it is highly stylized. It is drilled from end to end, that is, from mouth to tail, by means of two tubular drill holes which intersected just over 6 cm. from the mouth, or roughly two-thirds of the way back. These holes are about 1 cm. in diameter and have a very slight taper toward the base, presumably due to added wear during the drilling process at the upper part of each shaft. The holes almost fail to meet, having an overlap of about half a centimeter. Subsequently the object was sawed in half longitudinally. Whether this was done because the drill holes did not meet properly cannot be determined. The object is 9.1 cm. long and 4.7 cm. in maximum width. The original thickness prior to sawing was apparently a little under 2 cm. A small piece of pale-green jade was made by cutting what must have been a fairly good-sized cylindrical bead in half longitudinally (pl. 37, e, e’). The remaining piece is slightly less than a half cylinder in cross section. The lower end was broken off and repolished. A few shallow saw marks outline slanting elliptical eyes, the line across the muzzle, and what seems to be an indication of nostrils at the lower broken end. The top two biconical perforations, one on the face side, provide means of suspension and are connected by a sawed groove. The general effect is that of an animal head, but it is im- possible to try to guess the species represented. The present length of the object is 5.1 cm., the width is 1.8 cm., and the thickness 0.8 cm. A small buff-colored pebble, probably of serpentine, was slightly modified into the form of an animal head, possibly that of a dog (pl. 49, A). Shallow drilled pits indicate the eyes. A transverse biconical perforation runs from side to side at the base of the ears and another perforation goes through the lower lip and into the saw cut that represents the mouth. The length is 5.4 cm. A small carving representing a fairly realistic turtle, made, as I recall, of basalt or possibly diorite, and painted red, was found in the offering, but I did not find it with the collection in the Museo Nacional, and by some strange oversight no pictures of it seem to have been made in 1941. Nod)’ ~=©— «CERRO DE LAS MESAS OFFERING—DRUCKER 45 Two small representative objects of materials found with burial materials in Mound 30 may be added to our list, although they do not properly belong with the cache. One of these was a small carv- ing, possibly made from half of a subspherical bead, representing a monkey head (pl. 52, a, third row, middle). It has the characteristic wide crest on the head and protruding mouth parts. The eyes are small shallow drill pits. A transverse perforation runs through from side to side near the top of the head and two pairs of intersecting conical perforations connect sides and back just below the midpoint. The height of the object is 1.8 cm., the maximum width is 1.8 cm., and the thickness is 1.3 cm. The second little jade pendant from Mound 380 represents a highly stylized cicada (pl. 52, a, third row, middle). It is of bright-green jade with light-green mottling. In outline it is nearly elliptical with slight notches at the sides of the head and the base of the wings. Shallow saw cuts suggest the thorax of the insect, and two very shal- low drill pits near the upper rounded end, the eyes. A transverse biconical perforation pierces the object from side to side at about the level of the eyes. This little piece is well polished. Its length is 3.1 cm., the maximum width is 1.6 cm., and the thickness at the upper end is 1.0 cm. A little pebble of light-green jade was very slightly modified into a form that vaguely suggests the head of a wood duck (pl. 49, h). A slight projection at one end was bisected with a saw cut to suggest the beak, and a few light cuts on the other end suggest the feathers. A transverse biconical perforation runs through from side to side. The length is 2.9 cm., the width 1.8 cm., and the thickness 1.2 cm. PLAQUES There are a number of objects among the materials from the cache whose use is not known but which suggest by their more or less flat form and laterally balanced perforations that they were intended for suspension, perhaps hanging from necklaces like some of the objects portrayed on the various stelae. These pieces are grouped here as “‘plaques.”’ The first of the plaques is a small trapezoidal object of very light translucent green jade with a faint mottling of light green and white (pl. 38, a, a’, and fig. 3). The two upper corners are decorated with highly simplified animal heads in profile, formed by a very few sawed lines and with circular depressions made by a hollow drill to represent the eyes. The sides and lower edge of the stone is marked off with a border formed by a lightly sawed line. The central portion between the heads is also decorated with an angular pattern of sawed lines. 46 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 band is 220 mm. long, and is bent into an arc with a diameter of 140 mm, (fig. 11,@). Both ends are drilled from both sides and pierced. The transverse section is oval near the ends and circular in the middle. The convex surface is highly polished, and the concave surface is rough. ; A broken band 116 mm. long has a perforation drilled from both sides in a square end, and a groove in the concave surface that ex- tends for 60 mm. from the hole. The broken end is rounded and pol- ished, and it may have been reworked (fig.11,2). Four broken bands, none of which exceeds 80 mm. in length, have perforations in one end that are gouged from both sides. Surfaces are smooth but are not polished (fig. 11, g). One example is scored on both edges (fig. 11, kh.) A final fragment, a midsection, is 70 mm. long. A longitudinal eroove is cut into the convex face. Tines (4 specimens) The tips of these deer tines are lightly polished, with occasional longitudinal scratches near the tip. The proximal ends are irregular, and the fractures indicate that the tines were hacked from the body of the antler with some blunt tool. The lengths are 51 to 180 mm. Miscellaneous shaped objects (3 specimens) A rectangular piece of antler, measuring 41X17X5 mm., is cut transversely on both ends. The object was split longitudinally from the antler. The convex surface is scored lightly and smoothed; the under surface is rough (fig. 11, d). A second object measures 51X17X15 mm. One large end is sawed transversely through the compact outer layer, and snapped off. The object tapers to a squared, polished end. Its form is spuriously simi- lar to that of a modern pipe mouthpiece (fig. 11, e¢). The convex upper surface is smoothed on the high points, and the flat under side is longitudinally scored with deep gouges. Another item is broken, but originally exceeded 200 mm. in length. It consists of a straight shaft of antler 10 to 11 mm. in diameter. One end is squared, and the other end tapers to a broken tip. The sides of the object are scored, and five incised lines occur on either side. Four notches are cut into the sides of the shaft at the square end (fig. 11, f). Perforated elk teeth (2 specimens) Highly polished canines of adult ell have holes drilled from both sides in the roots. The holes are polished from a suspending cord or 47 WOOD AND WOOLWORTH PAUL BRAVE SITH Riv. Bas. Sur. Pap. No. 33] 1 teeth. ima Ficure 11.—Objects of antler and an 48 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 thong. One of them (fig. 11, 6) has three small incisions on one end. The other tooth is plain. Grooved incisor (1 specimen) A shallow groove, and part of a second, encircles the root of a large incisor, probably bison. It is not polished, and may not have been used (fig. 11, ¢). WORK IN SHELL Disk beads (10 specimens) Eight circular disk beads are 2 to 4 mm. thick and 10 to 12 mm. in diameter (fig. 12, a). The circular perforations are ordinarily drilled from one side of the bead. One specimen (fig. 12, 6) has a cylindrical hole that may have been drilled with a hollow reed. Two irregular specimens, larger than the finished beads described above, are 13 to 18 mm. in diameter. They are perforated, and the sides, although smooth, are not regular (fig. 12, c-d). Disks (10 specimens) Two disks are smoothly finished. One of them is 33 mm. in diam- eter and 5 mm. thick, and the other is 11 mm. in diameter and 4 mm. thick. The edges are smooth and rounded (fig. 12, #7). The eight remaining specimens are roughly circular pieces of shell 15 to 40 mm. in diameter. Edges are irregular or partially smoothed. These pieces may be rough-shaped blanks that were not brought to their final form (fig. 12, m,7). Pendant (1 specimen) A fragment of the shell of a Lasmigona complanata, with part of the hinge, is broken through a perforation drilled through a thin part of the shell (fig. 12,7). Fossils (22 specimens) Seven species of fossil shells were found that had been picked up elsewhere by the inhabitants and brought to the village. Five of the seven species are pierced or otherwise modified for suspension as beads or pendants. Three of the Z. nebrascensis shells are beads (fig. 12, ¢). The shell wall is pierced by a longitudinal cut that is smooth and rounded. A Riy. Bas. Sur. Pap. No. 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH Ficure 12.—Shell beads and disks 50 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLE 2.—Seven species of fossil shells examined Number | Number Genus and species of modified examples Lioplax nebrascensis (Meek and Hayden) __-_---------- 6 3 Viviparus retusus (Meek and Hayden)-_-..--.---_------ 2 2 Viviparus leidyt (Meek and Hayden)-_--._-----_-------- 1 1 “Goniobasis”’ nebrascensis (Meek and Hayden)-._-______- 4 2 Oxytrema insculpta (Meek and Hayden)-__-_--.-------- 5 2 Tancredia americana (Meek and Hayden)-_----___-_----- 3 0 Nucula planimarginata (Meek and Worthen) _--______-- 1 0 BEG Ter ne ba ee ee es aS = 22 10 thong or cord that had passed through the hole probably accounted for the polished edges on these and the remaining beads. Both of the V. retusus shells (fig. 12, f), as well as the example of V. leidyi (fig. 12, g), are treated in a similar manner. Circular perforations are drilled in the walls of the two shells of “Gondobasis” (fig. 12, h), and the upper parts of the O. insculpta shells (fig. 12, 7) formed beads after having been cut from the rest of the shell and smoothed. VEGETAL REMAINS Charred corncobs and a few seeds were found in Feature 15 and Feature 46. Identifiable remains consist of two charred corncob fragments (Zea mays), five charred corn kernels, and pieces of a fragmentary pumpkin or squash seed. Charcoal was present in large quantities in most of the pits, and several pounds of dry, compact material were saved for carbon-14 age determination. UNMODIFIED BONE AND SHELL All identifiable animal bones and mollusk shells were saved from the houses and features at the site and are listed in table 3. The mollusks were identified by Dr. J. P. E. Morrison of the U.S. National Museum; the bird bones by Mr. Loye Miller, University of California at Davis; the bison bone by Wood; and the remaining bones by Dr. J. Arnold Shotwell, Museum of Natural History, University of Oregon. The bison bone from Paul Brave was analyzed in a separate publication (Wood, 1962). The pelecypods, or fresh water mollusks, were probably taken from the bed of Four Mile Creek (see map 1). It is unlikely that any of them were living in the Missouri River, since they are species which like clear, slow prairie streams. They live in mud or sandy mud in deep holes and are not particularly tolerant of alkali water. Biv. Bas. 33], PAUL BRAVE SITE—WOOD AND WOOLWORTH 51 TABLE 3.—Species identified at Paul Brave Number Class Genus and species of indi- viduals Mollusks: GHBITODOGS ==. --- o-oo toca (Several fossil species; see table 2)_-------------------- 22 Poelecypods:..2..£226 2 to. Anodonta grandis plana Lea_..-.-.-----------------.-- 1 Lasmigona complanata (Barnes) .---------------------- 11 Lampsilis siliquoidea (Barnes)-_------------------------ 33 Lampsilis cardiwm Rafinesque------------------------- 1 Chordates: Uighe see. Sake Se 202 rea st Dee Catfish; probably Ictalurusi-v..-. 222-2 222-5223 -2t ee 1x Reptiles. 2.5... ses. 2 te Tartip (mdentined) seas. nos ea cae ee Birdsiof prey. s.2- ee eee Golden eagle (Aquila chrysaetos)_._..--.--------------- Marsh hawk (Circus cyanews).....-------------~------ Hawk (Buteo sp.) 22 2ecs 25s ss EE Wading and swimming birds------ Whistling swan (Cygnus columbianus) ----.------------ Crane (Gras canadensis) 222.2 22a ee Goose (Branta canadensis) -..=.-.=2---=-5--=2+-<==--=t= Ducks (size of teal and gadwall)--.-------------------- Obhor Dirds.— 2. tae Heath hen (Tympanuchus cupido)--------------------- Crow (Corvus brachyrhynchos) ------------------------- WATMIVOlCS = ose cas aes Dog, wolf, or coyote (Canis sp.) -------....--..--=-=-.- Gray fox (Urocyon cinereorgentus) --------------------- Artiodactys:..< 2. .<-22s2525--<-~2 Bison) CBison bis0n) 7-222. =e eo ee Deer or antelope (Odocoileus or Antilocapra sp.)------- Ik: (Cergue canadensis). =...--=-=--=.3--=---=~------- Modents/ sect ss ss2h8 besa Rabbit (Lepus and Sylvilagus sp.)--.----------- < Ground squirrel (Citellus sp.) .---------------- c Beaver (Castor canadensis) ---.---------------- 2 Muskrat (Ondatra (2ibethicus)..----------------------- Skunk (Mephitis hudsonica) --.------------------------ ENO WIE HIP WW OOH 1 Present but not counted. DISCUSSION STRUCTURES The long rectangular houses at Paul Brave are not consistent in size, but they are similar in form. Lengths are 37.5 to 46.0 feet, and widths are 24 to 82.5 feet. The long walls of the houses are not par- allel, and there is a variation in the end widths of 1.5 to 3.5 feet. This variation in width is present also at the Thomas Riggs site in houses 3 and 5 (Hurt, 1953, figs. 9, 11), and is reported in houses at the Huff site (32M011) according to Will and Hecker (1944, pp. 19-20). Houses 5 and 6 at the Thomas Riggs site (Hurt, 1953, pp. 7-8, figs. 11, 12) are similar in floor plan to House 2 at Paul Brave. There is a large post in the end of these houses opposite the entrance, and a large post centered on the house midline. These posts probably supported a ridge pole along the house midline. Houses 1 and 3 at Paul Brave (and apparently House 4 as well) have two intermediate rows of posts between the house midline and the walls. This feature is lacking at Thomas Riggs, but there seems to be evidence for a sim- ilar structural] pattern in the Over Focus Swanson site (39BR16) (Hurt, 1951, figs. 12-13, 16). In each of the houses at Thomas Riggs, a low bench of undisturbed native earth projects into the house floor between the two posts that mark the entrance. These ramps also occur at Paul Brave, but here they are small, not exceeding a foot in length, and might more appro- priately be termed “steps.” 661-932—64——_5 52 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Details of the house superstructure are rare. In House 3, timbers along one wall indicate that the wall posts were at least 5 feet high. There was no evidence of leaners on the bench of earth outside the house wall. These facts, together with the presence of small branches and twigs along the house walls, suggest that the walls were interlaced with branches. Some form of a wattle-and-daub wall may be rep- resented in this architectural form rather than the earth-covered lodge of later, historic tribes. Each of the houses had a line of center posts which apparently supported a ridge pole. In House 2, the roof may have been A-shaped or gabled, but in houses 1 and 3 the roof form is complicated by the presence of the intermediate row of posts. Bell-shaped pits predominate in the houses, but basin-shaped pits are also frequent. The bell-shaped pits in the house floors are not par- ticularly large, nor are they deep. A maximum depth of about 2 feet prevails. Exterior cache pits are larger; some of them are 6 feet in diameter and attain a depth of 5 feet below the present surface. Orig- inally they were probably no more than 4 feet deep. The additional depth is due to the soil accumulation over the site since it was aban- doned. All the pits contained refuse, and even the bell-shaped pits were used for rubbish disposal after their primary function as food storage pits was fulfilled. A long shallow trench in House 3, F70, may have an analogy in F67 in House 2, although this latter pit is irregularly shaped. These pits are similar to features in houses at Thomas Riggs (Hurt, 1953, p. 8, figs. 7-8, 10, 12). The pits at Thomas Riggs contained very little refuse, and some of them were lined with wooden slabs. . Although the pits at Paul Brave contained no wood, their refuse content was low. These pits are also similar in form to a pit in the midline of the long rectangular house at 32ME59, the site of “Grandmother’s Lodge,” al- though here the pit contained small stones (Woolworth, 1956, pl. 2). ARTIFACT COMPLEX The Thomas Riggs site, in central South Dakota, is the only ade- quately excavated site that compares closely with Paul Brave, although limited comparisons are possible with “Grandmother’s Lodge.” 'The pottery and other artifacts from Paul Brave refer to the “Archaic Mandan” period, the earliest known village culture on the Missouri River in present-day North Dakota. Comparisons with other sites are possible, but are rather restricted because of a lack of data. Ex- cavation in long rectangular house sites has been restricted primarily to testing in North Dakota, and space does not permit a detailed com- parison of the artifacts from Paul Brave with the numerous but lim- ited samples from sites listed by Will and Hecker (1944, pp. 118-121) as “Archaic Mandan,” An inspection of the various collections in the Riv. Bas. Sur. Pap. No. 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH 53 State Historical Society of North Dakota Museum indicates that Paul Brave resembles these sites in most particulars. The pottery from Paul Brave is remarkably uniform with respect to paste, surface finish, and form. Briefly, the pottery was probably made by means of lump modeling, with the use of a grooved paddle and an anvil. The use of a cord-wrapped paddle was less common. Decomposed or calcined granite was added as temper, and the pottery was fired to a hardness of 3.0 to 3.5, with a resulting color that ranged from light buff through black. The upper parts of most vessels were horizontally smoothed. Although some shoulders were vertically simple-stamped, the marks are usually obliterated. Incised decora- tions were applied to a smoothed shoulder area. Both decorated and undecorated rims were smoothed. The bases of the vessels were simple-stamped in a random fashion. A few vessels were polished. Vessels were globular, with rounded bases, and wide, apparently round mouths. Loop handles and a few strap handles were attached to rims. There are two rim forms. One of these is straight and vertical, with many rims outflaring. Vessels bearing such rims are herein termed the “Riggs Ware.” The other rim form is S-shaped, and the vessels with this character are herein termed “Fort Yates Ware.” The general characteristics of the Paul Brave pottery already enumerated apply for the most part to that from Thomas Riggs. Differences in the pot- tery from the two sites are found only when more detailed compari- sons are made. The rim sherd samples from both sites are nearly the same: 863 rims at Thomas Riggs, and 886 rims at Paul Brave. The body sherd sample from Thomas Riggs is more than six times that from Paul Brave. ‘The smaller sample from Paul Brave probably resulted from our practice of retaining only those sherds the size of a half-dollar or larger. Despite these quantitative differences, the percentages of types of body treatment at both sites are remarkably close. There is little more than a 5 percent difference between the two major groups of body sherds from the two sites. Cord-roughened body sherds are rare at both sites. The paste of the cord-roughened sherds at Paul Brave is identical with that of other body sherds, a circumstance that suggests that the sherds are indigenous. The rare check-stamped sherds at Paul Brave are not paralleled at Thomas Riggs, whereas the painted pottery at Thomas Riggs is absent at Paul Brave. The percentage of decorated shoulder sherds from both sites is nearly the same. The range of designs on Paul Brave pottery is lim- ited (figs. 1,2), and most of them occur on pottery at Thomas Riggs (Hurt, 1953, figs. 27-28, 830-31). The most popular patterns are com- posed of alternating elements of nested chevrons and triangular “ani- mal tracks.” Most of the designs are incised or trailed, but one ex- 54. BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 ample (pl. 6, e; fig. 1, 6) is composed of lines made by a stab-and-drag technique. The designs are almost exclusively composed of rectilinear elements. Curvilinear elements are present on only three vessels. The majority of shoulder designs are on the Riggs and Fort Yates Cross- hatched rims; only 18 rims of Riggs Plain are decorated at Paul Brave. Riggs Plain Rim comprises the bulk of the rims from both sites, averaging more than 85 percent of the total at Thomas Riggs? and nearly 70 percent at Paul Brave. Applique nodes are common on rims from both sites. On S-shaped rims, these nodes often serve as the apex for cord-impressed triangular elements (pl. 5,f). A unique example of applique work at Paul Brave is illustrated by Hewes (1949 b, pl. 6, right, s). The S-shaped rims comprise a larger percentage of the total rims at Paul Brave than they do at Thomas Riggs, and there is more varia- tion in the rim decorative elements. The presence at Thomas Riggs of the type Riggs Punctate, and an S-shaped rim with horizontally ap- plied cord-impressed lines lacking triangular or curvilinear rainbow elements may be significant. Conversely, the lack of crosshatched rims at Thomas Riggs (except for a possible single trade sherd), the absence of Riggs Wavy Rim, and the lack of rims of Example A may have equal significance. These differences in the pottery from the two sites, however, do not appear to be as significant as the overall similar- ities. Some local specialization is expected (see table 4). Site 32ME59, “Grandmother’s Lodge,” is a few miles downstream from the mouth of the Little Missouri River, in northwestern North Dakota. The pottery from this site may be classified as of types defined in this study. Example A is Fort Yates Cord Impressed Rim, and Example B is Riggs Plain Rim (Woolworth, 1956, pp. 90- 91, pl. 5, a-e). Only in the presence of the grooved ax and the large chipped stone projectile point does the site differ from the inventory at Paul Brave. The sample of 55 complete projectile points from Paul Brave is no- tably larger than the 18 from Thomas Riggs. There is no close correspondence in relative frequencies of projectile points, although most of the forms at one site occur also at the other. Most of the projectile points weigh 0.8 to 3.0 gm., are within the range of the small point tradition (Fenenga, 1953, p. 322), and may be interpreted as arrowpoints. Four points (fig. 3, 7-2) that weigh more than 4.5 g. are in the range of the large point tradition and are probably knives or dart points. The narrow knives from Paul Brave are similar to those from “Grandmother’s Lodge” (Woolworth, 1956, pl. 6, 7-7) and other sites 2The sherds of Riggs Straight Rim and Riggs Flared Rim from Thomas Riggs are herein classed together as Riggs Plain Rim. Riv. Bas. Sur. r Pap. No. 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH 55 TABLE 4.—Pottery frequencies at Thomas Riggs and Paul Brave sites Pottery type Thomas Riggs Paul Brave Number Percent Number Percent TRIPP SRP AOE en 5 ec soca te nee meas aetna 749 86.8 619 69.9 WrosssHatchegsee ts: 3 o> Morr Ree ee tes tee ef 8e8 1 22 54 6.1 Aerts RR ee Se ee ae ee ee 32 3.7 6 7 TEAS (AUG Gee he ig OS es NT SRR See a ee se edt ae ee | a en Se 26 2.9 TeeTare AD Pec ES eS ere eee oe Sees ae pee eee 32 Py A RS ESS [ae 2 AEE ee Fort Yates Cord Impressed 2____........-_..----------- 20 2.3 136 15.3 GH AIPA OO ROSS-TALCHOG stn re 2 8 See SES ee cs 30 3.4 Unnamed, cordimpressed’$= <=. 22: t he Cre Se be 16 VSB) | Fake se Feta 2S Er NURIGIASHENOG Sere fa ee | eee wee 8 a a 3 oe oe rere Sete eR ne eer Op. 70'0) (pf GAN DS LER eM a BEA oy Fs Ue Rr SR Sg A PE eee ee ee ee 10 ty Operas yi i ay Re a See ee ES RNA (2 TRE eS ae 1 mal Cobia 3h (Shu O) Ee GE A 6 2 SS Oe NE 6 ts ER Cs RECS | ee A Cee ee es 1 =p | Tope Veer Ol eID ICS | eee ee eo ee eee ee ee 10 det 3 .3 863 109.0 886 99.9 ‘Twelve ville. lack OnNGray 2. see. -- 8 - e 31 BOON sac tosceae Laan tes ae Srmido thee ste ae ees Say Lay ee eee Te ears 6, 620 64. 2 845 58.9 Bim ple-siami pede cee nt ere ee ee ee 3, 233 31.4 514 35.8 Cord-toushen edly Fe 2) 9 Saeed Fee es eee 39 .4 19 1.3 COU TEN eASy FTE TS gE VN OI 2 Sa RR ae RI ATR AOR] | ea a) | SEDER. Se 2 ai DE COTaTOG tees. - SR CE Ra pee TS ee eer Ey re Pe 374 3.6 54 3.7 Ger ECS 2225 Se Se ES eee 14 Pa i serene (ear ee “UNDA cies 1, OUR RRA Se ESE CE ae ee 10, 311 100. 03 1, 434 99.8 1‘‘Riggs Straight Rim” and “‘ Riggs Flared Rim’”’ from Thomas Riggs are grouped together here, 2 At Thomas Riggs, these sherds are designated ‘‘ Aldren Cord Impressed.”’ 3 At Thomas Riggs, these sherds are designated “‘ Fort Yates Cord Impressed.”’ _4 Several rims of this example at Thomas Riggs are designated “ Riggs Plain,” but the small sample from either site does not seem to deserve type status. along the Missouri River. The suggestion is made here that these knives were inserted into bone knife handles. Siltstone spheres were found at Paul Brave, as well as one circular sandstone concretion. The latter object may have derived from the famed Cannonball formation, which outcrops some distance to the north at the Cannonball River. Specimens similar to the siltstone spheres are reported by Wedel from sites in the vicinity of the Grand River, in South Dakota (Wedel, 1955, pp. 113-114, pl. 58, 7-7). A large block of sandstone was used as a mealing slab, and one mano, or handstone, may have been used on such a platform. In many respects the objects of worked bone at Paul Brave parallel those at Thomas Riggs, although there are some differences. In some classes of artifacts, such as scapula hoes and awls, the number of specimens varies but the relative proportions remain much the same. A point may be made concerning the serrated fleshing tools at Paul Brave. Tools of this nature occur in the Dodd and Phillips Ranch sites (89ST30 and 39ST14), in central South Dakota. Both of these sites contain trade goods (Lehmer, 1954, p. 110; figs. 33, a—c, and 49, j-k). They also occur in late Mandan sites (Will and Spinden, 1906, fig. 7 a-d). In the Central Plains, they are characteristic of early historic and historic complexes, and are rarely if ever found in pre- historic sites (Wedel, 1940, p. 316). Since Paul Brave lacks trade goods it seems that the usefulness of this tool as a late time marker is restricted to the Central Plains. 56 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The three bone disks from Paul Brave are similar in size and form to incised disks in the collections of the State Historical Society of North Dakota from the Motsiff, Slant, and Double Ditch sites. The disks from these three sites are incised in much the same manner as the bone disks used in the Mandan woman’s game of Sha-we (Libby, 1906, pp. 444-445). The specimens figured by Libby are cut from the walls of a heavy long bone and are carefully rounded. The identifi- cation of the bone as rib by Libby is erroneous. A number of bone tools are classed as “pottery modeling tools,” following the suggestion of Wheeler (1956, pp. 17-20). The bone tubes from the site may have been used as ornaments, but it is also possible that they were for medicinal use. In the collections of the State Historical Society are three bone tubes, two of which were col- lected by C. W. Hoffman on the Fort Berthold Reservation, N. Dak. They are said to have been used as emetics. These polished bird-bone tubes are 49 and 61 mm. long and 14 to 18 mm. in diameter. One of them has a hole in one side in the manner of a plume holder. A third bird-bone tube, collected by Frances Densmore prior to 1918 in north- ern Minnesota—probably among the Chippewa—is 64 mm. long and 12 mm. in diameter. This polished tube has two opposed holes in one end, and Densmore noted that it was swallowed “to be regurgitated.” The bone knife handles from Paul Brave are made from bison dorsal vertebrae spines, whereas most reported hafts are made from bison ribs. Specimens made of rib occur in many prehistoric and historic complexes in the Plains. They are known in Upper Republican (Kivett, 1949, p. 280; fig. 69, 6), and in the Mitchell and Twelve Mile Creek sites of the Over Focus (Hurt, 1952). The use of these im- plements continued into historic times, occurring in Mandan sites (Strong, 1940, p. 365), in the Sheyenne-Cheyenne site (Strong, 1940, p. 875), and in Arikara sites in the vicinity of Mobridge (Wedel, 1955, pp. 122-123, plate 61, f/-2). These specimens consist of a slot- ted rib with both ends cut square. The hafts from Paul Brave and from the Over Focus sites have a tonguelike projection on one end. In addition to the bone knife handles are the bone knives (fig. 7, ¢), which were used by historic tribes as “squash knives.” Most of these knives are rectangular in outline, but the form of one broken imple- ment (fig. 8, n) suggests that it was part of a hook-bladed knife similar to examples from the Dodd site (Lehmer, 1954, p. 68, fig. 26, m). Two perforated elk teeth from Paul Brave are identical with speci- mens found on costumes of historic Northern Plains Indians, as well as other groups to the west and south. The significance of these orna- ments has been discussed in detail elsewhere (Wood, 1957, pp. 381- 385), but it is relevant to state that they are prehistoric in the north- Riv. Bas. Sur. Pap. No. 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH 57 ern Middle Missouri area, but are lacking until the time of White contact in the Central Plains. Several antler bands from Paul Brave are made from thin strips of antler. The steps in the manufacture of these ornaments have been fully described by Steinbrueck (1906, p. 456-459). These ob- jects are known from the Double Ditch site (Will and Spinden, 1906, p- 172, pl. 36, w-z), where they are also made from antler. Similar specimens are in the collections of the State Historical Society of North Dakota from Slant Village, 32M026. Work in shell is relatively rare, the more common articles being disk beads and unperforated disks. The pierced snail shells from Paul Brave are made from local fossils, but otherwise are similar to specimens from Thomas Riggs. Shell disk beads also are shared by both sites. Circular disk beads are in the collections of the State Historical Society of North Dakota from the Badwater and North Cannonball sites. Circular disks, some of them probably game pieces, occur at Slant Village, Havens, Stanton Ferry, Hensler, and Demery. Pierced fossil snail shells at Boley, Motsiff, and Badwater include species identified at Paul Brave. Tt may be relevant to mention the grooved sandstone slabs found at Havens, Gaines Ranch, Upper Sanger, and Motsiff, that are in the Historical Society collections. These sandstone slabs have U-shaped grooves as much as 12 mm. wide, and the grooves are fre- quently so smooth that they might be classed as lightly polished. The width of the grooves approximates the diameter of most of the per- forated shell disk beads in the Historical Society collections, and it is possible that they may have functioned to smooth down rough-shaped shell beads. Such may not always have been the case, however, since disk beads occur at Paul Brave and grooved abraders of this character were not recovered. OTHER SITES Between Stanton, N. Dak., and Kenel, S. Dak., Will and Hecker (1944, pp. 118-121) list nearly fifty “Archaic Mandan” sites. Col- lections are available for a few of the sites in the vicinity of Paul Brave. In general, these collections are small, but those from the Robert Zahn and Havens sites are large enough for limited compari- sons with Paul Brave. ROBERT ZAHN SITE (32813) (MAP 1) Site 32S13, in the E14 SE14 sec. 9, T. 129 N., R. 79 W., covers about 10 acres on a level terrace and the adjoining slope of a low hill on the banks of a small stream. It is crossed by a gravel highway and much of it has been plowed. Will and Hecker (1944, p. 89) report that 58 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLE 5.—Comparison of traits at Paul Brave and Thomas Riggs sites I. Villages: II. Houses: Ill. Pottery (cf. table 4): IV. Miscellaneous objects of baked clay: V. Work in stone: Traits tone-rectanbular horses <2teud i hye ree. bein) ye yey eee Ee ns Boiltin shallows pitess si so ed oe he be ee De el Mntrance to'southwest-2 2 ete ee ee ee Ofiset fireplace:s) > 2223 sae wae a RE SS CaS alte es @enteriposts'in House midline. 23422 Perea eee Intermediate: 20ws Of POstsas= ove ee ey ee eT ee oh Closely-spaced’ wall posts:2.5=2 = U2 Sei Howseends Open. .t 3 «ee net we Es ne Bee ‘Trench iInvhoase rear e- = s e e 2 a Aee es ee Martinrvamp atentranep. fe heey 8 Ec yan Fo eee en Ts Howse wallsnot parallel=sc.ste see eee eee te IBFACAM DOSES Heese ko DE TY eel SE le Miterior, bell-shaped pitse.. 2-2 odo oo eee a Ices vb laine a sa aoe aan te) ew Oe eS eee Rises Cross-Hatched #-- 523". 2. Ss PL ee ae RUIsoS EN CISeM ss nee see eee een ee Netty en eee ee Dae eee we RispsPinche del © soe a a 8 ee al RIP SM CLATOL sree we ee tae ee ae. By emrees, SP NARS Fort: Yates Cord impressed. 592 ee Hort AWates'Crosssarchedus =. #s Ae see ie Nae St AE Mnnamed) Cordvinmpressedae tc 2. Be tw aio er Unclassified. Se ere RAT eae 2 ee ae 3 Ebb €7H 01) 0) (AE SL a OE a rohan ere Oem oe, TE tae ee Np. ta01) 0) (2) 3 Pee lags | AE Re A ROR ar LES Ss LOperieap ally 0 LANES: 2 Sa ee SOS Oe Ok: ARS ae SERN eee ee XSI Pleo ye eee 280 Oe I A ae a ee Twelve Mile BlackioniGray. 2-2-2 88022 Bmoutlied orplaines so. ore 2 3 ee eae 2 SOME) eee ne Sinnple-stam pede se ones tO ewe Ce we pe ad Cordoueiened fo sere 228 bo ae. SS hee @heck=ctamined omc sce 22 1! ara i ee ee ea imcisedioritratied!*se Sie ee sree bie Oe hel Unclassifiad S. = 5 8 ee 2 5 ee ee Se BHIpips fs) SE eee resi hs eee eee eee TE eS OEE RIDES a pees eae eae eae ne ee ee nee Re ee PTOICCTIO DOMES a es ee ae 55 NAbl1 AOXDaNG mp Ase Ls Tee se) Seer ik Seek Pa eS Siraignicshatted 2220 oo. | beyere Ph Danie SUS Fo MNGUSCTAPCES) a. suse ee ee Nes PY San te Mlaked oniunper surtace ss" #.- SE See eye eae Uppersuriace not flaked’ 2225.2 Ses 22a ke Phe ee Pe Broad bona PE, SDR A Se a ie 8 a 2 a ee 149 NE Le St OO SS aye Pe 2 ee Oma Se eee Nb ah JEG RS Sk NR A a ES Three-quarter groove. Sides grooved__- XXXKX XXXXX XXXKX XK KXKXXKX KXKXKXXKXKXXXXXX Paul Brave One an _ i bwSmom oon Of CO ss _ DoNeK Ore worn Thomas Riggs XX XX XXX KXKXKXXXKK KXKXXX KXXKXX XX KXXXXX coo XXoXX NSW on wreowovenwce wooo RNmory Riv. Bas. Sur. Pap. No. 33] PAUL BRAVE SITE—WOOD AND WOOLWORTH 59 TABLE 5.—Comparison of traits at Paul Brave and Thomas Riggs sites—Con. Traits Paul Brave Thomas Riggs V. Work in stone—Continued Stonespheressie tS ses oe A ER Bs es Boe 5 Shatt snioothersis sees bee ee eee eis 6 Grooved abraders_ ce cg ebraderse 312 ee Le 59 eads. VI. Work’ in bone: Seapila Noes) <2 82 We ee ee ae ke he a 66 Scapula hoes, notched glenoid......-........----.------------ 9 Scapula hoe, perforated blade.-...-..-.-.-.----=---------..-. 1 Scapnia hoe tracmontst.= 2 8s8h tok 2 a ae a 31 Scapula*thong)stretchers’?....--<-- 2 3. 2 | 2 0 _ oo co ol ho) ~ on SCAWORW Serrated fleshers._......-..---.----- Split metapodial scoops.___--.-.---- Horn core and frontal scoops ArWwIS mMmammalNlone DONC22 ia. 2 vee ence ea eo oe eee eee. 62 Headlot boneintacttette akc eat tei Fo teh eee 1 2 Headior DONG SDlitectoes. cae anes eno es oe eee ee 18 20 Head of bone worked down.2. 2 2226 22 iiss 12 12 HGad/ Of DONO TEMOVEd es an2cpcnee wae cece Coes easeeeoens 17 9 Sy ajliel 2) athe dh ee ee eS Are Sa Ae, ee eae ee = 14 WManimalrinawiSs= scans occ nacce erence se eon e ee eee eee 33 Split and worked & Splinter_..... SossSsecee 5 0 = a co bo _ o ISHS) a WS Se eee oe ae ei ou Se ee ene SS Pottery modeling tools.-------------------------------------- D> oo _ Sorco KFNWORD NWOCOK WOH HE WONWHORHY ROS Miscellaneous shaped objects... 26-5252 ee eae eke VII. Wee eer and teeth: Miscellaneous shaped antler 2202 22 28k Bao ees eee Perforateciclic teeth ce. oe ne ee oe ee ek Grooved! biSOMINCISON oc senees fee twee see eee ee seca eo senneces VIII. Work in shell: Disk beads--. 1 1 Disks eee es 1 Pendan x" res a an B ab Bo i=} nm o i) = log oO 1 Qu a — NOWHrA SCOoCce NO CKRKENWOCTOCSSCCCSCSONMOAW he house floors were once visible in the cut bank of the terrace and that the pottery was “Archaic Mandan.” A test pit excavated south of the gravel road by the State Historical Society field party in 1955 recovered the artifacts described below from a shallow, irregular pit. Three of the seven rim sherds are similar to the type herein described as Riggs Plain Rim; the lips of two sherds are indented. Two small rim sherds represent the type Fort Yates Cord Impressed. One rim, 30 mm. high, has a fillet on the mid-rim, the crest of which is 12 mm. below the lip. The rim profile resembles that of the type Riggs Punctate (Kleinsasser, 19538, p. 29, fig. 26, 2), but the fillet lacks the punctates that are characteristic of this type. A final rim is unclassi- fied. Seven of the 18 body sherds are simple-stamped, and 7 are smoothed. 60 BUREAU OF AMERICAN ETHNOLOGY (Bull, 189 Work in stone includes two pieces of plate or vein chalcedony with bifacially prepared edges, two faceted scoria abraders, two end scrap- ers, and the bases of two arrowpoints of form NBb1 made from Knife River flint and white chert. A bone awl made from a split long bone and an antler tine tip complete the list of specimens. ; There are similarities to Paul Brave apparent even in this small sample. Five of the seven rims are classified as of types described from Paul Brave, and the remaining artifacts are similar to examples from that site. The two variant rims, however, indicate that the Robert Zahn site has elements that are lacking at Paul Brave. Further testing and perhaps extensive excavations should be carried out here, since the site will be inundated by the Oahe Reservoir. The site may aid in determining the range of variation in the “Archaic Mandan” complex. HAVENS SITE (32EM1) (MAP 1) This site, in the SW14 NEY, sec. 3, T129N, R79W, is situated on a low, rolling terrace and covers about 15 acres. It is on the east bank of the Missouri River across from the Paul Brave site. About 40 widely spaced house depressions are visible. The surface is in native sod, with about 8 or 10 inches of soil accumulation over the former village level. No refuse middens are apparent, and there is only a thin layer of village refuse. There is no indication of a ditch, and the oval house depressions suggest that the dwellings are long-rectangular houses (Will and Hecker, 1944, p. 76). A large number of artifacts from the site are in the collections of the State Historical Society. Notes are not available for this sample, but it is likely that they were derived from the surface or from limited testing by Will and Hecker. The majority of the rim sherds may be classified as of types described from Paul Brave: PRS OSM Ta ote otc Ae taiene ang amet inen s «ndeaiye Tes MOMS AMBRE 134 RIP eS CnOss= Hat Gued tere a cnet ee ee eee 3 i SSh ee INGISeG Bk se cn eA eS eee ee eee 2 IRIS ER PINCH EC sss sae elise ayy: Je ca eh ee be es 10 WorteVates COL. lin pResse die 22a era re eaten tee ee eens aoe af TONGIVATES OTORS-Liatcned 2 eke es his ed Eee ee oF WGA SCM HIE Ly Pe NCTE ONSET ree geet Pye ck ge ie ee eee a Ty C27) ERA Ri, GUN SE ERMNENY “INR Me PO oe Cs Nee DS 162 Body sherds are classed as follows: Nimple-stamped 27-4 ssid share Awe ek ITE SI) aed dy Seti 10 Smoot hed {72-24} TROT Rare: BET PINS CE ee 49 @ord-roughened s2ce! . 6 BO hee ea be PGT 1 Ineiscdy Orsenaibed £2325 2-2 bie we a ee 13 Riv. Bas. Sur. sel 5 Lap. No 33], PAUL BRAVE SITE—WOOD AND WOOLWORTH 61 Stonework includes several flake knives, a number of end scrapers, and an arrowpoint of form NAb2, measuring 33 X 15 X 3 mm., made from Knife River flint. Worked bone includes two long bone splinter awl fragments and part of a split rib pottery modeling tool. An irregular shell disk is 15 mm. in diameter and 5 mm. thick. ] ‘ " , nee AY oe j Fy Aaiats vis Bons ae | ‘ ¥ fora’. wh \ *¢ . 4 inn > bendy F4 ’ anit ie vile itis cial (TX. wilyetnee tl AeA We We PS &F «at ao . 2 wy Bethe x ft afoatmsgil Hnw Wists % aly teak isin ila, et ona ak Pitan’ u! Sy, a hah .lvoeder , ny 2 rs 4 pe ‘ q Gasggaecn’ ‘ 3 i TEU (+ . ry RI a stk ds a igs phate: WEL, x ave P 204 ; 4%4 PAG ANG et HP i it» To NR ANB} 6 Bagh ofa) ttasen! Laeies lon dyety, xf BIOyT 7 =. foe laifs 5 1 seer. See PLATES BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE a, Aerial view of the Paul Brave site and environs. 6, House 2. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE a, Feature 70 in House 3. 6, Burned timbers in House 3. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 3 n l Pottery rim sherds. a, Fort Yates Cord Impressed strap handle; b, d—h, Riggs Plain Rim: c, Riggs Incised Rim; 7, Riggs Pinched, Rim. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 4 Cross-hatched rim sherds. a—b, f, Riggs Cross-Hatched Rim; c-e, Fort Yates Cross-Hatched Rim. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 5 Pottery examples A-D, and cord-impressed rim sherds. a, Example A; b, Example B; c, Example C; d, Example D; e-j, Fort Yates Cord-Impressed Rim. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 6 OQ. <, cm h Pottery disks, body sherds, and vessels. a—b, Pottery disks; c, cord-roughened body sherd; d, check-stamped body sherd; e, Riggs Plain Rim miniature vessel; f-g, plain miniature vessels; h, Fort Yates Cross-Hatched Rim vessel. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 189 River Basin Surveys Papers, No. 34 The Demery Site (39CO1), Oahe Reservoir Area, South Dakota By ALAN R. WOOLWORTH and W. RAYMOND WOOD 661-932— 646 67 ‘ y i a ‘ 0 L ( i : A. \ ‘4 W ut ) Ci ue ymin See atta tias ime ed I sie 4h reget hae 0p Aa tel sapegcaens — SrA Cana arin Mai ag t patege a hg npn LIA mek st ite om tee ee ke 9 cpm rendinaeietery fare aig & chron yet enae Mies. Aa WW) leaped 78 Li STC QE Sy 2) cape gS ee rt Sf ee AN 85 HEAT UEESTO DSCL VEO ie tge ess yey gees ey ete ee eae en eens, 86 San el UIEEY Gf HEA te ee Oi eee a se apes ch Sp fa hep hte etn Bae ht ict 88 LCi Ha) Aon alata ta pregame me nays ow te S ohh foe nD hal .- imams herpepnelneethnnne skayte lene 88 (Goeneralicharacveritshics oe oer ee ee re et eee ae ees 90 E910 gy SLL 6 a man am ed a ean aye Tag te ape aes pa a = en ae lg 9 1 a 90 Deseriptive categories and ty pes=—~—.. = ts Moe eee eee 92 CSIR ROOT ee Nee cn ere oe Se ee ee 105 GIGUTTeS CONG eset ee ee ees eee. See nee Lew. Lee eee 113 BOS CTEL UL CUS een te an ee Oe ee ee ee ee ee Se eee re eee 116 AWTDICTS ATTILCUSS 2 ashe a see Fe ee ES ee rk ey Oe wiAge OY ee eaenye 123 Shelivarhiftacis eee aes ete Ree Ah rahe ERS ERE he ey 4 by Ae cee Ss © 124 Périshable remianrsesevet 3) IBRD 2) ORS UPN O00 tO Be eAss 124 Patinalreniams. 9Ues0 _ cuViey it pg) _ _ ROG ela) SPOR oy) _ rh, 126 Piscussion==<=- SS ees Se eer fei se eset eee Ee 126 StTUcuuresieat. te - Fe CMM, EAN LA. eh ad a ad STN Se. Ee 126 Pottery sate yeh hy. Pate che ne ety ery es as pA ei 128 Othenmartifacts:. wens aaSkh atte el ee Ben nit ep eat a tk he eel ae 132 GONCHISIONS os deh ar at pe ee SS ate 134 WHIGCTAUUTE CLC Mia =. aise Sere ee etre aii oa ae ae Se 135 TABLES 1. Features recorded during the excavation of Demery site, 39CO1______ 86 2. Probable number of vessels represented by the rim sherd sample_____~_ 105 ILLUSTRATIONS PLATES (All plates follow p. 138) . @, House 2. The view is southwest; excavation is beginning in House 8 in background. 06, House 4. The view is southwest, toward the entrance. . ad, The entrance passage of House 1; the view is southeast. 6, Collapsed wall poles on the north wall of House 4. . Pottery rim sherds and handles. . Horizontally incised rim sherds. Note the stab-and-drag treatment to the left of the handle in j. . Cord-impressed rim sherds. . Incised and plain rim ‘sherds. . Miscellaneous rim sherds. . Restored pottery vessels. a, Vessel 1, House 2. b, Vessel from Feature 93, House 4. 69 70 13. 14. 15. 16. ta. 18. 19. 20. 21. 22. 24, ooanon 10. btw 12. 13. ILLUSTRATIONS TEXT FIGURES PAGE Pottery, rim: profiles (A-X 22 b eo ea a ee 91 Pottery rim. and shoulder design elements__————--====_3--= ee 93 Pottery rim and shoulder design elements___---________________-_____ 94 Pottery shoulder design elements! + ACV)... 95 Projectile points: << 2552 ee ee a eee ee 107 Chipped Stone Knives... =. 222 02-2 Le es et eee oe ee 108 Chipped and ground stone artifacts. a—b, Drills. c—h, End scrapers. i-j, Game pieces. k, Catlinite disk-bowl pipe_____________-______ 109 Basketry, woodwork, and bone artifacts. a, Charred, twilled basketry fragment. 06, Slotted bone knife handle. c, Charred, slotted wooden knife handle. d-e, Bison ulna tools. f, Bison radius pick. g, Bison Scapula woes. sos si he4 2S" an Dae See et ae eee eee 111 Ground stone artifacts. a,d, Arrow shaft smoothers. 6, Catlinite pipe bowl. c, Sandstone disk. e, Diorite celt. f, Grooved abrad- bit aes] 20) 6(: eee ee ee a eee a ee 114 Artifacts of bone, antler, and shell. a, Scored bison rib. 0b, Serrated rib tip. ec, Arrow shaft wrench. d-e, Shell scrapers. f, Antler cylinder. g, Scapula “cleaver.” h, Bone knife_____________-_____ ibicg . Artifacts of antler and bone. a-—b, Antler bracelets. c, Miniature bone awl. d-e, Fishhooks. f, Split rib awl. g-h, “Rib-edge” awls. i, Bone tube. j, Pottery modeling tool. k, Bird bone awl_-----~__ 119 Artifacts of bone and shell. a, Pendant. 0, Modified shell. c, Shell “face.” d, Shell disk bead. e, Cancellous tissue abrader. f-j, MESO TLO SWS yca e e ea re s O 120 MAPS . Location of the Demery site and other sites discussed_______-____--_ q2 . The Demery site (39CO1), showing excavations________--__________ 76 b Hxcavations: 5>7;and 8=-222--.-3-= << soon sen eee 17 . Upper: Feature 1, excavation 2. Lower: House 5, excavation 5 (facing) ee ee eee 78 IGUSE =I excavation as. == ene ee ee 79 JTOURE.2, (OXCAVALION O-——— oe ee ae ee 81 EPOUSG a /EXCAVATION 0.623 22 oo ee ee 83 House 4 exeavation 02025226 ee 84 THE DEMERY SITE (39001), OAHE RESERVOIR AREA, SOUTH DAKOTA? By Auan R. Wootwortn and W. Raymonp Woop INTRODUCTION In the summer of 1956 an archeological field party from the State Historical Society of North Dakota carried out excavations at the Demery site, in the upper part of the Oahe Reservoir, in Corson County, South Dakota. Funds for the project were provided under a cooperative agreement with the National Park Service, Department of the Interior, and through appropriations by the North Dakota State Legislature. The excavations were conducted between June 18 and August 31, 1956, under the supervision of Alan R. Woolworth and W. Raymond Wood, then staff archeologists with the State His- torical Society, with Frederic Hadleigh-West serving as assistant archeologist. The major reason for selecting Demery for excavation in 1956 was that the site is the northernmost manifestation of the “Category B” complex on the mainstem of the Missouri River. In 1949, Paul L. Cooper had described the Category B pottery complex from sites along the Missouri River in South Dakota (1949, pp. 303-306). The Category B rim design consists of a solid band of horizontally incised or trailed lines on rim exteriors. Cooper demonstrated the wide spatial distribution of this form of pottery rim decoration, and pre- dicted that the delimitation of the temporal and spatial scope of this archeological complex would provide one of the more important chapters in Plains prehistory. The Demery site is only one of the many sites bearing pottery of this character in the Central Plains and Middle Missouri areas. Because of its extreme northern geographic setting, it was anticipated that Demery would provide data on a peri- pheral variant of the complex as it was known from work in sites in central and southeastern South Dakota. (See map 6.) Much of the credit for the successful completion of the work at Demery is due Russell Reid, Superintendent of the State Historical 1 Submitted February 1962. 71 72 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 © Swan Creek ane = eney® Myers”) Spotted Beor Cl Arzberger 2 a. Totking Crow Lo Roche © Scalp Croek Minorik NJ Redbird Gray-Wolfe Burkett ° Location of the Demery Site and other sites discussed Map 6.—Location of the Demery site and other sites discussed. Society of North Dakota, for his unfailing assistance in every phase of the work. The field party was established in the town of Fort Yates, N. Dak. Mr. J. W. Wellington, then Superintendent. of the Standing Rock Indian Agency, in Fort Yates, did everything in his power to make the summer a pleasant and successful one. The Missouri Basin Project, Smithsonian Institution, loaned photographic equipment and was, as always, helpful in many other matters. Ber- nard Weinreich, of Bismarck, N. Dak., prepared the photo- Riv. Bas. Sur. Pap. No'3i],. DEMERY SITE—WOOLWORTH AND WOOD 73 graphic illustrations; their quality is a tribute to his skill. Frederic Hadleigh-West was in charge of mapping the site and the excava- tions, a task of no mean proportions considering the distances in- volved and the equipment available. Permission to excavate on tribal land was kindly granted by the Tribal Council of the Standing Rock Indian Reservation. Mr. J. Dan Howard, then Chairman of the Tribal Council, aided the field party in many ways. Crew members included Robert P. Barr and Stephen W. Robinson of Grand Forks, North Dakota; George E. Archambault and William C. Gipp of Fort Yates, N. Dak.; Mark F. Blum of Lincoln, Nebr.; and Craig Gannon of Bismarck, N. Dak. The assistance of each of these contributors is gratefully acknowledged. We are also grateful for the comments and constructive criticism of the manuscript provided by Robert L. Stephenson, Warren W. Caldwell, and Robert W. Neu- man, of the Smithsonian Institution, Missouri Basin Project. The laboratory analysis of the pottery was carried out by the junior author, who also prepared the final maps and the line drawings. The remaining artifacts were processed by the senior author, and both of us are jointly responsible for the discussion and conclusions. The field notes, maps, photographs, and artifacts from Demery are on file in the museum of the State Historical Society of North Dakota, in Bismarck, for preservation and further study. ARCHEOLOGY OF THE SITE DESCRIPTION The Demery site, 89CO1, is in the SE% of sec. 21, and in the SW, of sec. 22, T. 23 N., R. 29 W., Corson County, S. Dak. It is on the west bank of the Missouri River about 16 miles downstream from the town of Fort Yates. The site lies on a large, flat. terrace overlooking the floodplains of the Missouri River and John Grass Creek (map 7). A large gully separates the site area from the ter- race to the south, and a low swale to the west leaves the site nearly surrounded by low land. To the west, the terrain is level for nearly half a mile, where the Missouri River bluffs rise gradually from the valley floor to the treeless plains. A small intermittent stream, locally known as Black Eagle Creek or as John Grass Creek, is just north of the site. The reputed cabin site of the Dakota chief, John Grass, is a few hundred feet upstream from the bridge spanning the creek bear- ing his name on the Kenel to Fort Yates road. The North Dakota- South Dakota boundary is just north of John Grass Creek. Fire Heart Butte, a high, flat-topped prominence, rises above the surround- ing upland plains about 5 miles northwest of the site. The 1,610-foot contour line roughly delimits the area containing cul- tural remains, but the majority of surface material and most of the 74 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 excavations were carried out within an area covered by heavy vege- tation (map 7). This area, consisting of about 2 acres, appears to have been the major village area. About 10 percent of the area was stripped of overburden by a bulldozer and a road patrol. Five houses were located by this method and subsequently excavated, and two other probable house sites were located. A speculative projection, based on the distribution of these houses, gives a total of about 30 dwellings in the village. Demery appears to have been an open, unfortified site. A test trench dug from the east wall of House 4 in excavation 5 to the ter- race edge revealed no evidence of fortification, and a close inspection of the terrain and of aerial photographs revealed nothing suggesting the presence of a fortifying ditch. In an effort to find burials, test pits were dug along the terrace edge south of excavation 7, but no positive evidence of either burials or occupational debris was found. Most of the area bearing heavy vegetation had been cultivated, which obliterated indications of surface features. Only a few acres of grassland near the creek are still in undisturbed sod, and somewhere in this area Will and Hecker (1944, p. 87) reported a number of cir- cular lodge ruins. Early in the field season a single circular depres- sion was noted here, but excavation 2, which explored this feature, yielded the remains of an irregular pit, Feature 1, which was inter- preted as a borrow area (map 9). A few flint chips and body sherds were found near the bridge across John Grass Creek (designated by an “x” on map 7). Surface inspec- tion of the terrace beyond the large gully south of the site, near the remains of the former Demery home, revealed no sherds or other evi- dence of aboriginal occupation, although there are a few sherds in the State Historical Society collections that are said to be from this vicin- ity. These sherds resemble those from the major occupation of the Demery site in all details, and they are probably the source for the statement by Will and Hecker (1944, p. 88) that there was a small site contemporaneous with Demery south of the gully. A number of grassy surface irregularities, as well as many clumps of dense buckbrush, occurred throughout the site area. The excava- tions revealed no correlation between either these irregularities or the brush patches and the house sites. In fact, the houses appeared to have been on level, grassy areas. Toward the end of August, how- ever, when a dense cover of sunflowers and weeds had overgrown the site, there were spots on level, grassy areas where this cover was ex- ceedingly dense, and such areas we feel were house sites. Since the field season was over we did not have the opportunity to investigate them, although it would clearly have been desirable. Riv. Bas. Sur. Pap. No. 34] +DEMERY SITE—WOOLWORTH AND WOOD 75 EXCAVATIONS Six excavation units in different parts of the site were stripped to the approximate base of the plow zone by means of a bulldozer and a road patrol. At this depth the houses, pits, and other features became clearly visible as mixed earth in otherwise sterile soil. Features were marked by stakes while the soil was still moist and their outlines dis- tinct, and later excavated by hand tools. Two other units, excavations 3 and 4, were excavated wholly by hand methods. Houses appeared as large circular areas of mixed earth which some- times contained charred timbers. Four of the houses found were fully excavated anda fifth house was cross sectioned. The presence of two additional houses, designated as Houses 6 and 7 on the site map (maps 7, 8), is postulated on the basis of charred timbers noted on the floor of the road patrol cuts, but time did not allow the investigation of these features. In view of the large areas stripped by the road pa- trol, houses were infrequent and widely separated. It is not probable that many houses were missed by the excavators, since soil disturbance was rather clearly defined in the dense, buff soil below the plow line. The site maps were prepared by West with the use of a plane table and alidade, and distances were chained. The individual houses were mapped using a stake in the center of the primary fireplace as a datum. EKecavation 1 (map 7).—This unit, in the northwest part of the site, was originally a series of test pits laid out on a 5-foot grid. Features 40, 41, 42, and 43 were recorded here, respectively : an irregular trash- filled pit, two basin-shaped pits, and a bell-shaped pit. This area was later stripped with a bulldozer and subsequently smoothed by a road patrol. Numerous postholes and fireplaces were found, but no definite dwelling units were detected. The area apparently lay on the periphery of the village. The unit was 275 feet long and about 75 feet wide. Excavation 2 (maps 7 and 9).—This excavation was designed to explore a circular depression in the northwest part of the site, which appeared from surface observations to be about 10 feet in diameter. It was heavily overgrown with weeds and buckbrush and superficially resembled an earth lodge depression. A test trench was dug from the north edge into the pit, and the bulldozer was then used to strip the overburden from an area 80 feet long and 55 feet wide. Feature 1, a large irregular pit nearly 20 feet in diameter, is interpreted as a borrow pit. The floor of the feature was littered with broken animal bones, pottery, and stone fragments. Features 48 and 49, both of them bell-shaped pits, were outside of this feature. EHecavation 3 (map 7).—This excavation was made in the north part of the village area, and consisted of a test trench 28 feet long and 5 feet wide. A single basin-shaped pit, Feature 50, was re- [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 76 *sUOIJBAROXS SUIMOYS “([QDGE) e318 AtoWAaq xy, —'Z AVIV ayig Asaweg ayy lOD6E ulojdpoo; 4 s ‘US Voor eben Aacoy Ni 40 sfw)) aan) HORy JOALY lunossiw Riy. Bas. Sur. Riv. Ras. Sur. DEMERY SITE—WOOLWORTH AND WOOD 77 ‘House 6 se? Cie 3 F59 e Fug F75° Fope ode | X5, 7 and 8 O Excaveted pit % Unexcavated pit @ Unexcavated fireplace if } Probable earth lodge ! i Lal Map 8.—Excavations 5, 7, and 8. 78 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 corded, and the miniature shell “face” (fig. 24, c) was recovered near the south end of the unit. Excavation 4 (map 7)—A low earthen mound, about 50 feet in diameter and 1.5 feet high, was along the terrace edge north of the village area. ay + Aik wa OTP 4 | » tH al EOI LZ. I De ig) ee be Riv. Bas. Sur. aes Bap. No 34], 4 DEMERY SITE—WOOLWORTH AND WOOD 79 o KX? L\ * 39COl Y House I, X7 ke # 9 ee Pit Bojl-shoped pit ne Fireplace Posjhole Probable poslilon of roof beams Map 10.—House 1, excavation 7. There were three basin-shaped fireplaces in the shallow, dish-shaped house floor. Feature 4 was centrally located and was probably the hearth for the original lodge; it was 3 feet wide and 0.6 foot deep. Feature 5 was offset to the southwest from the midpoint of Feature 4, and was the same size as Feature 4. Feature 6 was a small auxiliary fireplace between Feature 5 and the entrance. The four large center posts around Feature 4 form a square 8.5 feet on aside. Another set of postholes, around Feature 5, may be the re- sult of rebuilding the house and setting the posts nearer the entrance. 80 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 These latter posts form an irregular square 7 to 8 feet on a side. In this instance, however, the corners of the square are formed by three postholes rather than by a single post. These posts were much larger and deeper than any of the remaining auxiliary posts in the floor... Center posts of both sets were 0.6 to 1.2 feet in diameter, and 1.5 to 2.0 feet deep. Five bell-shaped pits were in the house floor. One of them, Fea- ture 10, partly intersects Feature 11, another bell-shaped pit. Fea- tures 14 and 15 are in the house wall line; the latter pit, which is bell shaped, has an orifice shaped like the figure 8... Two of the five small circular to oval pits with shallow U-shaped cross sections intersected the wall line. These and the remaining pits contained refuse and charcoal. The pit in the wall opposite the en- trance, Feature 103, may have had a special function, since it is in a position analogous to that occupied by ceremonial altars in certain historic earth lodges. House 2 (map 11, pl. 7, a2) —This house was in excavation 5 in the north central part of the site. The structure was the simplest in plan of those excavated, although its pattern is complicated by the prob- able presence of a second structure. The house consists of a roughly circular basin-shaped depression 22 feet in diameter, and about 2 feet deep in its center. The entrance faced the southwest, and was marked by two short rows of three posts each, separated by a distance of 4 feet. The posts extended only 3 feet out from the house wall, and per- haps the rest of the passage was not located ; rodent disturbance in this area made the definition of features difficult. A basin-shaped fireplace, 2.6 feet wide and 0.4 foot deep, was in the house center. Around this hearth were four large supporting posts in the form of an irregular square 8 feet on a side. These posts aver- aged 0.9 foot wide and 1.6 feet deep. Three bell-shaped pits were in the house. One of these, Feature 23A, had a smaller bell-shaped pit in its floor; Feature 21 partly un- dercut the northwest wall line; and Feature 20 was in the southeast part of the floor near the house wall, adjacent to an irregular, basin- shaped trash pit, Feature 112. Two circular, basin-shaped pits were in the house. One of them, Feature 24, was in the house wall opposite the entrance, in a position analogous to that of Feature 103 in House 1. The other pit, Feature 18, was between the two center posts facing the entrance; 1t contained several artifacts. ‘Two shattered pottery vessels were near the north wall, opposite the entrance. Both of them were smashed flat, and must have been broken by the weight of the collapsed roof. One of them contained charred fragments of animal bone. Vessel 1 was restored (pl. 14, a). Feature 22 was a fireplace in the southwest house wall line. It had a series of seven posts set about it, approximately in the form of a Riv. Bas. Sur. Pap. No 34], DEMERY SITE—WOOLWORTH AND WOOD 81 waa N b . & a / x Vessel 2 -=~ x Vessel | % o O (me? Os e / ~ 39CO! @ ia House 2, X5 ig On iO} Bell-shoped pit Storile soll —— \ Plow zona—— Fireplace Posthole Probable position 8 of roof beams Map 11.—House 2, excavation 5. square, 6 to 8 feet on a side. It appears to be a second and smaller structure overlapping with the floor of House 2, but whether it was earlier or later could not be determined. House 3 (map 12).—This structure was in excavation 5, in the north central part of the site. The house was contained in an oval depression 2 feet deep, 32 feet long, and 30 feet wide. The long axis was in line with the orientation of the entrance. The shallow entrance passage, facing the southwest, was 8 feet long, 4 feet wide, and 0.5 82 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 foot deep; it was lined by a series of small, irregularly-set post holes. Feature 61 was a large oval fireplace in the center of the house; it was 6.6 feet long and 4 feet wide. This feature was composed of two linked basin-shaped fireplaces, each of which was about 4 feet wide. and 0.8 foot deep. This fireplace was surrounded by seven large center posts braced with bone and stone. Four of these posts, in the form of a square 8 to 10 feet on a side, were probably the posts which supported the roof. Three other braced posts were in the same gen- eral area, and may have served as braces for the center posts. It is possible, however, that some of them belong to a rebuilding of the house in which the structure was shifted in one or another direction along the long axis. Braced posts are unique to this house within the site, in any event, and this may be a function of size, since House 3 was the largest of those excavated. Perhaps this larger structure needed a more stable foundation than the other houses, which was afforded by jamming bone and stone beside the posts. The center posts averaged 0.9 foot wide and 1.7 feet deep. The seven bell-shaped pits, and one pit south of the house, were of radically different sizes. Four of them were quite large, and would have provided adequate storage space. The rest of them were much smaller, and may have served another function. All of them, how- ever, contained refuse, particularly Feature 65 in the north wall of the house. This feature contained most of the perishable remains recovered at the site. There were three small auxiliary fireplaces in the house: Features 25, 46, and 47. One large, irregularly shaped refuse-filled pit, Fea- ture 79, intercepts the south wall of the house, and there are three small, basin-shaped pits in the house floor. One of the latter pits, Feature 29, was between the center posts facing the entrance in a position analogous to that of Feature 18 in House 2. House 4 (map 13; pis. 7, b, and 8,6) —This house was in the east end of excavation:5, near the terrace edge adjacent to the Missouri River flood plain. The house pit was quite irregular in outline, averaging about 24 feet in diameter. It might be described as an oval with one rather flat side. The house pit was 1.5 feet deep along the walls and attained a maximum depth of 2.8 feet near the fireplace. An elongated, depressed entrance passage was in the southwestern perim- eter of the house wall, measuring 6.5 feet long, 3.5 feet wide, and 0.4 foot deep. Feature 91, the basin-shaped central fireplace, was 3.4 feet wide and 0.6 foot deep. It was an oval pit surrounded by and filled with white ash and burned earth. Another centrally located basin-shaped fireplace, Feature 104, joined Feature 91 on the northeast edge. The latter pit was oval in outline, measuring 1.8 by 2.1 feet, and was 0.4 Riv. Bas. Sur. ae; Riv. Bas. Sur. DEMERY SITE—WOOLWORTH AND WOOD 83 ‘ ee ae LO Cee ee 27 The crown of the incisor is worn down and the pulp exposed. The pulp cavity, however, had receded with the attrition of the tooth. 86 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The fifth sacral vertebra of the sacrum and part of the fourth are lacking and, since the borders of the bone are broken, length and breadth are not obtainable. Since all the segments are fused, the individual was more than 25 or 30 years old. The cranial fragment . appears to be part of the occipital bone; one edge shows part of a complex lamboidal suture. FEATURES OBSERVED The 118 features recorded during the excavation of Demery are tabulated below in table 1. While most of the feature numbers were assigned to pits in the houses and in the village area, a number of them were used to designate houses, entrance passages, and other structural details. These features are excluded from table 1 since they are described in detail in the preceding text. Maps 7-9 give a general picture of the major areas investigated at the site, and the location of all features not in or immediately contiguous to the houses. TasLe 1.—Features recorded during the excavation of the Demery site, 89CO1 Feature | Excava-| Depth Dimensions Identification No. | tion No. Feet 1 2 2.0 | Max. length, 21.0; Borrow pit. width, 20.0. 4 7s B,,| Diamh., 3.002 -s0d4-hs Central fireplace, House 1. 5 a Bey he i Ti | ea Auxiliary central fire- place, House 1. 6 i AN) ADIN. AO 2 he AUP EY, fireplace, House 7 ri - SH Diani, 20025. iv. Basin-shaped pit, House 1. 8 7 2.0 | Orifice, 1.8; base, 3.2__-| Bell-shaped pit, House 1. 9 € 2.0 | Orifice, 1.8; base, 3.5__| Bell-shaped pit, House 1. 10 7 2.1 | Orifice, 2.8; base, 3.6__| Bell-shaped pit, House 1. 11 7 1.2 | Orifice, 2.2; base, 2.8__-| Bell-shaped pit, House 1. 12 7 1.1 | Length, 1.7; width, 1.1_| Oval pit, House 1. 13 7 S60) ADiani0.505. 20.9150 Basin-shaped pit, House 1. 14 7 10") Omi. Son oo ee Basin-shaped pit, House 1. 15 7 3.5 | Orifice, 2.0; base, 3.8__| Bell-shaped pit, House 1. 16 7 2.2 | Orifice, 1.2; base, 2.2__| Bell-shaped pit, House 1. 18 5 £37) Siam 0:9). sect e eS Circular pit, House 2. 19 5 a Diam, ee oS Pet fireplace, House 20 5 1.9 | Orifice, 2.0; base, 2.5__| Bell-shaped pit, House 2. 21 5 1.9 | Orifice, 1.8; base, 2.4__| Bell-shaped pit, House 2. 22 5 6 f- Diam. 17X26 22 222 Fireplace, House 2 (with secondary structure?). 23A 5 1.9 | Orifice, 2.4; base, 2.7__| Bell-shaped pit, House 2. 23B 5 1.3 | Orifice, 1.5; base, 2.6_-| Bell-shaped pit, in base of Feature 23A. 24 5 1, 1-} -Diam,, 4,3. 207204 oa wolbennt 4 pit, House 25 5 oh by Diba AN ae bs, See Auxiliary fireplace, House 3. a p 4.6 | Orifice, 2.4; base, 5.5--| Bell-shaped pit, House 3. 7 1.5 WII i ee ee Deep, circular pit, House i Riv. Bas. Sur. Dap. No. 34] DEMERY SITE—WOOLWORTH AND WOOD 87 TABLE 1.—Features recorded during the excavation of the Demery site, 89CO1— Continued Feature | Excava-| Depth Dimensions Identification No. | tion No. Feet 28 5 2.7 | Orifice, 2.6; base, 4.5_--| Bell-shaped pit. 29 5 122.4) Bain! PS). OSs OS Deep, circular pit, House 3. 30 7 of | Dram... b.0% 6:5...) Irregularly shaped pit. 31 7 .3 | Diam., 4.3 6.0__-.-_--| Irregularly shaped pit. 32 7 1.2 | Diam., 3.8X6.3__-_----| Irregularly shaped pit. 33 7 1.2 | Orifice, 1.3; base, 1.6__] Bell-shaped pit. 34 7 1,0.) Diant. $2.81). vat! 4. Circular pit with steep sides, House 1. 35 7 1.0 | Diam., 12.0 18.5_.___| Irregularly shaped pit, House 1. 36 71 4,9 | Orifice, 2.3; vase, 4.7--| Bell-shaped pit, House 1. 37 7 $8 -|. Diam.o ASA _,.custl Basin-shaped pit, House 1. 39 7 1.2 | Diam., 12.0 15.0_____| Irregularly shaped pit. 40 Ye Oe71. 0 -|. Diam.; eat 3.0.42 52 Irregularly shaped pit. 41 1 1} O-) Diam.) 2:9. ace. Basin-shaped pit. 42 1 .4 | Diam., 3.0 4.0.-___--| Basin-shaped pit. 43 1 3. 1 | Orifice, 2.4; base, 5.5__| Bell-shaped pit. 44 7 .7 | Diam., 12.0X14.6__--- Irregularly shaped pit. 46 5 + 1 «} Diam.,. L165 cult Auxiliary fireplace, House 3. 47 5 + 1.) Diam,;. 1:21 so .cuutlo_ Auxiliary fireplace, House 3. 48 2 3.0 | Orifice, 2.5; base, 4.5_--| Bell-shaped pit. 49 2 2.1 | Orifice, 2.6; base, 3.7__| Bell-shaped pit. 50 3 .5 | Diam., 2.3X3.0__-..--| Basin-shaped pit. 51 5 1 GO: Diam. 1501 or Ck Irregularly shaped pit, House 5. 52 5 2.9 | Orifice, 2.8; base, 3.8__-| Bell-shaped pit, House 5. 53 5 29 | Diam, 15.4 19.8__-4- Irregularly shaped pit. 54 5 3.0 | Orifice, 2.7; base, 4.0__| Bell-shaped pit, House 5. 55 5 2 5.) Diam.,' 11.1 K14:520. -- Irregularly shaped pit, House 5. 56 5 t 33)..DiantSG0UC sui. Irregularly shaped pit. 57 5 1. 0} Diam.,; 6:3'X 9.0.21: —- Irregularly shaped pit. 58 5 : 2-- Diany. 6:0 10.0cL__-- Irregularly shaped pit. 59 5 1.0 | Diam., 14.0 22.0__--- Irregularly shaped pit. 61 5 .8 | Diam. 4.0X6.6...._4- Two linked fireplaces, House 3. 62 5 1.2 | Orifice, 1.0; base, 1.4__| Bell-shaped pit, House 3. 63 5 1. 2 | Orifice, 1.2; base, 1.8_-| Bell-shaped pit, House 3. 64 5 Verh rae Gs 2 SSS & Basin-shaped pit, House 3. 65 5 3. 8 | Orifice, 2.4; base, 4.0__| Bell-shaped pit, House 3. 66 5 1. 6 | Orifice, 1.4; base, 2.0__| Bell-shaped pit, House 3. 67 5 3. 0 | Orifice, 3.2; base, 4.2__| Bell-shaped pit, House 3. 68 5 3.0 | Orifice, 2.2; base, 3.6__| Bell-shaped pit, House 3. 69 5 5 Ae | EELS SS OR PO eee a Basin-shaped pit, House 3. 71 vd 3.8 | Orifice, 1.9; base, 4:45.) Bell-shaped pit, House 1. 72 7 2.5.) Orifice, 2.8; base, 3.5_-| Bell-shaped pit. 73 a Zea | OTANI, 200 oc 0 an Oval pit, House 1. 74 § .8 | Diam., 3.8X4.0_---2-- Irregularly shaped pit, House 4. 75 5 ro 1 ise, 0-0 Xx 0... Irregularly shaped pit. 76 5 . &)| Diam; 2.4X 2.9. 22 Irregularly shaped pit, House 4. 77 5 oT Palit, o:0 © 9-0~—-— 2 Irregularly shaped pit. 78 5. 3. 3 | Orifice, 3.7; base, 3.9-.| Bell-shaped pit. 88 BUREAU OF AMERICAN ETHNOLOGY (Bull. 189 TABLE 1.—Features recorded during the excavation of the Demery site, 389CO01— Continued Feature | Excava-| Depth Dimensions Identification No. _ |tion No. Feet 79 5 1.5 | Diam., 4.8X8.0___..-- Irregularly shaped pit, House 3. 81 5 . Sts Digi 215 oa: eek Central fireplace, House 5. 82 5 1.5 | Orifice, 1.8; base, 2.4__| Bell-shaped pit, House 5. 83 5 1.5 | Orifice, 1.1; base, 1.8__| Bell-shaped pit, House 5. 91 5 $6 | Diamiy 3:42 toa pee Central fireplace, House 4. 92 5 2)0 .|. Diam.,. 1.0X3,2e4icL-- Irregular pit, House 4. 93 5 1.6 | Orifice, 1.8; base, 2.8__| Bell-shaped pit, House 4. 94 5 2.7 | Orifice, 1.6; base, 3.0__| Bell-shaped pit, House 4. 95 5 2.5 | Orifice, 2.2; base, 3.8__| Bell-shaped pit, House 4. 96 5 2.6 | Orifice, 2.8; base, 4.6__| Bell-shaped pit, House 4. 97 5 2.3 | Diam.,. 4.9% 5.82 -- Irregularly-shaped pit, House 4. 98 5 270 |. Diam s (15843. Oraif bot Irregular pit, House 4. 99 5 2? 0 |. Diam.,$£'1-5362.Qeci2t Irregular pit, House 4. 100 5 1 2.|. Diam.,. 4.0% 7.Ouici Irregular pit, House 4. 101 4 1.6 | Orifice, 1.5; base, 2.5__| Bell-shaped pit, House 1. 102 7 OF 7 |. Diamysc 168 - sess Basin-shaped pit, House 1. 103 7 1:0 | Diam., 1.0% 1.9_2-L-_ Basin-shaped pit, House 1. 104 5 .4 1 Diam,,. 2i02-_ereilk Auxiliary central fire- place, House 4. 105 5 .. §_|_ Diam.,. . 18! —-areicl 4- Basin-shaped pit, House 4. 106 5 3. 0 | Orifice, 2.2; base, 4.4__| Bell-shaped pit, House 4. 107 5 18 | Diiamesd 1238 as8iw2 4 Basin-shaped pit, House 4. 108 5 14.) Dilamg | 292° ef as) - Basin-shaped pit, House 4. 109 5 PS.) Diante SGoe- penict o1.. Basin-shaped pit, House 4. 110 5 D4,\.Diam,, Vile s_ormitt 4 Basin-shaped pit, House 4. 111 5 |Ca. 3.0 | Orifice, 2.2; base, 4.0__| Bell-shaped pit. 112 5 . ¢.| Diam; 5: S28 7. Bilal) 4. Trregularly shaped pit, House 2. 113 5 2.2 | Orifice, 2.0; base, 4.4__| Bell-shaped pit, House 3. 114 5 .6 | Diam., 1.4X1.8...___-| Basin-shaped pit, House 2. 115 65 1.0 | Diam., 1.21.6_..____| Basin-shaped pit, House 2. 116 5 .8 | Diam., 2.2 2.5_..____| Basin-shaped pit, House 2. 117 5 1.3 | Orifice, 1.0; base, 1.4__} Bell-shaped pit, House 3. 118 5 1.2 | Diam., 5.26.0_._____| Irregularly shaped ‘pit, House 4. ARTIFACTS POTTERY The pottery from Demery is abundant and, despite-an overall uni- formity in. paste and surface finish, is quite varied in the form of the rim and the type of decoration applied to the outer rim. The heter- ogeneity of the rim forms and decorative elements suggested, prior to excavation, the possibility that the site may have had.more than one occupation, and this possibility was one of the reasons the site was chosen for excavation. During the fieldwork, however, and through the laboratory analysis of the materials it, became increasingly appar- ent that only one major occupation was present.’. A second and minor Riv. Bas. Sur. Pap. No. 34] DEMERY SITE—WOOLWORTH AND WOOD 89 occupation by a Thomas Riggs Focus group is suggested by a very few sherds. The analysis of the pottery proceeded as follows. The rim and body sherds, sacked together from the same find-spot, were first matched for fits, and subsequently segregated. Body sherds were classed according to surface finish, these classes including: smoothed, simple-stamped, cord-roughened, check-stamped, and _ decorated (incised or trailed). Red-filmed sherds were also separated, but this filming occurred on three surfaces: smoothed, simple-stamped, and decorated. During the sorting, each sherd was individually inspected for temper. A single cord-roughened sherd which appears to con- tain crushed shell provides the only exception to the statement that all sherds are grit-tempered. In the descriptions to follow, the unit of description is, insofar as possible, the whole vessel. Since the rim forms and the types of decoration are so varied, it is probable that the estimated number of vessels represented by each of the groups is reasonably accurate. In addition to the reasons cited by Krieger (Newell and Krieger, 1949, pp. 75-77) and by Spaulding (1956, pp. 180-131) for studying “vessel types” assembled from “sherd types,” is the fact that there is often a discernible—and sometimes substantial—difference in the percent- age of types represented in a count of rim sherds and a count of vessels from a site. In spite of efforts to describe vessel types here, the pottery described below should probably be regarded as rim types, since so few restorable vessels were recovered. Only a few of the 17 pottery groups from the site are identified as previously named types. The unnamed groups might have been named as types by some workers, and perhaps tentative type names could have been given them here. The permanent nature of “tempo- rary” type names, coupled with the fact that the complex at Demery is involved and geographically diffuse, suggests that type names would not be advantageous. At this time, when many sites closely related to Demery remain unexcavated and unreported—from the mouth of the Grand River in South Dakota to the mouth of the Nio- brara River in Nebraska—it does not seem proper to attach type names to the ceramics from a single site, particularly when the range of variation in pottery at sites down the river is unknown, at least to the writers. Except for the sherds attributed to the Thomas Riggs Focus occupa- tion of Demery, the pottery is sufficiently similar in paste, surface finish, and form to permit a general description of these features that applies to all the ceramics from the major occupation of the site. 90 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 GENERAL CHARACTERISTICS PASTE: Method of manufacture: The vessels were probably lump modeled and fin- ished with a paddle and anvil. The vessel walls are quite uniform in thickness from the lip to the bottom, but the pots do not appear to have been scraped, since traces of paddle impressions are visible over most of the vessel surface. Temper: Grit, composed of calcined or decomposed granite, consisting of particles of quartz, mica, and feldspar, ranged in diameter from .5 to 2.0 mm. Temper is nearly always sparse, and many sherds—particularly those from the smaller vessels—appear to have contained no purposefully included grit. Texture: Smooth to medium fine. The paste is compact and well worked, although many vessels have a tendency to split parallel to the vessel wall. Hardness: 2.5 to 3.0, the majority about 3 (calcite). Color: Light buff to black, ranging through several shades of brown and gray. Characteristically, the sherds are a mottled light buff or gray. Smoke clouding is common, and black, charred organic matter sometimes adheres to the interior or exterior surfaces. SURFACE FINISH: The entire vessel seems to have been first paddled with a grooved or, more rarely, a cord-wrapped paddle. On the large restored vessel (pl. 14, 6), simple stamps extend from the neck to the base; a number of vessels have vertical or oblique stamps on the outer rim. Shoulders are ver- tically stamped, and an incised or trailed pattern was sometimes applied to them over a partly smoothed surface. In nearly all cases the stamps are at least partly smoothed ; in about 5 percent of the vessels they are unmodified. A few bases are lightly polished, but surfaces are generally dull, with little tendency to reflect light. FORM: Lip: Flat to rounded; a few are pointed and a small number are extruded (fig. 18). Rim/neck: Characteristics of these areas vary with the individual groups and types (fig. 13). Orifice: The vessel mouth is wide and round in all observable instances. Shoulder: This area is rounded and steeply sloping, joining the body in a smooth curve. On a few small to medium-sized vessels the shoulders are sharply angular at the juncture with the body (pl. 10, 7). Appendages: The loop and strap handles are riveted to the upper shoulder and welded to the lip. Handles on large restored vessel segments are paired on opposite sides of the rim. A number of rim projections (castel- lations) and some applique nodes occur on some groups. Appendages of this character occur in fours, equally spaced around the rim. BODY SHERDS The majority of the sherds are either simple-stamped or smoothed. Most of them show some smoothing although few may be classed as even lightly polished. Every degree of transition from unmodified stamps through partly smoothed stamps to lightly polished surfaces was observed, sometimes on a single large sherd. All sherds that showed any evidence of stamping were classed as simple-stamped. Sherds were classed as smooth when the surface was so smooth that Riy. Bas. Sur. Pap. No 34] DEMERY SITE—-WOOLWORTH AND WOOD 2») russ Le) 5 cm. (S954 Ficure 13,—Pottery rim profiles A-X. irregularities were lacking. A third class was composed of sherds with incised or trailed lines executed on simple-stamped or smoothed sherds. The 1498 sherds with distinguishable simple stanvps show that the stamps were applied vertically to the shoulders and rims, but near the base they are random or oblique. These stamps are 5 to 8 mm. wide and are of varying length, but in no instance do they exceed 30 mm. Vessel walls are 1.5 to 6 mm. thick, averaging about 3 mm., although rims are sometimes thicker. There is little variation in the thickness 92 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 of the vessel wall in most of the restored vessels and large sherds. For example, the thickness of the base was the same as that over most of the body on the largest restored vessel (pl. 14, 6) ; the neck and rim were only slightly thicker than the base. The 1736 smoothed sherds share all of the characteristics of the simple-stamped sherds except . that they show some striations and a tendency to reflect highlights, although typically they have a dull surface. One check-stamped sherd (pl. 9, 2) was recovered from the back dirt in excavation 8. The 896 decorated sherds which carry incised or trailed lines are from the vessel shoulder. Except for a single curvilinear element (fig. 16, 7) all patterns are rectilinear. Several patterns and pattern elements are recorded (figs. 14-16). Decoration was applied to the vessel shoulder prepared by some horizontal smoothing, and on most vessels the vertical simple stamps are only partly obliterated. On the thinner shoulders, with a thickness of 2 mm. or less, the exterior designs resulted in raised impressions on the interior. The 239 red-filmed sherds are generally simple-stamped or smoothed, although a few of them are decorated. In all instances the film oceurs on the interior of the vessel. The 61 cord-roughened sherds are from vessels which were malleated with a paddle wrapped with a fine vegetal cord with an S twist. The cord used was somewhat less than a millimeter in thickness (pl. 18, j-k). Two large sherds are from miniature vessels. One of them, from House 4, is evenly smoothed and rounded, and is from a vessel about 55 mm. in diameter; wall thickness is 5 mm. - The other sherd, from the surface, is from a vessel about 65 mm. in diameter, with a base 10 mm. thick and a shoulder 7 mm. thick. The exterior is irregularly smoothed; the interior is likewise rough. Neither vessel was decorated. DESCRIPTIVE CATEGORIES AND TYPES 1. TALKING Crow SrraicutT Rim (pl. 9, e-g, i-k, and pl. 14, @). SAMPLE: 483 vessels, one of them restored. SURFACE FINISH: The rims and shoulders are vertically simple-stamped and im- perfectly smoothed, although a few shoulders, particularly near the neck, are wholly smoothed. Four rims are vertically brushed on the outer surface. __ DECORATION : Lip: The majority of lips are modified by tool impressions : 297 are obliquely impressed ; 93 are transversely impressed ; 66 are punctated ; 2 are cross- hatched; and 25 are plain. Tool impressions occur on the interiors of 79 rims (pl. 9, i-j) ; 57 are obliquely impressed and 22 are transversely impressed. Rim: The rim exterior is plain. Shoulder: This area is decorated with incised or trailed lines on 5 vessels. On 4 of them, decoration occurs on rims with oblique tool impressions on the rim interior. There are at least 2 occurrences of opposed diagonals Riy. Bas. Sur. ae: Pap. No. 34]. DEMERY SITE—WOOLWORTH AND WOOD 93 ot tla eo a AMANAVAAYANAS I AALALLEE ANAAAAAAAYS YLT SSSI WAAAY AAYS Ficure 14.—Pottery rim and shoulder design elements. (fig. 14, a) and one elaborate variant of the same pattern (fig. 15, a). The latter pattern is on a vessel with 2 opposed, vertically pierced lugs. ForM: Lip: Rounded to flat. On rims with deep or broad tool indentations there is some extrusion due to displaced clay. Rim: Characteristically straight to slightly outflaring, including rim forms A-B, D-G, and I (fig. 13). Height varies from 15 to more than 55 mm., with a mean of about 35 mm.; thickness is from 4 to 8 mm., with a mean of 6 mm. 94 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 SSN Y EZ tee Ceoeescees ee, Ficure 15.—Pottery rim and shoulder design elements. Neck: Constricted on outflaring rims. Shoulder: This area is usually about 45 degrees from the vertical, although some are more steeply sloping or more nearly horizontal. Size: The projected ares of 9 large rim sherds indicate vessel mouths of 46, 132, 158, 180, 190, 202, 210, 212, and 218 mm. diameter. Some vessel mouths may have been less than 40 mm. wide but probably only a few were more than 225 mm. wide. Riy. Bas. Sur. Pay. No. 34]. DEMERY SITE—WOOLWORTH AND WOOD 95 ‘a Ficure 16.—Pottery shoulder design elements. Appendages: A single plain loop handle was attached to the lip and shoulder of one rim, and there are 4 rim projections (pl. 18, @) and 49 lugs in this series. The lugs project out 5 to 8 mm. from the lip, generally at a 90- degree angle (pl. 9, f-g). Nine of the lugs are vertically or obliquely pierced (pl. 9). The rim projections, as well as the lugs, are com- monly decorated by a continuation of the lip decoration. 96 BUREAU OF AMERICAN ETHNOLOGY -. [Bull. 189 Restored vessel: One vessel of this type (pl. 14, a) has the following dimensions: Height. 22250825 eA EE ee Ee 213 mm Diameter at:shoulders.-< CC 5 i524 ib 6 22 Mee 265 Dinmetet.at neck= =. 2. SVAN ANS A ee See m by A Mouth, diameter = + at ters Oe SN 175 PTC TCKNENS ee eee ee ee 5 Rim /body thiekniess_£ 5 >- = 52-5 bs Se 2-6 Capacity (liquid) 220 — = oe ee ee 5 qts. ComMeEntTs: This type most closely resembles the type Talking Crow Straight -Rim from the Talking Crow site (Smith, 1951, pp. 36-37, pl. 8, g-k). Dif- ferences between the Demery and the Talking Crow samples consist of minor deviations in rim size and the presence of different shoulder patterns at Demery. A comparison of the Demery pottery with the Talking Crow Straight Rim from the Spain site (Smith and Grange, 1958, pp. 101-102, pl. 30, h-7) reveals more distinctive differences: interior lip notching and handles (pres- ent at Talking Crow and Demery) are lacking in the Spain site. The Demery sample is similar to most of the pottery designated by Cooper (1949, pp. 306-308) as Category C. Smith and Grange (1958, p. 102) remark that some of the Wheeler Ridged and Wheeler Plain rim sherds from Scalp Creek (Hurt, 1952, pp. 738-75) are “indistinguishable from the type” as described from the Spain site. There are also similarities between the Demery pot- tery and certain of the noncollared rims from Arzberger (Spaulding, 1956, pp. 149-164). 2. FLARING RIMS, HORIZONTALLY INCISED (pl..10, cj). SAMPLE: 476 vessels. SURFACE FINISH: Undecorated shoulders were vertically simple-stamped, but decorated vessels were smoothed before the application of the pattern. Many rims show evidence of vertical stamp impressions under the horizontal smooth- ing and incising. DECORATION : Lip: The 411 rims with clear evidence of treatment include 256 indented with oblique tool impressions; 59 with tool impressions transverse to the lip; 22 with vertical or oblique punctates ; 28 with oblique tool impressions on the interior rim; 14 with low oblique, elongated punctates along the lip; and 82 plain rims. Rim: The rims are horizontally incised or trailed between the lip and the neck. Bands consist of 2 to 13 lines; 90 percent of the vessels have 3 to 10 lines, with an average of 7 to a vessel. Twenty rims have stab-and- drag lines, either alone or in combination with incised lines (pl. 10, 7). On nine rims the horizontal lines are interrupted by incised triangles , fr" 10, h). Shoulder: Twenty-nine vessels are decorated with incised or trailed shoulder patterns. Three complete patterns observed include: Fig. 14, a, six ex- amples; fig. 14, b, one example; and fig 14, c, one example. One large sherd from a small vessel is unique in having a continuous series of nodes at the angular break of the shoulder (pl. 10,7). Form: Lip: Rounded to nearly: flat. There are exteagony on a few rini$ due to coarse tool indentations: \ Rim: Characteristically straight to slightly eatintiahe ‘Height varies from 8 mm. on small vessels to a maximum of 65 mm. on a large rim decorated Riv. Bas. Sur. we Pap. No. 34]. DEMERY SITE—WOOLWORTH AND WOOD 97 with stab-and-drag lines; the average height is between 35 and 40 mm. Thickness is 4 to 8 mm., averaging between 5 and 6mm. The thickness of the lip and the mid-rim is nearly the same, with the neck about 1 mm. thicker. Rim forms include A—B, H-F, and H (fig. 13). Neck: Constricted on outflaring rims. Shoulder: Most of the shoulders are about 45 degrees from the vertical, al- though some appear to have been more nearly horizontal. Body: Several partly restorable rims indicate vessels with a globular body and wide, circular mouths. Size: Three large rims have mouths of 118, 141, and 180 mm. diameter. A few sherds suggest that mouths may have attained widths of about 200 mm. Appendages: The 18 handles include 14 loop handles and 4 strap handles; in each instance they appear to have been paired on the rim. They were riveted to the mid-rim or upper shoulder and welded to the lip. Loop handles are invariably decorated with horizontal lines, in effect a continuation of the rim design (pl. 10, j) but strap handles are often decorated with herringbones or other incised designs (pl. 10, g). Thirty- seven strap handles, detached from the rims, are arbitrarily assigned to this group since handles are lacking on most of the remaining groups; many of them are horizontally incised. There is much variation in the size and form of the strap handles, with widths varying from 12 to 65 mm. Three rim projections are decorated by a continuation of the lip decora- tions; they extend vertically from the lip to a height of 5mm. Twenty- four lugs, extending out at right angles from the rim (pl. 10, h) are decorated on the upper surface by the same impressions occurring on the lip, but here the impressions are larger and deeper. One lug is centrally incised (pl. 10, e). ComMMEnTs: This pottery very closely approximates the type Grey Cloud. Hori- zontal-Incised from the Spain site (Smith and Grange, 1958, pp. 102-103). It differs from this type in having a larger number of horizontally incised lines, in lacking the shoulder patterns common to Grey Cloud vessels, and in the presence of appendages. Grey Cloud Horizontal-Incised grades into Iona Horizontal-Incised at Spain, the latter type having protruded T- or L- shaped lips, lacking at Demery. The Demery sample is similar to most of the pottery designated by Cooper (1949, pp. 303-306) as Category B. This rim category is also related to the type Wheeler Horizontal-Incised from the Scalp Creek and La Roche sites (Hurt, 1952, p. 76) ; to certain of the hori- zontally incised noncollared rims at Arzberger (Spaulding, 1956, pp. 153-157) ; and to the type Evans Incised, from sites of the Redbird Focus in north central - Nebraska (Wood, 1956). 3. FLARING RIMS, OBLIQUELY INCISED (pl. 13, D, g). SAMPLE: 15 vessels. SURFACE FINISH: The finish is not discernable on most sherds due to horizontal smoothing on the rim, and to the fact that most rims are broken from the body at the neck. On one partly restorable vessel (fig. 15, c) the entire body below the neck is cord-roughened. Above the angular shoulder the cord impressions are vertical, but below the shoulder they are random. ‘The cord used was less than a millimeter in diameter; it was made from two twisted fibers, but smoothing has obscured the direction of the twist. DECORATION Lip: Oblique tool indentations occur on 11 rims; on 1 rim the lip is punctated, and the other 2 are plain. 98 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Rim: Opposed oblique incised lines form a continuous band around the rim of 11 specimens (fig. 15, c) ; vertical lines occur on 2 rims together with oblique lines; and 1 rim is lightly incised with chevrons (pl. 13, b). Shoulder: One of the two incised shoulders is decorated (fig. 15, ¢) over a cord-roughened surface. The other and fragmentary decoration was im- posed over a surface of indeterminate nature. Form : Lip: Flat to somewhat rounded. Rim: Straight to somewhat outflaring, including rim forms B and F (fig. 13). Height ranges from 14 to 52 mm., averaging about 30 mm., and thickness from 4 to 8 mm., averaging about 6 mm. Shoulder: The shoulder is flat and steeply sloping, with an angular break at the lower border on the cord-roughened vessel. Size: Several measurements of the cord-roughened vessel include: ds (2) G21 as opr tl A ce cg. nl i tol I Bb ein, SY ca. 1380 mm, DIGMELOL ab NOOUIG CT te ee ee en a ee ee ca. 150 <0 Yes ECS WE CO SY AE eh i a pce i ek ABE I ts OL ca. 100 MET O/T ae LC KOSS eee ee ee a £4 EET ELL N bes e e em eae a yA | Ody Will sGOICKMCSS == esr oe ee eee ee eee 3 Appendages: One rim has a pierced lug extending out at a right angle 17 mm. from the rim, with a vertical perforation 9 mm, in diameter. It is on a lip decorated with punctates, although the top of the lug is decorated with transverse tool impressions. CoMMENTS: These rims are related to the recurved rims from Demery (Group 9) bearing obliquely incised lines, although they lack the pinched border below the incised rim design on the latter specimens. This group of rims clearly resembles some of those of the type Iona Diagonal-Incised from the Spain site (Smith and Grange, 1958, p. 100, pl. 30, e-g), but the Demery specimens lack the rim protrusion typical of Iona Ware. 4, FLARING RIMs, BEVELED OR BRACED (pl. 13, cf, and pl. 14, 0). SAMPLE: 105 vessels, one of which is restored. SURFACE FINISH: The interior and exterior rims are horizontally smoothed. This is the case with the restored vessel (pl. 14, b), where the entire surface of the vessel is smoothed, nearly obliterating the simple-stamping, which extends from the neck to the base. The stamps occur in a somewhat spiral fashion around the vessel body. DECORATION : Lip: On four vessels the lip interior is indented by oblique or transverse tool impressions. Rim: Oblique to transverse tool impressions occur on the uppermost part of the exterior rim; some rims are decorated with oval or circular punctates. Form: Lip: Usually pointed, although some are rounded. Rim: The straight to outflaring rims are beveled at the lip, the lower edge of the bevel sometimes being extruded by decorative impressions. Rim form J is typical, but on three rims, paste extruded by the pressure ap- plied during the decoration is pushed down, resulting in rim form EK (fig. 13). Riv. Bas, Sur. — bap. No’ 34] DEMERY SITE—WOOLWORTH AND WOOD 99 Shoulder: The rims join the body in a smooth curve. On the restored vessel (pl. 14, b) the shoulders are rounded and steeply sloping, and the body is globular with a round base. Size: Measurements of the restored vessel include: TENT ee ee cede ae ea gece Shee reall ed as See 313 mm Mismeterg at (SUOUIOCE 2.2 oe ae ee eee 355 Orifice; lsMeCtens = 2 ace eee ee eee 255 iy NI CKNOSS= Sa o8 eee oe ee nee Se 8 DE yOu MEA DESI sat ae as ae oe ak iy tra be sehen 45 Rim/body, thickness == ==> 8 os Ae eee 3-5 @Wanacity «(liquide ons ee he 18 qts. Appendages: Lugs on three vessels project out from the lower limits of the lip bevel. They are decorated by a continuation of the rim design. Comments: Some of the attributes of this group are shared by pottery from Biesterfeldt, a Cheyenne site on the Sheyenne River in southeastern North Dakota (Strong, 1940, pp. 370-376; Wood, 1955, pp. 3-8). At the Biesterfeldt site the rim is characteristically beveled or “wedge-shaped” in cross section ; this rim form occurs on a few of the Demery rims. A few of the rims from the Huff site of the type Huff Plain and Huff Braced Rim (Wood, MS. a) recall this rim form at Demery. 5. Fuarine Rims, Corp-ROUGHENED (pl. 13, j-k). SAMPLE: Six vessels. Paste: Generally similar to that described under the heading “General Char- acteristics,” except that the grit is somewhat coarser and the clay is not as well worked. The edges crumble easily on the more coarsely tempered sherds. A single sherd from Feature 51 contains particles of what appear to be shell, which effervesce under the action of diluted hydrochloric acid. It is possible that the shell was originally in the clay, and some of it has been destroyed by chemical weathering, since some of the particles effervesced only slightly even when treated with undiluted acid. SURFACE FINISH: The exterior rim is vertically cord-roughened, this treatment extending down over the shoulder. There is some horizontal smoothing on necks, although body sherds may or may not be smoothed. The cord used had an S-twist. DECORATION : Lip: One rim is marked with oblique tool impressions; another has oblique punctates applied at a low angle to the lip top by a pointed tool; the other four are plain. Rim/shoulder: No decoration observed. ForM : Lip: Fiat. Rim: Slightly outflaring, most rims closely resembling rim form J (fig. 13). Rim thickness is 7 to 8 mm.; height is indeterminate, but in excess of 30 to 50 mm. Body: One large sherd built up from fragments suggests that the shoulder is rounded, probably blending smoothly with the rim. The even thickness of all body sherds, about 4 mm., indicates that the vessel walls were of nearly constant thickness. Appendages: None observed. CoMMENTS: The cord-roughened surfaces and certain characteristics of the paste set this group of rims apart from the rest of the Demery pottery. From what can be observed, however, vessel form does not appear to depart signifi- 661-932—64——_8 100 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 cantly from that observed of the pottery relating to the major occupation of the site by the Demery Component. Since most of the Woodland pottery from North Dakota appears to be sand-tempered (Wood, 1962 a) there seems no good reason to attribute this pottery to a Woodland occupation of the site. Perhaps the nearer affiliations of this pottery may be found in the Campbell Creek types, defined at the Talking Crow site (Smith, 1951, pp. 37-39). A major distinction between the Demery pottery and the Campbell Creek types is the presence of a clearly defined rim-shoulder juncture on most of the Campbell Creek pottery, and the lack of such a juncture on the Demery vessels 6. Fort Rice Corp-IMPRESSED (pl. 11, d-h, j-1). SAMPLE: 79 vessels. SURFACE FINISH : The shoulders are vertically simple-stamped and are smoothed Lips, rims, necks, and interiors are horizontally smoothed. DECORATION : Lip: On 6 vessels the lip is obliquely tool-impressed; on one it is obliquely cord-impressed ; the remaining lips are plain. Rim: Horizontally applied cord-impressed lines alternate with curvilinear or triangular elements. From 2 to 8 horizontal lines occur, most vessels having 5 or 6 lines. In the 16 examples of triangular elements, the center of the triangle is plain. At the apex of each triangle is a raised tab (pl. 11, f-7) ; a single exception in which the tab is absent occurred on a curvilinear element, or “rainbow.” ‘Ten curvilinear designs were applied over a small applique node on the mid-rim; if a wholly restored vessel rim is typical (fig. 15, 0), four nodes occurred on each vessel. A horizontal series of punctates occurs on the lower rim of two vessels (pl. 11,d). Three vessels, decorated with cord-wrapped stick impressions, are otherwise similar to the above specimens (pl. 11, 7). Shoulder: The shoulder of only one vessel is decorated. The pattern con- sists of a complex rectilinear design (fig. 15, 0) bordered by oblique punctates. ForM : Lip: Predominantly fiat, with a few rounded examples. Rim: Hight forms were observed, of which six are recurved, including rim forms L-Q (fig. 18). Two rims, of forms V and W, are somewhat more angular and might be termed collared. Rims range from 30 to 60 mm. in height, and are 4 to 7 mm. thick. Neck: Constricted to varying degrees and, except for rim form WN, they join the shoulder in a smooth curve. Body: Apparently globular, with round bottoms. Size: The orifice of the reconstructed rim (fig. 15, 6) measures 152X156 mm. Since the rim was badly shattered, the difference between the two measurements need not be interpreted that the vessel mouth was oval. The projected arcs of three large rim sherds indicated mouths between 182 and 250 mm. in diameter. Appendages: There are no handles; nodes are discussed above, under decoration. CoMMENTS: These rims are similar to the type Fort Yates Cord-Impressed from the Paul Brave site (Wood and Woolworth, 1964, pl. 5; Hewes, 1949, pp. 65-66), but they differ in the conformation of the rim, the pres- ence of the appliqued nodes, and the smaller cords used to impress the design. Fort Yates Cord-Impressed rims have a wedge-shaped cross sec- tion due to a thickening near the rim base; only two rims from Demery are so thickened. The type Fort Rice Cord-Impressed, described from the Huff Be NS ay): DEMERY SITE—WOOLWORTH AND WOOD 101 site (Wood, MS. a) is more nearly like the Demery pottery. There are only minor differences between the Huff pottery and that from Demery, prin- cipally in such matters as the number of cord-impressed lines and the degree of rim curvature. This pottery group is, therefore, tentatively identified as Fort Rice Cord-Impressed. 7. Fort Rick TRAILED (pl. 12, e—h, j-m). SAMPLE: 132 vessels. SURFACE FINISH: Shoulders are vertically or obliquely simple-stamped and partly smoothed. Necks and rims are horizontally smoothed, although traces of vertical stamps are often visible under the rim designs. DECORATION :! Lip: The 128 rims with clear evidence of design include 45 with oblique tool impressions ; 10 with transverse tool impressions; and 3 with punctates. The other 70 lips are plain. Rim: The rim is horizontally incised or trailed with 2 to 11 lines. Of the 72 rims carrying complete series, 57 have 4 to 7 lines. Nine rims have stab-and-drag lines (pl. 12, #) either alone or in combination with incised lines. Triangular elements interrupting the horizontal lines occur on 28 rims (pl. 12, k-m). On 22 of these rims the apex of the triangle is near a small applique node or a rim projection. Shoulder: The shoulders of two vessels are embellished with incised patterns consisting of alternating chevrons and horizontal and vertical lines bordered by oblique lines (fig. 14, d). TorM : Lip: Predominantly flat, although a few plain rims are rounded. Rim: Gently curving recurved rims of nearly constant thickness from the lip to the shoulder include rim forms Z—P, and FR (fig. 13). A few rims thicken toward the base of the design area, and are more aptly termed collared. The latter rims include forms T-V (fig. 13). Height ranges from 35 to 65 mm., averaging about 45 mm.; thickness, from 4 to 9 mm., averaging about 5 mm. Neck: Constricted. Shoulder: Most of the shoulders are at about a 45-degree angle from the vertical, although a few were more nearly horizontal. Size: The projected arcs of several large rim sherds indicate vessel mouths of 227 and 253 mm. Appendages: Applique nodes applied to the mid-rim were on 9 vessels, and applique or pinched-out lugs or rim projections occur on 28 vessels. A few of the applique nodes on the mid-rim bear central, vertical punctates. CoMMeENTS: The group of rims is related to the flaring rims with horizontally incised lines (Group 2). The latter group, however, lacks nodes applied to the mid-rim. The nearer affiliation of this pottery is with the type Fort Rice Trailed from the Huff site (Wood, MS. a). The only difference between the Huff pottery and that from Demery is the lack of collared rims at Huff (they are rare in Demery) and the smaller number of horizontal lines in the rim design at Huff. This pottery group is accordingly tentatively identified as (ort Rice Trailed. 8. Recurvep Rims, Horizonrarty INCISED AND PINCHED/PUNCTATED (pl. 10, a-c). SAMPLE: 23 vessels. SURFACE FINISH: The interior and the exterior rim and the neck are hori- zontally smoothed. 102 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 DECORATION : Lip: On the 11 rims with lips, 8 are impressed with oblique tool impressions ; 1 has oblique tool impressions on the interior lip; 1 has circular punctates ; and 1 is plain. Rim: On the 11 rims complete from neck to lip there are 2 to 7 horizontally incised lines, with more than half of them (6) bearing 5 lines. Below these lines are finger-pinched indentations on 18 vessels, and punctates on 10 vessels (pl. 10, a—b). One rim with two horizontally incised lines above a line of punctates is incised with six lines below them; and one atypical rim (pl. 10, c) lacks either pinching or punctates. ForM : Lip: Flat to somewhat rounded. Rim: Three forms occur. Rim forms U-V (fig. 13) are classed as collared, and comprise six of the eight classifiable rims. One rim is of form M, and another is of form X. Neck: Constricted. Body/size: Sherds are generally small, but one large rim indicates an orifice of about 200 mm. in diameter. Appendages: None observed. Comments: A number of characteristics of this pottery, including rim form and decorative elements, are shared with the type Arzberger Horizontal Incised (Spaulding, 1956, pp. 139-141, pl. 8), although none of the Demery vessels appear to have cord-roughened shoulders. 9. RECURVED RIMS, OBLIQUELY INCISED AND PINOHED/PUNCTATED (pl. 18, h). SAMPLE: 12 vessels. SURFACE FINISH: Originally vertically simplestamped, then horizontally smoothed on rims and necks. Two rims bear nearly obliterated impressions that suggest they were malleated with a cord-roughened paddle, then smoothed. DECORATION : Lip: Oblique tool impressions occur on five lips, and two are plain. The lipis absent on the remaining rims. Rim: Oblique incised lines occur on 11 of the 12 rims; on one rim a series of 3 oblique lines, forming a chevron, interrupt a series of 10 horizontally incised lines. On the eight sherds retaining the lower rim, three have a finger-pinched border and five are punctated. Shoulder: One rim bears traces of an incised pattern. ForM : Lip: Flat to somewhat rounded. Rim: The gently recurved rims are of nearly constant thickness from the lip to the neck, except where the finger-pinchings thicken the cross section. Rim forms are U-V (fig. 18). Rim heights range from 25 to 47 mm. ; thickness, from 4 to 8 mm. CoMMENT: These rims are related to the flaring rims with obliquely incised lines (Group 3), although the latter rims lack the rim profile and the modified rim base occurring on this group. 10. Recurvep Rims, PINCHED IN Mip-Rim (pl. 13, 4). SAMPLE: 8 vessels. SURFACE FINISH : The rims are horizontally smoothed on the interior and exterior. DECORATION : Lip: Oblique tool impressions. Rim: A horizontal series of finger-pinched impressions occur on two rims at the mid-rim; the remaining rim is obliquely tool-impressed (pl. 13, i). Riv, Bas. Sur} DEMERY SITE—WOOLWORTH AND WOOD 103 Form : Lip: Flat. Rim: Gently recurved, with constant thickness from the lip to the neck, of form Z (fig.13). Neck: Constricted. Body size: No data. CoMMENTS: These sherds seem to be variants of the rims with obliquely or horizontally incised lines with finger-pinched or punctated rim bases (Groups 7-9), but these rims are less complex. 11. REcURVED RIMS, PLAIN (pl. 12, a—b,1). SAMPLE: 18 vessels. SURFACE FINISH: The rims and shoulders were first simple-stamped and then horizontally smoothed. DECORATION : Lip: Oblique tool impressions occur on five lips; one has transverse tool impressions; one has oblique tool impressions on the inner rim; and the rest are plain. Rim: Plain. Shoulder: Incised chevrons and oblique incised lines occur on the shoulder of one vessel (fig. 15, f). Form: Lip: Round to somewhat flat. Rim: Gently recurving rims of nearly constant thickness from the lip to the neck include forms Z—M and P-Q (fig. 13). Height varies from 25 to 60 mm., averaging about 40 mm.; thickness, from 4 to 8 mm., averaging about 6 mm. Neck: Constricted. Shoulder: On two vessels it is about 30 degrees from the vertical. Appendages: An applique lug occurs on the mid-rim of one rim, and a lug scar appears on another rim. CoMMENTS: There are close similarities between this group and Examples D from the Paul Brave site (Wood and Woolworth, 1964), and “Riggs Plain” from the Thomas Riggs site (Kleinsasser, 1953, p. 28). It also resem- bles Example D from the Huff site (Wood, MS. a). 12. Corp-IMPRESSED RiMs, PxaMpie A (pl. 11, @). SAMPLE; 1 vessel. PASTE: Temper: Sparse, minute particles, including some sand. Texture: Well-worked clay with a laminated core. Color: Light brown. SURFACE FINISH : Interior and exterior horizontally smoothed. DECORATION : Rim: Four horizontally cord-impressed lines on exterior. ForM: Lip: Pointed. Rim: Straight, with an added fillet below the lip, of form K (fig. 13). ComMMENTs: This rim does not closely resemble any other reported pottery. 13. Corp-IMpressep Rims, EXAMPLE B (pl. 11, bd). SAMPLE: 1 vessel. PASTE AND SUBFACE FINISH : See Hxample A. DECORATION : Lip: Oblique tool impressions. Rim: Three horizontally applied cord-impressed lines. 104 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 ForM: Lip: Flat. Rim: Straight, with the rim blending smoothly with the shoulder, of form B (fig. 18). CoMMENTS: This example may be a cord-impressed variety of the horizontally incised, flaring rim sherds (Group 2). 14. Corp-IMPRESSED Rim, HxAMPLE C (pl. 11, 7). SAMPLE: 1 vessel. PASTE AND SURFACE FINISH : See Hxample A. DECORATION : Lip: Chevrons, composed of shallow, oblique tool impressions on the interior and exterior. Rim: Seven horizontally applied cord-impressed lines. Shoulder: Herringbone pattern, the only occurrence of this design in the site sample. ForM: Lip: Pointed. Rim: Straight to slightly recurved, of form D (fig. 13). Shoulder: Flat and sloping, with an angular break at the juncture with the body. Size: The orifice measures 95 mm. in diameter. CoMMENTS: This example may be another cord-impressed variety of the hori- zontally incised vessels with flared rims (Group 2). 15. Corp-IMPRESSED R1iM, EXAMPLE D (pl. 11, ¢). SAMPLE: 1 vessel. PASTE: Temper: Coarse grit. Texture: Rather crumbly, with loosely compacted clay. SURFACE FINISH: Horizontally smoothed on interior and exterior. DECORATION : Lip: Crosshatched. Rim: Five horizontally applied cord-impressed lines. Form: Lip: Beveled down and out. Rim: Straight, of form Jd (fig. 13). CoMMENTS: This rim does not closely resemble any previously described pottery. 16. Riecs Cross-HATcHED RIMs (pl. 12, d). The sample consists of five sherds from four vessels. The lip of one vessel is carefully crosshatched, but the same design on the outer rim is irregularly applied. The paste is coarse and the grit is large-grained. This single sherd is from House 4, while the rest of them are from the surface. The other sherds are more carefully crosshatched, with oblique tool impressions on the lips of two vessels, and a plain lip on the other vessel. The paste and tem- per of these sherds is less coarse than that of the vessel from House 4. Hach of the sherds of this type is identical to examples as decribed from the Paul Brave site (Wood and Woolworth, 1964), a component of the Thomas Riggs Focus a few miles north on the Missouri River. 17. Rices STRAIGHT RiM (pl. 138, 1). The 12 rims from the 9 vessels of this type are distinguishable from the other sherds from the site in having (a) a coarse temper and a crumbly, granular texture, (b) cross sections 8 to 14 mm. thick, as opposed to cross sections 4 to 8 mm. thick in the majority of rims, and (c) contorted cores, in contrast to the more commonly laminated cores. Vertical rim projections occur on two rims. Biv. Has. Si DEMERY SITE—WOOLWORTH AND WOOD 105 Four rim sherds are from the surface, and one each is from Feature 51 and Feature 55, irregularly shaped pits near House 5; sherds from 3 vessels were in House 4. A number of fragments of one large vessel were in Feature 97, a large irregular pit in the south side of House 4, and a decorated body sherd (fig. 16, f) was in the same pit. This sherd may be duplicated among those from Paul Brave (Wood and Woolworth, 1964, fig. 1). One of the center posts of House 4 was intrusive into the fill of this pit, providing evidence that the pit was dug and filled before the house was built. This circumstance suggests an occupation of the site by a group predating the occupation repre- sented by the majority of excavated remains. This occupation, attributed to the Thomas Riggs Focus, is presumed to predate the Demery Component. TABLE 2.—Probable number of vessels represented by the“rim"sherd sample Number] Percent of vessels Demery Component (i alkine Crowesiwalcht Rima sea Fe Les eee 483 359.6 2} Marine rims, horizontally incised: .2.-22.22. 525-5. 476 35. 1 Soohlaring mms wobliquely incisedas = === ==) s—= = aa 15 ial 4. Flaring rims, beveled or braced _-_----------------- 105 ews bay aring rims. cord-roughened....2 f= 52S. = seco = 6 4 GayHorninRice @org-lmpresseda2 2 sna = ee = 79 5. § ( Pore Rice Urailed Oe Ul ee fhs Jn See ee ares 132 9.7 8. Recurved rims, horizontally incised and pinched/ PUMCUDCC Late eek Sees eee ec 23 ie 9. Recurved rims, obliquely incised and _ pinched/ punctate de Sra st 2 ae cea a es Sie A 12 .9 10. Recurved rims, pinched in mid-rim________-_------ 3 es HIESIeE curved rims tp lalme = sae hese ee Senn eee hare a 18 1.3 12. Cord-impressed rim, Example A____-------------- 1 207 13. Cord-impressed rim, Example B____---_----------- 1 07 14. Cord-impressed rim, Example C____--__-_-_------- 1 . 07 iS. Cord-impressed rim, Example DoJ. .~.2-6.5---=- 1 07 ARO tales ees Serpe RR eT ES EOS 2k ie AE Latin 1, 356 99. 78 Thomas Riggs Focus Component IG Riges Cross-Hatehed Rim. o3.. 2202.2 225-5 oo ane 4 30. 8 River bee SCEAN Glu? Turin ens ae eet mete ees SOC NE 4 9 69. 2 aD eke a a Na ae SC a cat NL eS aN 13 100. 0 CHIPPED STONE Projectile points (130 specimens) : The points in this sample are all essentially triangular in outline, with straight to somewhat convex edges. They were made by the careful removal of small pressure flakes from both faces, with a final product having evenly flaked sides and a lenticular or biconvex cross section. Three groups are distinguished on the basis of the shape of the base and the presence or absence of side notches. Seventy-nine of them are made from a light-gray chert, and twenty-two are made from Knife River flint (chalcedony). The remaining 29 specimens 106 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 are made from quartzite (7), variously colored cherts (14), Badlands chalcedony (6), and agate and petrified wood (2). Triangular blade, straight base, unnotched (75 specimens).—These points have straight to somewhat irregular bases, with gently convex blade edges. They range in length from 20 to 58 mm., although 80 percent of them are less than 40 mm. long. Width is 12 to 26 mm., with a mean of about 19 mm. Thickness ranges between 8and6mm. The smallest point has a weight of 0.4 gram, and the largest weighs 16 grams, but only 5 points exceed 2 grams in weight (fig. 17, a-d). Triangular blade, straight base, two side notches (27 specimens) .— These points have fairly straight, regular bases, with straight to gently convex blade edges. They range in length from 20 to 41 mm., al- though 80 percent of them are less than 31 mm. long. Width is 10 to 18 mm., with a mean of about 14 mm. Thickness is from 2 to 6 mm. The smallest point has a weight of 0.6 gram, and the largest of them weighs 2.3 grams, most of them averaging about 1 gram (fig. 17, e-h). Triangular blade, concave base, two side notches (15 specimens) .— These points have even, concave bases, with straight to gently convex blade edges. Length ranges from 21 to 44 mm., although 80 percent of them are less than 33 mm. long. Width is from 12 to 18 mm., with a mean of about 14mm. They are 3 to 7 mm. thick. The smallest point has a weight of 0.7 gram; the largest weighs 3.0 grams, but only one of them exceeds 1.6 grams in weight (fig. 17, 2-7). Point fragments (13 specimens).—These fragmentary points con- sist of the tips of points which have been broken through side notches; the form of the base is indeterminate. E'nd scrapers (226 specimens) : The outlines of these specimens tend to be triangular to rectangular, although many of them are irregular in form. The working edge, on the end opposite the bulb of percussion, is generally steeply flaked. Most of them bear large flake scars on the convex upper surfaces (fig. 19, d-g), but a few of them are flaked on the upper surface (fig. 19, c,h). The smallest scraper measures 18X19 X6 mm., with the largest of them measuring 76X46 X15 mm.; the average is about 40 X24 10 mm. (fig. 19, e-h). Expanding base drills (4 specimens) : One complete specimen, made of petrified wood, has a short, tri- angular point on one side of a leaf-shaped flake. The specimen is unifacially flaked; it is 32 mm. wide and 29 mm. long (fig. 19, a). The other three specimens have nearly parallel-sided shafts which are broken off a few millimeters from the base (fig. 19, 6). Bases are 18 to 832 mm. wide, and the shafts suggest lengths originally in excess of Riv, Bas: Sr) DEMERY SITE—WOOLWORTH AND WOOD 107 Figure 17.—Projectile points. BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 108 Ficure 18.—Chipped stone knives. Riv. Bas. Sur. bay. No ga] DEMERY SITE—WOOLWORTH AND WOOD 109 k' Ficure 19.—Chipped and ground stone artifacts. \ a-b, Drills. e-h, End scrapers. i-j, Game pieces, &, Catlinite disk-bowl pipe. 110 BUREAU OF AMERICAN ETHNOLOGY {Bull. 189 20mm. These shafts are alternately beveled, with rhomboidal cross sections. Material includes chalcedony, quartzite, and petrified wood. Broad knives (117 specimens) : Bipointed, convex blade edges (2 specimens) —Leaf-shaped knives with pointed ends are bifacially flaked, having lenticular cross sec- tions. One of them measures 115 X 36X10 mm.; the other 10040 8 mm. They are made from a gray chert and from fine-grained quartzite (fig. 18, e). Triangular blade, convex bases (35 specimens).—Blade edges are strongly convex; tips are pointed. Each is bifacially flaked, with a lenticular cross section. The three complete specimens are 78 to 95 mm. long, 31 to 33 mm. wide, and 8 to 9 mm. thick, although broken specimens suggest original lengths exceeding 100 mm.; widths of 46 mm.; and thicknesses of 15 mm. (fig. 18, f). Triangular blade, straight bases (3 specimens).—Blade edges are gently convex; tips are pointed. The bases of these bifacially flaked knives are thinned on only one side, resulting in beveled bases. They are 57 to 70 mm. long, 31 to 43 mm. wide, and 8 to 12 mm. thick. Materials were vein chalcedony, Knife River chalcedony, and gray chert (fig. 18, a). Oval knives (77 specimens) .—Thirty-one of these knives are oval to round, with convex edges and rounded to somewhat flattened ends. All are lenticular in cross section and are bifacially flaked. The six complete knives are 40 to 115 mm. long and 32 to 68 mm. wide. Thick- ness ranges from 7 to 11 mm. Broken specimens suggest widths were sometimes as low as 28 mm., and thicknesses ranged up to 14 mm. Knife River chalcedony, quartzite, petrified wood, and gray chert served as raw materials (fig. 18, ¢). Forty-six specimens have more nearly parallel edges. The six com- plete specimens range from 60 to 95 mm. long, and from 28 to 30 mm. wide. Thickness ranges from 10 to 13 mm. Broken specimens sug- gest that original lengths sometimes exceeded 100 mm., and widths exceeded 45 mm. Materials are petrified wood, vein chalcedony, and quartzite (fig. 18, g). Narrow knives (18 specimens) : These knives are similar in form to the broad knives described above as bipointed, with convex blade edges, but they are narrower and better made than the broad group. These differences may have resulted from their use in a special context. Their size and form is well adapted for insertion in the slotted bone or wood knife handles from the site (fig. 20, b-c). The five complete specimens are 62 to 83 mm. long, 15 to 24 mm. wide, and 6 to 10 mm. thick. One broken knife may originally have been 150 to 170 mm. long. Knife River chalcedony, quartzite, and variously colored cherts were used as raw materials (fig. 18, /). Riv. . Sur. Biv. Bas. Su DEMERY SITE—WOOLWORTH AND WOOD 111 wow orrr wm Fs ry Foy ec e aro ee Ae os ere A BO taka, ory Ficure 20.—Basketry, woodwork, and bone artifacts. a, Charred, twilled basketry frag- ment. b, Slotted bone knife handle. c, Charred, slotted wooden knife handle. d-e, Bison ulna tools. f, Bison radius pick. g, Bison scapula hoe. 112 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Asymmetrical knives (10 specimens) : The form of these specimens is best shown in the illustration (fig. 18, 6). They have one strongly convex edge, and one edge more nearly straight. The ends are convex to pointed. The three com- plete specimens are 77 to 102 mm. long, 36 to 55 mm. wide, and 8 to 11 mm. thick. Materials are chert and quartzite. Vein chalcedony knives (45 specimens) : These knives were made from slabs of vein chalcedony, one or more edges of which were bifacially flaked to form a sharp cutting edge. They are 20 to 80 mm. long, 15 to 67 mm. wide, and 5 to 10 mm. thick (fig. 18, d). Flake knives (89 specimens) : Flakes with bifacial flaking on one or more edges are classed as flake knives. The edge is generally even and regular, with carefully de- tached pressure flakes forming the cutting edge. Flakes are of almost every shape, with lengths up to 102 mm. Materials are quartzite, petrified wood, and chalcedony. Flake scrapers (154 specimens) : Flakes with regular, unifacial pressure flaking along one or more edges are classed as flake scrapers. These tools were made from con- choidal flakes of Knife River chalcedony, quartzite, colored chert, petrified wood, and a few pieces of stone which resemble Bijou Hills quartzite. This stone, common in sites along the Missouri River in southern South Dakota and northeastern Nebraska, may have been traded into the area. Modified flakes (170 specimens) : Flakes of varying size and form, having one or more edges bearing minute flakes detached by use, are classed as modified flakes. These flakes appear to have been modified by use and not by pressure flaking. Material is chalcedony, petrified wood, chert, Knife River chalcedony, and quartzite. A single flake of smoky obsidian (the only scrap of this material recovered at the site) resembles samples from Wyoming. Choppers (19 specimens) : Circular or oval slabs of quartzite, chert, and basalt, detached from the parent material by percussion, may have been used as choppers. Most of them have convex to straight edges modified by coarse, uni- facial percussion flaking, although some of them are bifacially flaked. These implements are large, measuring 90 to 260 mm. long and 55 to 30 mm. wide. One of them has shallow notches flaked into the edges and could have been hafted for use as an ax. Riv. Bas. Sur. Pap. No. 34] DEMERY SITE—WOOLWORTH AND WOOD 113 GROUND STONE Grooved mauls (19 specimens) : Eighteen of these tools were made from oval granite pebbles, with one specimen of fossiliferous stone composed of calcified organic mat- ter. Basically, they are oval stream pebbles modified only by the pecked groove that encircles them. These grooves are centered on 13 mauls, but on 6 of them they are offset toward one end. Only one of them has a polished groove. The ends are usually heavily battered. They range in size from 60 X55 mm. to 135 X 180 mm.; weight ranges from 14 to 914 pounds, with a mean weight of 4 pounds. Diorite celts (4 specimens) : Three of these tools are subrectangular in outline and in cross sec- tion, having been fully shaped by pecking and grinding (fig. 21, e). Lengths range from 115 to 145 mm.; widths, from 55 to 56 mm.; thick- ness, from 31 to 40 mm. The cutting edges are about the same width as the maximum width, and are blunted from use; polls are battered from use as hammers. A large piece of diorite, pecked over most of its surface, may be a celt in an early stage of manufacture. Pebble hammerstones (46 specimens) : These implements were made from granite and quartzite pebbles, and occur in three forms. Five of them, about the size of baseballs, were made from pebbles which have been battered on all surfaces. Another group of six circular hammers are also battered on all sur- faces, but they are consistently smaller and are evenly shaped. These latter implements are 50 to 65 mm. in diameter. The remaining 35 hammers are irregularly shaped river pebbles, tending to be oval in shape. Most of them were battered on one or more faces, and usually on the ends, but none of them were purpose- fully shaped. They range from 30 to 100 mm. in diameter. Discoidal hammerstones (8 specimens) : These tools are disk-shaped sandstone and quartzite stones; sides are flat and cross sections are oval. A shallow pit was pecked into each of the flat surfaces, possibly for finger holds, and the entire circumference of each stone is battered. Abraders (61 specimens) : Irregular pieces of fine-grained sandstone, scoria, shale, diorite, and granite either have shallow, cup-shaped depressions or flat, smoothed areas on one or more surfaces that suggest they were used as abraders. Grooved abraders (67 specimens) : These objects are composed of scoria and medium- to fine-grained sandstone. The irregular width and form of the grooves precludes 114 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 e f Ficure 21.—Ground stone artifacts. a, d, Arrow shaft smoothers. 0, Catlinite pipe bowl. c, Sandstone disk. ¢, Diorite celt. f, Grooved abrading stone. the possibility that they were used as shaft smoothers; it is more likely that they were used to sharpen awls or other tools. Many of them are small, with shallow V-shaped grooves which suggest that they were used for delicate abrading. They are 30 to 70 mm. long, and average slightly less in width. Other pieces of irregular sandstone have larger, more prominent U-shaped grooves. They vary greatly Riv. Bas. Sur. Dap. No. 34] DEMERY SITE—WOOLWORTH AND WOOD 115 in size, ranging from 40 to 170 mm. long and 25 to 110 mm. wide (fig. 21, f). Shaft smoothers (21 specimens) : Although most of these are fragmentary, they are generally elon- gated, boat-shaped pieces of coarse- to fine-grained sandstone with convex ends. All specimens have a U-shaped groove of varying depth along one flat side. One complete example is 100 mm. long and 40 mm. wide (fig. 21, d). Catlinite pipes (2 specimens) : One complete disk-bowl pipe is 140 mm. long and 33 mm. high. The prow projects 60 mm. beyond the forward edge of the platform. The forward end of the prow is serrated, and the platform has 14 paired, incised lines radiating out from the orifice in the platform center. The bowl was drilled with a tapered drill (fig. 19, %). A sec- ond pipe is a tubular fragment of a cylindrical bowl. The piece is 40 mm. long and 16 mm. in diameter (fig. 21,0). Sandstone disk (1 specimen) : One-half of a disk of fine-grained sandstone, 97 mm. in diameter and 10 mm. thick, has a lenticular cross section. The surfaces are smooth and even, and there is a shallow depression in the center of one face. The edges are scored by small notches, and one face is smudged with red ocher (fig. 21, ¢). Game pieces (2 specimens) : One of these objects is a smooth, oval stone 24 mm. long and 15 mm. wide, with an encircling groove at its midpoint (fig. 19, 7). The second stone resembles a smooth, fossil fruit stone; it is 8 mm. long and 7 mm. wide, with a raised ridge along the edges. One surface bears a deeply incised X (fig. 19,7). Mealing slab (1 specimen) : A large slab of granite with one smooth, depressed surface measures 421 mm. in length, 254 mm. in width, and 127 mm. in thickness. The depressed surface was ground smooth; it was probably used after the manner of a mealing slab, although no manos or mullers were recovered. Pigments (6 pieces) : Six pieces of sandstone and chalk may have been sources of paint. Four of them are of very fine-grained sandstone, impregnated with hematite; three of them are brick red in color and the other is light buff. Two small pieces of light yellow chalk have smooth, rounded surfaces. 661-932—64—9 116 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Unworked stone (2 pieces) : These items consist of a piece of unmodified calcite 65 mm. long, 22 mm. wide, and 12 mm. thick; and a piece of sheet mica 75 mm. square, with rounded corners. These two items may have been picked up and carried to the site as oddities. BONE ARTIFACTS Scapula hoes (100 specimens) : The scapulae of adult bison were used in the manufacture of these implements, although some smaller specimens may be from young bison or from elk. The supra-scapular border was beveled on the side bearing the scapular fossae, and these fossae are hacked away so that the surface of the bone is nearly flat. In a few instances there are suggestions that the edges were roughened by chopping at points about 100 to 150 mm. from the cutting edge, probably to provide a rough surface for binding a handle to the tool. The cutting edge may either be rounded or straight; the more heavily worn tools are rounded, whereas the others are more nearly square. The articulating end is retained without modification (fig. 20, 7). Sixty-four hoes are made from the right scapula, and 36 are from the left scapula. The range in length of the 46 complete hoes is 270 to480mm. Theshortest hoe, however, was about 40 mm. shorter than any other in the series, and the longest was about 40 mm. longer; ex- cluding these two atypical specimens, the range in length is 310 to 440 mm. Scapula knives (19 specimens) : These specimens were made from various parts of bison scapulae. Two groups are identified in the series. The first group consists of 10 specimens made from parts of the scapula on which the spinous processes were hacked down, retaining part of the thin flat bone in the scapula center as a blade. In general, they are paddle shaped, with roughly parallel or convex edges. Parts of the medial or lateral border often project beyond the blade and may have served as handles. Some of them may have been made from broken scapula hoes, since they often have a beveled edge on one end, on the side bearing the scapular processes; this bevel is characteristic of hoe blades. The edge of the scapula opposite the scapular process is sharpened on each of them. The blades are either convex or concave, the latter specimens resembling a scythe blade in some respects. Lengths of complete specimens range from 220 to 390 mm. (fig. 22, 7). The other nine specimens classed as knives are oval, triangular, or rectangular scapula fragments sharpened on one or more edges. They resemble the familiar bone tools generally called “squash knives,” and are 120 to 165 mm. long (fig. 22, A). Bap. NS! Say: DEMERY SITE—WOOLWORTH AND WOOD Luly Ficure 22.—Artifacts of bone, antler, and shell. a4, Scored bison rib. 0», Serrated rib tip. c, Arrow shaft wrench. d-e, Shell scrapers. f, Antler cylinder. g, Scapula ‘“‘cleaver.” h, Bone knife. Shaft wrenches (13 specimens) : These tools were made by drilling one or two roughly circular holes through a large rib, and rounding the rib ends. The three complete specimens are 240 and 295 mm. long; one of these has two oval holes spaced 75 mm. apart (fig. 22, ¢). The other two wrenches have a single, centrally located hole. All holes are oval, and all except one have beveled or rounded edges; diameters range from 11 to 16 mm. 118 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Five of them have lightly incised decorations, usually on the con- cave side of the rib. On one, the incisions consist of random diamond- shaped incisions along the midline; on another, the design is a wavy line with angular crests and troughs. Random crosshatching and X marks occur on two of them. Twelve shallow notches are cut into the edge of one of the latter pieces (fig. 22, c). The fifth incised wrench has two incised X marks on one edge. Knife handle (1 specimen) : A piece of large rib, 137 mm. long, is cut into rectangular form and smoothly polished. There is a small slot at one end, 60 mm. long and 13 mm. deep, made by removing the cancellous tissue. A V-shaped notch is cut into the handle behind this slot, and 15 mm. from this notch is a smaller one (fig. 20, 0). A similar specimen, of wood (fig. 20, c), is described under the heading, “Perishable Remains.” Fishhooks (2 specimens) : These implements are made from segments of mammal long bones. On one of them, 40 mm. long and 16 mm. wide, the sharp point forms an angle of about 40 degrees from the shank. The end of the shank is notched to assure a firm attachment to the line (fig. 23, d). A sec- ond object, superficially resembling a fishhook, probably had another function. This object, 60 mm. long and 20 mm. wide, has a rounded “point,” and the end of the shank shows no modification for attaching a line (fig. 23, e). Tubes (5 specimens) : Both ends of two eagle wing bones are cut, but they are not smoothed and the cancellous tissue is intact. They may represent a stage in the manufacture of a whistle or a tube; they are 62 and 111 mm. long. Both ends of another tube, made from the long bone of a deer or antelope, are cut square. The cancellous tissue is removed and the ends are scored (fig. 23, 2). Two other bones seem to be stages in the manufacture of bone tubes. One of these is the proximal end of a deer or antelope meta- podial. It is 36 mm. long, and the distal end was transversely cut to a depth of 4 mm. and snapped off. Another specimen appears to be the proximal end of an eagle humerus. At a point below the humeral head, where the bone is oval, it was transversely cut and the end snapped off. Neither of these latter two bones was otherwise modified. Cancellous tissue abraders (3 specimens) : These oval pieces of cancellous bone have rounded surfaces which suggest they were used as abraders. They do not seem to be “paint brushes” since they retain no paint in the interstices. Each seems to be made from a femoral head; they are 35 to 40 mm. in diameter (fig. 24, e). 119 DEMERY SITE—WOOLWORTH AND WOOD Riv. Bas. Sur. No. 34] Pap. j, Pottery c, Miniature bone awl. 4, Bone tube. a-b, Antler bracelets. g-h, ‘‘Rib-edge” awls. f, Split rib awl. k, Bird bone awl d-e, Fishhooks. Ficure 23.—Artifacts of antler and bone. modeling tool. 120 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 J Ficure 24.—Artifacts of bone and shell. a, Pendant. 6, Modified shell. c, Shell “face,” d, Shell disk bead. ¢, Cancellous tissue abrader. *j, Bone awls, RNS ga] DEMERY SITE—WOOLWORTH AND WOOD 121 Scored ribs (4. specimens) : These objects were made by cutting 8 to 16 transverse notches, 3 to 10 mm. apart, across the convex surface of large ribs. All of these objects seem to be broken, since both ends are fragmentary. They range from 95 to 253 mm. in length (fig. 22, a). Avwils (79 specimens) : The awls from Demery generally have sharp, tapering points, although some of them are rather blunt. Each of the tools with points sharp enough to have served as hide perforators is classed as anawl. Theclassification of these implements follows that of Kidder (1932), with a few necessary modifications as required by the sample. The awls are divided into six groups, based on the material from which they are derived. The amount of work done to bring them to their final form is the basis for the following classification. Mamimarleloney WONeS 2.6 Sete A toe pepe alls nee he ot fe eed 29 Head of bone left intact except for splitting (24). Distal ends of metapodials (18). Proximal ends of metapodials (6). Head of bone partially worked down (2). Circular, tapered awls (2). Splinter awl (1). SHO EUP ENTE! |p DON Os ee eee 2 ee Se DS oe anne eee SOs ey seneeT 7) 5 Neuralyspine awis (Tip sed ee!) 2226. 2S a es Be ee 39 Scapula: SPHNtCTSs. eee nee ee ee een Se SP a, Se 2 BIT elONE? DONES See He RL Ae eta ie heat ioe SE Ark aad Tk ES eb ee al. Miniature /awls’ (source: uncertain )it-suot 282 rie es eter ore sed 3 ROT ETN IEP (ete SE i SN ee ae TAS eee eer SRly ee re ye 79 Long bone: head of bone left intact except for splitting (24 speci- mens) .—AII of the specimens in this class were made from the meta- podials of deer or antelope. They were split by longitudinally sawing the posterior side of the bone and by wedging apart the diastema in the proximal end, and using the resulting splinters, with the ends as butts. A complete metapodial was 237 mm. long, and the longest complete awl is 156 mm. long, revealing that the longest awls were about two-thirds the length of the bone. The butts were rounded and polished by use. Eighteen awls were made from the distal end of the metapodial (fig. 24, 2). An unfinished awl is 190 mm. long, but finished speci- mens are 66 to 156 mm. long. The longer awls are slender and evenly tapered; the shorter ones, thick and stubby, probably from repeated sharpening. Six awls are from the proximal end of the bone. The longest of these is 97 mm. long; the shortest, 69 mm. long; they were highly polished. One of them is incised (fig. 24, 7). Long bone: head of bone partially worked down (2 specimens) .— 122 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The two specimens of this class, made from the proximal end of metapodials, are of similar size and proportions. They are 98 and 101 mm. long, with long tapering shafts. The butts are smooth and rounded, having been worked so that articulating facets and project- ing bone were removed, leaving a smooth, swelling grip (fig. 24, ). Long bone: circular, tapered awls (2 specimens).—These two awls, made from sections of long bones, are 145 and 225 mm. long. There are faint traces of cancellous tissue along one side of the larger awl. The butt of the shorter awl is broken (fig. 24, 7) ; the butt of the other is rounded. Long bone: splinter awl (1 specimen) .—The single aw] of this class consists of a splinter of long bone which was sharpened on one pointed end. It is smoothed on all surfaces except the butt. Length, 88 mm. (fig. 24, f). Split mammal rib awls (5 specimens).—The five complete speci- mens in this class were made from split bison ribs, and are 80 to 178 mm. long. They have sharp, tapering points; most of their surfaces are smoothed, although only one of them has a smooth, rounded butt (fig. 28, f). Neural spine awls (39 specimens) .—Specimens of this class, usually called “rib edge” awls, were almost certainly made from sections cut from the neural spine of bison thoracic vertebrae (see Wedel, 1955, pp. 119-120). In some cases, the edge is ground down so that the can- cellous tissue is nearly gone; in others, part of it remains. They are usually triangular in section and the butts are rounded or faceted. A few of them are long and slender, tapering evenly to form a slender shaft (fig. 23, 2), but more often they are short and stubby (fig. 23, 7). Scapula splinter awils (2 specimens).—Fragments of bison scapula spines, as much as 40 mm. long, are fashioned into awls by grinding one end smooth to serve as the butt, and the other into a sharp point. Bird long bone awl (1 specimen).—The single awl of this class is made from what appears to be a radius, with one end brought to a sharp, stubby point. Length, 129 mm. (fig. 23, /). Miniature awls (8 specimens).—These awls are made from uniden- tifiable, thin-walled bones. Two of them are slender and tapered, with sharp points. The third awl has nearly parallel edges and a blunt, stubby point (fig. 23,c). They are 32 to 62 mm. long, and 5 mm. wide. Serrated rib tip (1 specimen) : One end of a large rib, 230 mm. long, is serrated; the bone adjacent to the serrations is polished and beveled from wear (fig. 22, 5). Pottery modeling tools (5 specimens) : These five incomplete items were probably made from bison ribs. Four of them have one convex and one broken end; the fifth one is Riv. Bas. Sur. Pap. No. 34]. DEMERY SITE—WOOLWORTH AND WOOD 123 fragmentary. Surfaces are smooth but they are not polished. The original lengths of these tools exceeded 68 to 160 mm. (fig. 23, 7). Ulna chisel (1 specimen) : The proximal end of a bison ulna and part of the shaft were used for this tool, which is 195 mm. long. The humeral articulating facet islacking. The proximal end hasa chisel-shaped edge which is turned at about a 45-degree angle from the flat surface of the bone (fig. 20, e). Ulna picks (6 specimens) : The proximal ends of right bison ulnae have rounded and smoothed tips, perhaps indicating use as digging tools. Three of them are cut or hacked on the posterior surfaces below the distal ends. One of the latter tools has a smooth anterior surface below the distal end; the other two show signs of hacking. This modification probably had a function in hafting these tools (fig. 20, d). Bison radius picks (4 specimens) : These tools were made by cutting away part of the shaft of a bison radius and sharpening the midsection to a chisellike edge. Three of the four specimens have oval holes 25 to 40 mm. in diameter extending longitudinally through the articulating surfaces into the marrow cavity. They range from 247 to 180 mm. in length (fig. 20, f). Digging tools (6 specimens) : This residual category includes tools with polished, blunted points that are not readily identifiable as to function. The blunted to pointed, polished tips suggest that they may have been used as digging tools. Portions of spinous processes from bison scapula were used for three of them, which are 120 to 310 mm. long. One of them is from the medial edge of a left bison scapula; it is 170 mm. long. One tool is part of a bison rib, 134 mm. long, and a final specimen is part of a heavy long bone, 180 mm. long. Pendant (1 specimen) : The rib of a small mammal has been cut into a segment 85 mm. long; it tapers from 9 mm. in width at one end to 8 mm. at the other. Both ends are convex; the smaller end has two grooves, by means of which it may have been suspended (fig. 24, a). ANTLER ARTIFACTS Cylinders (11 specimens) : These objects, sometimes called “tapping tools,” were made from the proximal ends of mule deer antlers by cutting around the antler to a depth of 4 to 6 mm. and snapping the section off, about 40 mm. below the lowest tine. The attaching burrs are rounded and reduced ; 124 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 the cut, distal ends are rounded and smoothed. Lengths range from 110 to 150 mm., with diameters from 20 to40 mm. Only one of them is smooth and polished (fig. 22, f). Tine flakers (15 specimens) : These tines were cut or broken from the racks of deer, and range in length from 25 to 120 mm. Seven of them consist simply of tips broken from a tine. Thirteen tips have beveled ends or the ends are striated in such a manner that they may have served as a flaking or knapping tool. Bracelets (8 specimens) : A small piece of incised and pierced plate antler is probably part of a bracelet or some similar ornament (fig. 23, a). Two pieces of long, thin, pierced antler may also be bracelets. They were made from thin strips of the compact outer layer of antler; the ends were perforated by a tapered drill (fig. 23, 5). SHELL ARTIFACTS Miniature shell “face” (1 specimen) : This object was made from a thin piece of mollusk shell, and is similar in some respects to large gorgets found in North Dakota mounds (Howard, 1953) and elsewhere; differences seem to be largely a function of the much smaller scale of this object. It is 20 mm. wide and 26 mm. high (fig. 24, c). Disk beads (2 specimens) : One disk bead is 12 mm. wide and 5 mm. thick, with a conical hole drilled from one side (fig. 24,d). A roughly circular piece of mollusk shell, 18 mm. wide and 1 mm. thick, may be an unfinished bead. Scrapers (23 specimens) : These tools are mollusk shells of a size and form that suggest they were used as scrapers, with the lip of the shell acting as the scraping edge. The shell edges are rounded or beveled, some of them having been worn into a straight edge. Five of them were purposefully shaped into triangular forms (fig. 22, d), and three of these are so worn that the hinges are now lacking. Two small, roughly triangular pieces of shell, 30X40 mm. and 20X82 mm., have smoothly worn edges (fig. 24, 5), and one shell, retaining the hinge, has a nearly flat scraping edge bearing two shallow serrations (fig. 22, e). PERISHABLE REMAINS The charred remains of several food plants were in the bell-shaped pits in Houses 3 and 4. The most common of these was corn: kernels, Riv. Bas. Sur. ep NS Sty «~DEMERY SITE—WOOLWORTH AND WOOD 125 pieces of cobs, roots, stalks, husks, and leaves were especially plentiful in Feature 65 in House 3. A fragment of braided cornhusk from Feature 65 suggests the treatment of ripened corn in which the ears were husked and braided into long strings. The kernels from Feature 67 in House 3 resemble ethnological specimens which were picked green, roasted on the cob, shelled and dried; those from Feature 65 seem to have been brought to maturity. The size and the internal structure of a bean from Feature 67 resembles red beans of the variety collected by Oscar H. Will on the Fort Berthold Reservation in North Dakota (descendants of which are in the collections of the State His- torical Society). The charred, peeled roots of Pomme blanche, or Tipsina (Psoralea esculenta Marsh) were in Feature 65 in House 3, and in Feature 94 of House 4. Gilmore (1919, p. 92) states that these roots were dug in June or in early July. A few items of material culture were also preserved by charring. Perhaps the most significant and interesting of these is a piece of basketry (fig. 20, a). The elements are woven under-three-over- three, creating a diagonal pattern known as twilled plaiting. Gil- more, in discussing Arikara baskets made from the inner bark of black willow and boxelder, illustrates two baskets which were apparently woven in the same manner as the Demery fragment (1925, pp. 89-95, figs. 41-42). A piece of peeled wood underlying the basketry may be part of a post foundation; if this is the case, the original basket was about the size of the sdtwa, a large Arikara work basket that stood knee high (Gilmore, 1925, p. 94). The elements in the Demery speci- men are similar in size and form to those in Arikara workbaskets in the State Historical Society collections. A piece of wood, rectangular in cross section, is cut and slotted in the same manner as a bone knife handle, and is probably correctly identified as a wooden knife handle. A second piece of wood, rounded on one end and bearing a deep notch, is from the same pit. The growth rings are identical in both cross sections, and the pieces obviously belong together: the object is tentatively restored (fig. 20,.¢):. A final perishable specimen is a fragmentary buckskin bag from Feature 65. It was originally 150 mm. or more long and 100 mm. or more deep. It was made by folding a rectangular piece of hide back upon itself, piercing the edges, and lacing it together. Since the top is now gone there is no way to determine how it was closed; since it was found lying flat, it was probably closed by a flap rather than by a drawstring. It was empty when found save for a few ounces of sterile earth, and the surface of the bag indicated that it carried no decoration that involved piercing the hide. 126 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 FAUNAL REMAINS The following species have been identified among the artifacts and the food refuse at Demery : Animal: Number of Artiodactyls: individuals TBI SON Se is Ya Se a ae ed ee 95 Meer and antelope. == 2 2 22 ee eee 5 DO pn a las A mela Ee oat MAS a ee Rk ee Bgl ot al Carnivores: Ganids: 22) Stew i. 9 ee eee Le A ee ee a 12 Bad ger yest fs ea. tale at Be Cores ts saen pees) bE ee ee ee 1 Rodents: RAD Dib eseres, Locks a bi iy, ult. 4 eke, AR a, kee 1 Grounds SQUMTC let’ 20a fn ea Bo Se ee eee 2 Fish and reptiles: Cath tn tines Fee” Eek) Ae eee ee SE ee _ Ed ee ee x TM Urtle ys Se: 6 Se he ee ee a a Reo 2 lo Set 4 Birds: White ;nelicane. 2 2 Sob ne Dos Bee ee oe ee ee il WY 0 Wats. aha] oes B98 oes Ia ate pos a awa Rhee Setanta med! le eh a i American rough-ler naw Ko ae eee ee eek a eee 1 Merrucinous rough-les@ hawk. 0 see eee 1 ATT ENT No alee we heeded ie ad fe Ooh SE so ag oo Dt ehh a Orowie 2 22k 2 Set eee mide By Ei! oe oe ou ee ee 1 The avian bones were identified by Mr. Loye Miller, Department of Zoology, University of California at Davis; the bison, fish, and turtles, by Wood; and the remaining animals by Dr. J. Arnold Shot- well, Museum of Natural History, University of Oregon. Comments on the bison bone from Demery, Paul Brave, and Huff have appeared elsewhere (Wood, 1962 b). The remains of other species were too small for an analysis of butchering technique. DISCUSSION STRUCTURES The houses at Demery stand in rather sharp contrast with those of other circular earth lodge village complexes in the Middle Missouri area. The lack of vertical wall posts along the house shoulder implies a style of roofing the dwellings that differs from the technique most commonly illustrated in the historic Mandan, Hidatsa, and Arikara earth lodges in this same general area. The Demery lodges are small, ranging from about 20 to 30 feet in diameter. They are circular to oval, with the long axis of the house in line with the orientation of the entrance. The entrances, facing the southwest, overlooked the broad expanse of river terrace to the south- Riv. Bas. Sur. TOR Pap. No. 34] DEMERY SITE—WOOLWORTH AND WOOD 197 west, rather than the Missouri River or John Grass Creek. The houses were built in pits dug about a foot and a half into the ground, with shallow, dish-shaped floors. A central, unlined basin-shaped fire- place is in the house center, around which are four roof supports set in the form of a square with the posts oriented to the four cardinal directions. There are no postholes on the edge of the house floor; the walls and roof were probably composed of poles set along the edge of the house pit and leaned in against stringers resting on the center posts. Supporting evidence for this inference was observed in the charred timbers in House 4, which extend up to the shoulder of the house pit, where they were removed by the road patrol (pl. 8, 0). House 2 presents the simplest floor plan; the other three houses fully excavated were rebuilt one or more times, and the posthole patterns are therefore more complex. All of the houses were destroyed by fire, since there were charred beams on the floors, with a film of ash cover- ing the last floor level. The entrances were usually elongated basin-shaped depressions 4 to 7 feet long, 2.5 to 4 feet wide, and 0.2 to 0.5 foot deep. Both sides of these depressions were lined with small postholes that formed the passage walls; in several of the houses these posts continue into, or originate in, the house floor. Bell-shaped pits were most common along the house walls; in six instances, they undercut the house shoulder, and their mouths were bisected by the house wall. A variety of irregular and basin-shaped pits were also in the house floor, as well as a few auxiliary fireplaces. From the foregoing description, we may infer that the houses were nearly conical in form, with a covered entryway projecting from the southwest side of the structure. Evidence from House 4 suggests that the roof poles were covered with grass over twigs, and the mottled and disturbed nature of the soil overlying the floor seems indicative of an earth cover over the grass. The house fill, that is, resembles that of structures known to have been earth covered at Like-a-Fishhook Village. The type of house just described is clearly analogous to the “eagle- trapping lodge” of the historic Mandan, as reported by Bowers (1950, p. 282, fig. 25). The same sort of structure has also been described for the Hidatsa by Wilson (1934, pp. 405-409, 411-415, fig. 40). Such dwellings conform in all major details with the type of house inferred from the floor plans of the Demery houses, although the structures at Demery are larger than those described by Bowers and Wilson, since they served as the principal dwelling type, rather than as an ad- junct to it. The lack of wall postholes was also noted in houses at the Spotted Bear site in central South Dakota (Hurt, 1954, pp. 4-8, figs. 6-10). 128 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The house walls at Demery were distinctly visible as a line separat- ing the mixed earth of the house fill from the undisturbed native soil outside the house. The irregular nature of the house walls as mapped is a reflection of the actual situation at the site; the oval shape of the house floor area in Houses 1 and 3 may have resulted from rebuilding, but the flattened arc on the southeast side of House 4 may have some other explanation. The house with the simplest floor plan, House 2, was not rebuilt nor unduly complicated, and here the floor was roughly circular. The source for the innovation of building houses without vertical wall posts at Demery is unknown. These houses stand, therefore, as an interesting variation from a more nearly uni- versal mode of construction. POTTERY The ceramics from Demery are varied and complex. Fifteen groups of rim sherds are described that probably relate to a single, major occupation termed the Demery Component. Two pottery types, Riggs Straight Rim and Riggs Cross-Hatched Rim, relate to an earlier occupation by a group which left the remains designated as the Thomas Riggs Focus Component. The pottery of the major occupation is characteristically very thin and quite hard, and vessels are of excellent quality. The largest re- stored vessel (pl. 14, 6) seems to be typical in form and execution. The malleating, shaping, and firing of this vessel required extremely fine control of the techniques of pottery making. This vessel weighs 714 pounds, and even larger vessels are represented among the sherds. In contrast, the extremely thick sherds of Riggs Straight Rim (pl. 13, 1), with their contorted cores and Jess compact paste scarcely approach the pottery of the Demery Component in skill of manufacture. The thinness of the Demery pottery is a characteristic feature of the ceramics, a feature shared by sites of the La Roche Focus and related foci farther south along the Missouri River. The Demery Component pottery was probably made by building up the vessel by lump modeling, and shaping the walls by malleating it with a grooved paddle or, more rarely, by a cord-wrapped paddle. The six vessels represented by the Group 5 pottery are uniformly and conspicuously cord-roughened from the lip to the base. One vessel of Group 3 and two rims of Group 9 carry nearly obliterated markings that suggest they were originally cord-roughened. Only one-half of 1 percent of the Demery pottery was so treated; the remainder was simple-stamped, or so smoothed that the original surface finish is indistinguishable. There are 1,356 vessels attributed to the occupation by the Demery Component. The range of variation in rim form, illustrated in fig. Riv. Bas. Sur. es Riv. Bas. Su DEMERY SITE—WOOLWORTH AND WOOD 129 13, includes both flaring rims and recurved (“S-shaped” and “col- lared”’) rims, as well as a few residual, unclassified examples. Highty percent of the rims are flared or straight, with recurved rims of all varieties comprising the remaining 20 percent. There is more varia- tion in design on the recurved rims than on the flaring rims, largely because the cord-impressed designs occur nearly exclusively on the recurved rims: cord impressions occur on flared rims only on the four rims described as Examples A—D. Decoration frequencies are as fol- lows: horizontal incising, 44.8 percent; oblique incising, 3.7 percent ; cord-impressing, 6.1 percent; and plain or indented 45.4 percent. The affiliations of pottery groups 1 to 15 are not susceptible to ready generalizations. Typological considerations intimated, even before excavation, that the site was heterogeneous, and that it contained a mixture of ceramic traits typical of two geographically separated and culturally distinct groups. While superposition did establish the fact that 2 pottery types of the Thomas Riggs Focus—Riggs Cross- Hatched and Riggs Straight Rim—predate the major occupation of the site, the remaining 15 groups still convey an impression of hetero- geneity, yet they appear to be characteristic of the major occupation by the Demery Component. Three of these fifteen groups are tentatively identified as types described from other sites. Group 1 resembles the type Talking Crow Straight Rim, as described from sites in central South Dakota (Smith, 1951; Smith and Grange, 1958). Groups 6 and 7 are tentatively identi- fied as the types Fort Rice Cord-Impressed and Fort Rice Trailed, as described from the Huff site (Wood, MS. a) of the Huff Focus. Many of the other types from Demery, including Groups 2-3 and 8, seem to be related to pottery from central South Dakota, including that of the La Roche and Shannon Foci, and the Arzberger site. Ceramic- ally Demery is “transitional” between sites of the Chouteau Aspect and those of the Huff Focus. This is not to say that Demery is simply a fusion of these two complexes, but it seems obvious that its predecessors drew heavily upon sources both to the north (Huff Focus) and to the south (Chouteau Aspect) for its roster of material culture. The Demery site is on the west bank of the Missouri River immedi- ately south of the North and South Dakota boundary. Village sites are plentiful along both banks of the Missouri River to the north and south, but collections from nearby sites yield no pottery that suggests they are closely related to Demery. The material from the site cannot be correlated with any known site within the limits of North Dakota. Demery may be regarded as a site-unit intrusion (Willey, 1956, pp. 9-11) into the area, since it appears as a distinct complex in an area 130 BUREAU OF AMERICAN ETHNOLOGY [{Bull. 189 previously dominated by the Thomas Riggs and Huff Foci (Wood, MS. b) and is not closely related to either of these complexes. There are no cultural predecessors for Demery to the north, but there are no complexes yet described to the south which provide sources for most of the traits that set Demery apart from the Thomas Riggs and Huff Foci. The flared-rimmed pottery at Demery carrying hori- zontally incised lines (Group 2) is related to pottery in sites down- stream along the Missouri River in South Dakota and Nebraska. In 1949, Paul Cooper described a pottery complex along the Missouri River in South Dakota which he called Category B (1949, pp. 303- 806), and demonstrated the wide geographic and temporal distribution of the rim design occurring on this pottery. The importance and the complexity of Category B are being revealed as the excavation and analysis of the many sites in the Fort Randall and Oahe Reservoirs in South Dakota have proceeded. A number of sites and foci have been described that are related to this complex and to Demery, includ- ing the Arzberger site (Spaulding, 1956) ; the La Roche and Scalp Creek sites, assigned to the La Roche Focus by Hurt (1952; see also Meleen, 1948) ; the Myers site (Hoard, 1949) ; the Shannon Focus, as represented at the Spain site (Smith and Grange, 1958) ; the Akaska Focus, as represented at the Swan Creek and the Payne sites (Hurt, 1957; Wilmeth, 1958) ; and sites of the Redbird Focus, north central Nebraska (Wood, MS., 1956). This brief review of related sites makes it evident that the primary orientation of Demery is to the south (see map 6) ; only a few traits at Demery indicate relationships to the north. Arzberger appears to predate sites of the La Roche Focus and prob- ably is at least in part ancestral to that complex. The pottery from Arzberger is complex and heterogeneous, but his analysis of the pottery modes at the site led Spaulding (1956, pp. 111-168) to the con- clusion that there is only one occupation represented in the remains. The situation at Demery is comparable in that there are a great many varieties of pottery, but the majority of excavated material is at- tributed to a single occupation. The pottery at Arzberger was simple- stamped, cord-roughened, or check-stamped. Al of these techniques occur at Demery, but only simple-stamping was of any importance: only one-half of 1 percent of the pottery was cord-roughened, and there was a single check-stamped body sherd. The Arzberger pottery was divided into two groups, the Arzberger Group (collared rims) and the Hughes Group (straight or outflaring rims). Certain collared rims of the type Arzberger Horizontal In- cised are similar to the horizontally incised rims from Demery with recurved rims and a pinched or punctated lower border (Group 8), and the collared Arzberger Opposed Diagonal rims are similar to the Riv. Bas. Sur. (aes ee oa DEMERY SITE—-WOOLWORTH AND WOOD 131 recurved rims from Demery decorated with opposed diagonals and finger pinchings or punctates (Group 9). There are even greater cor- respondences between certain of the Demery pottery and the Hughes Group, as the horizontally incised and plain rims of this group are closely analogous to the Group 1, or Talking Crow Straight Rims, and the Group 2 rims from Demery. In brief, there are enough simi- larities in rim form and decorative elements in these sites to indicate participation in a common tradition, if not contact or contempo- raneity. Since the cord-roughened pottery comprises such a small percent- age of the Demery sample, and since cord-roughening seems to be an early trait, we may infer that Demery is later in time than Arzberger. The large number of rims decorated with cord-impressed lines at Demery clearly sets the site apart from Arzberger, since the five cord- impressed rims from Arzberger appear to be trade sherds from a “Middle Mandan” source. The angular and curvilinear “rainbows” on the cord-impressed and incised Demery rims may be duplicated in both the Thomas Riggs and Huff Foci, but it is more likely that Demery acquired these designs from the latter source: the recurved rims carrying these designs at Demery more closely resemble the S-shaped Huff rims than they do the collared Thomas Riggs rims. In sum, Demery appears to postdate Arzberger, for which there are two carbon-14 dates: A.D. 1461 and 1529 (samples M-1126 and M- 1126a, run by the Radiocarbon Laboratory, University of Michigan, from samples provided by the University of Nebraska as part of the Missouri Basin Project Chronology Program). The pottery complex at Demery is most closely related to sites of the Chouteau Aspect (Stephenson, 1954). Among the foci assigned to this aspect are the La Roche (Hurt, 1952), the Shannon (Smith and Grange, 1958), and the Akaska (Hurt, 1957). The outstanding: ce- ramic trait of this aspect is the horizontally incised design applied to straight to outflaring rims. The Group 2 rims from Demery, which comprise 35.1 percent of the site sample, carry this rim design in its most characteristic form. The Demery pottery most closely resembles the type Grey Cloud Horizontal-Incised at the Spain site, which has protruded T- or L-shaped lips that are lacking in Demery. The type Wheeler Horizontal-Incised from La Roche Focus sites (Hurt, 1952, p. 76) similarly differs from the Demery pottery in having protruded lips. The type Nordvold Horizontal-Incised, from the Akaska Focus Swan Creek and Payne sites (Hurt, 1957, pp. 44-45; Wilmeth, 1958, p. 5) has brushed necks, a trait wholly lacking in the Demery sample. The Group 1 sherds at Demery are tentatively identified as Talking Crow Straight Rim, a common type in sites of the Chouteau Aspect (Smith and Grange, 1958). At Demery, this type comprises 35.6 661-932—64——10 132 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 percent of the rims; together with Group 2, these groups make up 70.7 percent of the site sample. A number of the minority types at Demery, included in the remain- ing 29.3 percent of the rim groups, are comparable to minority types in Chouteau Aspect sites. The Demery Group 38 rims, for example, resemble the type Iona Diagonal-Incised Rim from the Spain site, al- though the Demery specimens lack the rim protrusion typical of Iona Ware. Again, the Group 5 rims resemble in a general way some of the Campbell Creek types as defined at the Talking Crow site (Smith, 1951, pp. 87-39). Groups 8 and 9 resemble some of the pottery from Arzberger (Spaulding, 1956, pp. 189-141), and the types Wheeler Horizontal-Incised and Wheeler Incised-Triangle from the Scalp Creek site (Hurt, 1952, pp. 75-76). In none of these instances do the groups from Demery and the types from the other sites approach identity. These correspondences, together with the presence of the type Talking Crow Straight Rim, are, however, indicative of the cul- tural affiliation of Demery with certain of the South Dakota sites and foci. OTHER ARTIFACTS The bone artifacts at Demery are abundant and of good quality. Scapula hoes are common, and except for the removal of the spine and the posterior border, the implements were modified only by the preparation of the broad end for use as a hoe. Other scapula tools include knives made from the thin bone from the scapula center and which resemble the historic “squash knives.” A cleaverlike imple- ment is more distinctive, although its function is unknown: it may have been used as a knife. This latter tool is also known from Arz- berger (Spaulding, 1956, p. 49, pl. 4, 0). The bison radius picks from Demery resemble those from the Dodd site (Lehmer, 1954, p. 65, fig. 30, g) in having a hole through the articulating surface. They differ from those at the Paul Brave site (Wood and Woolworth, 1964, p-. 88) and the Thomas Riggs site (Hurt, 1953, p. 34) since the picks from the latter two sites are longitudinally split. The 79 bone awls from Demery are, for the most part, forms which are common to most sites in the Middle Missouri area. The principal exception is the class commonly termed “rib-edge” awls, but which appear to be made from the neural spine of bison vertebrae (Wedel, 1955, pp. 119-120), rather than from rib edges, as originally suggested by Kidder (1932). These implements, sometimes also called pins or flakers, are lacking at Paul Brave, Thomas Riggs, and Huff, and as far as present evidence is concerned they appear to be lacking also in historic Mandan and Hidatsa sites. Wedel, however, reports them from sites of probable Arikara origin in the vicinity of Mobridge, S. Dak. (1955, pp. 119-120), and Lehmer illustrates them from Riv. Bas. Sur. Pap. No. 34]. DEMERY SITE—WOOLWORTH AND WOOD 133 the Dodd and Phillips Ranch sites, in the vicinity of Pierre, S. Dak. (1954, pp. 65, 110, fig. 31, 7-m). Similar specimens occur in the La Roche site (Meleen, 1948, pl. 3, 8-11), the Spain site (Smith and Grange, 1958, p. 111, pl. 34, m), in sites of the Redbird Focus in north central Nebraska (Wood, 1956), and in the Lower Loup Focus in cen- tral Nebraska (Dunlevy, 1936, p. 197, pl. 18, 6-d). In central South Dakota, the La Roche Focus probably predates A.D. 1600, but the Redbird Focus is estimated to date somewhat later, from about A.D. 1600 to 1700. The known Lower Loup sites contain trade goods and probably date after 1700, although Wedel (1947, p. 155) has suggested that the complex was in existence by about A.D. 1550. On an earlier time level, Spaulding (1956, pp. 56-57, pl. 4, d-e) found them at Arzberger. Wedel (1955, p. 119) has suggested that these tools may be a late time marker in the Central Plains, but in the Middle Missouri area they are present in prehistoric times. They do not occur, how- ever, in sites suspected of affiliation with the historic Mandan or Hidatsa. The serrated fleshing tools made from bison or elk metapodials were lacking at Demery, but there appears to be a substitute in the form of a serrated rib tip. The wear on the working end of this tool is the same as that on the blade of metapodial fleshers. Two antler artifacts are worthy of comment: these are the antler cylinders, or “tapping tools,” and the pierced strip bracelets. The antler cylinders from Demery are identical to those from sites in the Central Plains. Moving from the Demery site to the south, the first instance of these tools appearing is at the Payne site (Wilmeth, 1958, p. 10, fig. 22), and they occur also at the Spotted Bear site (Hurt, 1954, p. 18, fig. 22, VIII), and at the Scalp Creek site (Hurt, 1952, p. 42, fig. 25,9). “Tapping tools” also occur in sites of the Redbird Focus (Wood, MS., 1956), in the Leary Oneota site (Hill and Wedel, 1936, pl. 10, a), and in sites of the Oneota, Nebraska, and Upper Republican aspects in the Central Plains. Short antler cylinders, usually only slightly longer than their maximum diameter, are in the Historical Society collections from the Biesterfeldt Cheyenne site and the On-a-Slant Mandan village, as well as at the Huff site (Wood, MS. a), but they are not known to be present in Thomas Riggs Focus sites. The affiliations of the “tapping tool” are felt to be with Central Plains complexes and with the later sites in the Middle Missouri area. The two pieces of long, thin, pierced antler may be from bracelets. Similar specimens from Paul Brave (Wood and Woolworth, 1964, pp. 45-46, fig. 11, a7), Double Ditch (Will and Spinden, 1906, pl. 36, w-z), and Slant Village are of essentially the same form. We find no record of them south of the vicinity of Mobridge (Baerreis and Dallman, 1961, pp. 316-327, figs. 88-90), and it is possible that the 134 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 affiliation of this particular artifact class is with complexes that shared in the development of the Mandan or were in contact with the Man- dans. As far as we can determine, these items are uniformly made from antler, although they are often erroneously identified as of bone. About half of the mollusk shells from Demery show wear on one or more edges opposite the hinge, and are probably scrapers. A similar use of mollusk shell is reported from the La Roche site (Meleen, 1948, p. 18, pl. 4, 7-6), and from the Arzberger site (Spaulding, 1956, p. 59, pl. 5, 7-¢). Shell scrapers also appear in the Huff site (Wood, MS. a), and there are 51 specimens from the Biesterfeldt site. A carved shell “face” is similar in many respects to large gorgets recovered in mounds in North Dakota and elsewhere (Howard, 1953). Differences seem to be largely a function of the small scale of the Demery specimen. This is the only Demery artifact which appears to show a “Southern Cult” motif; the oblique incised lines between the raised nose and the eyes seem to be a simplified rendition of the “weeping eye” motif executed on Southern Cult items. In contrast to the majority of local sites, perishables were rela- tively abundant at Demery. Among the charred food remains are corn, beans, and Tipsina root. Of special interest is the fact that evi- dence for braided corn was apparently present, and some corn kernels seemed to have been cut green, roasted on the cob, and shelled, giving insight into methods of food preparation and preservation. Other perishable objects included a fragment of twilled basketry, possibly from a large work basket, and a slotted knife handle. Leather work is represented by the remains of a small, rectangular buckskin bag. CONCLUSIONS The excavation of the Demery site provides a statement for the northernmost known manifestation of the Chouteau Aspect. The peripheral position of Demery to other sites of this aspect is a factor which in part accounts for the presence of traits that set it apart from other sites of the aspect. The proximity of Demery to another and distinct cultural entity, the Huff Focus, is advanced as an expla- nation for some of these traits. Demery is not closely related to sites in its immediate geographic locality, neither those inferred to be earlier in time nor those that postdate it. Despite its proximity to sites which participated in the cultural stream that culminated in the historic Mandan, there are only a few ties with the prehistoric sites relating to that group. The closer relationships of Demery are to the south, downstream along the Missouri River, although there are details of the architectural pattern, the ceramic complex, and some other traits that set the site apart from the most closely related sites. For this reason the major Riy. Bas. Sur. mo Pap. No. 34] + DEMERY SITE—WOOLWORTH AND WOOD 135 occupation at the site is designated the Demery Component of an un- named focus. The village was built and occupied by a group of people who ap- parently moved north along the Missouri River from some point south of the mouth of the Grand River. These people lived at Demery for some years, probably peacefully, to judge from the lack of forti- fications, and seem to have acquired some artifacts from the nearby and culturally distinct Huff Focus. A lengthy occupation is indicated by the fact that some of the houses were rebuilt by the occupants. In brief, Demery appears to be a site-unit intrusion into Huff Focus territory by a group of people from central South Dakota, who retired to the south after an extended occupation of this peripheral position. The time of this occupation and subsequent dislocation is inferred to have been between about A.D. 1550 and 1650. The presence of a few sherds identified as types found in Thomas Riggs Focus communities indicates an occupation of the Demery site by this focus, previous to the time of the major occupation by the Demery Component. LITERATURE CITED BAERREIS, Davin A., AND DALLMAN, JOHN HB. 1961. Archaeological investigations near Mobridge, South Dakota. Soc. Amer. Archaeol., Archives of Archaeol., No. 14. Univ. Wisconsin. Bowers, ALFRED W. 1950. Mandan social and ceremonial organization. Univ. Chicago. Cooper, Pavun L. 1949. Recent investigations in Fort Randall and Oahe Reservoirs, South Dakota. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 300-310. DuUNLEvy, Marion L. 1936. A comparison of the cultural manifestations of the Burkett and Gray- Wolfe sites. Chap. Nebr. Archaeol., vol. 1, No. 2, pp. 147-247. Lincoln, GrIuMoRE, MELVIN R. 1919. Uses of plants by the Indians of the Missouri River region. 33d Ann. Rep. Bur. Amer. Ethnol. for 1911-12, pp. 43-154. 1925. Arikara basketry. Mus. Amer. Ind., Heye Found., Indian Notes, vol. 2, No.2. New York. HEWEs, GoRDON W. 1949. Pottery from the sites excavated by the 1947 North Dakota field ses- sion. Proc. Fifth Plains Conf. Archeol., Notebook No. 1, Lab. Anthrop., Univ. Nebraska, pp. 58-67. Hum4, A. T., AND WEDEL, WALDO R. 1936. Excavations at the Leary Indian village and burial site, Richardson County, Nebraska. Nebraska Hist. Mag., vol. 17, No. 1, pp. 3-73. Hoarp, LYon J. 1949. Report of the investigation of the Meyer site, Stanley County, South Dakota. South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 2. 136 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Howakrgp, JAMES H. 1953. The Southern Cult in the Northern Plains. Amer. Antiq., vol. 19, No. 2, pp. 130-138. Hort, WESLEY R., JR. 1952. Report of the investigation of the Scalp Creek site, 39GR1, and the Ellis Creek site, 89GR2, Gregory County, South Dakota, 1941, 1951. South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 4. 1958. Report of the investigation of the Thomas Riggs site, 39HU1, Hughes County, South Dakota, 1952. South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 5. 1954. Report of the investigation of the Spotted Bear site, 39HU26, and the Cottonwood site, 389HU43, Hughes County, South Dakota, 1953. South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 6. 1957. Report of the investigation of the Swan Creek site, 389WW7, Wal- worth County, South Dakota, 1954-1956. South Dakota Archaeol. Comm., Archaeol. Stud., Circ. No. 7. Kipper, A. V. 1932. Artifacts of Pecos. Phillips Acad., Southwestern Exped. Pap., No. 6. New Haven. KLEINSASSER, GLENN. 1958. Thomas Riggs pottery types. In Hurt, South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 5, pp. 22-31. LEHMER, DONALD J. 1954. Archeological investigations in the Oahe Dam area, South Dakota, 1950-51. Bur. Amer. Ethnol. Bull. 158, Riv. Bas. Surv. Pap., No. 7. MELEEN, ELMER HF. 1948. A report on the investigation of the La Roche site, Stanley County, South Dakota. Univ. South Dakota Mus., Archaeol. Stud., Circ. No. 5. NEWELL, H. PERRY, AND KrikcerR, ALEX D. 1949. The George C. Davis site, Cherokee County, Texas. Soc. Amer. Archaeol. Mem., No. 16. SMITH, CARLYLE S. 1951. Pottery types from the Talking Crow site, Fort Randall Reservoir, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 3, pp. 32-41. SMITH, CARLYLE S., AND GRANGE, ROBERT T., JR. 1958. The Spain site (89LM301), a winter village in Fort Randall Reser- voir, South Dakota. Bur. Amer. Ethnol. Bull. 169, Riv. Bas. Surv. Pap., No. 11, pp. 79-128. SPAULDING, ALBERT C. 1956. The Arzberger site, Hughes County, South Dakota. Univ. Michigan, Mus. Anthrop. Occas. Contr. No. 16. STEPHENSON, ROBERT L. 1954. Taxonomy and chronology in the Central Plains-Middle Missouri area. Plains Anthrop., No. 1, pp. 15-21. Strone, WILLIAM D. 1940. From history to prehistory in the Northern Great Plains. Smith- sonian Mise. Coll., vol. 98, No. 10. WEDEL, WALDO R. 1947. Culture chronology in the Central Great Plains. Amer. Antiq., vol. 12, No. 3, pp. 148-156. . Bas. . Biv. as. y)~) DEMERY SITE—WOOLWORTH AND WOOD 137 WEDEL, WALDO R—Continued 1955. Archeological materials from the vicinity of Mobridge, South Da- kota. Bur. Amer. Ethnol. Bull. 157, Anthrop. Pap., No. 45, pp. 73-188. WILL, GrorcE F., and Hecker, TuHap. C. 1944. The Upper Missouri River valley aboriginal culture in North Dakota. North Dakota Hist. Quart., vol. 11, Nos. 1-2. WILL, GEoRGE F., and SPINDEN, HERBERT J. 1906. The Mandans; a study of their culture, archaeology, and language. Pap., Peabody Mus. Amer. Archaeol. and Ethnol., Harvard Univ., vol. 3, No. 4. WItey, Gorpon R. 1956. An archeological classification of culture contact situations. In: Robert Wauchope, ed., Seminars in archaeology: 1955. Soc. Amer. Archaeol. Mem., No. 11, pp. 1-30. WILMETH, ROSCOE. 1958. Report of the investigation of the Payne site, 39WW302, Walworth County, South Dakota. South Dakota Archaeol. Comm., Archaeol. Stud. Archaeol., Cire. No. 8. WILSON, GILBERT L. 1934. The Hidatsa earthlodge. Arr. and ed. by Bella Weitzner. Amer. Mus. Nat. Hist., Anthrop. Pap., vol. 33, pt. 5, pp. 341-420. Woop, W. RayMonp. 1955. Pottery types from the Biesterfeldt site, North Dakota. Plains Anthrop., No. 3, pp. 3-12. — The Redbird Focus. MS., M.A. thesis, Dept. Anthrop., Univ. Nebraska, 1956. MS. a. The Huff site, 32M0211, Oahe Reservoir, North Dakota. MS. a sub- mitted to Nat. Park Serv., Reg. 2, on file at the State Hist. Soc. North Dakota, 1961. Bismarck. MS. b. An interpretation of Mandan culture history. Ph. D. dissertation, Dept. Anthrop., Univ. Oregon, 1961. Eugene. 1962 a. Cord-roughened pottery in north central North America. Plains Anthrop., vol. 7, No. 18, pp. 232-236. 1962 b. Notes on the bison bone from the Paul Brave, Huff, and Demery sites (Oahe Reservoir). Plains Anthrop., vol. 7, No. 17, pp. 201-204. Woop, W. RayMonp, and WooLtwortH, ALAN R. 1964. The Paul Brave site, 32SI14, Oahe Reservoir area, North Dakota. Bur. Amer. Ethnol. Bull. 189, Riv. Bas. Surv. Pap. No. 33. RY ager? Pes. vg oe ile ty mas Aner wy Chie ie Moh an wee, : ea lea hae . fie rae wh me) ey a aly sived Y WN ee Spi pees $< gi WAN eer skh File © a vee «et — Aad Muth 008 Onli; ‘epileatubyea ai welds! deaths oop oeteee "Mit dug Pte rivslé mers ar ative SBP Seo nies eRe fata st) “ihe hk alee cis nodal dtvem aoe ; ats a if ee | | | oy Ri cca We eid mith tie sles of be fitte aA Setbh tee wath « fea ne se a, ill? aes earnest A StF jin 2k ; ATA wittet Vignette” iin Sitatioteta ait)’ ary” wiiret men oy ; Aiello eae 2 mre, it’ aoa dé eed. slat) A u ott . pee a pe a aoe ae ei! aes aie Ted nck . ofP te SOA oe re ud .alkt etRie ( io 2 aah rad Y 2a 6F Gesries : MA gitiattt WL airal (thu¥ Holt tome AT Se ee See ORI TO eee as oe > pat a. 22 auf) I Ao HOGA, Agst OF FS Sa we got Pe ae Ot 12 : ‘oa di i aves i IO Wine i rut} 7 a aP ohkl erry nt . r ws wood makt i ade: Mit. BA Art ioctiicsé .2aetS nimi ada actiz- Fee co enoawtl: dt sear thre gun {tbs WA vi Rie ie per ‘Seat vist lal ek “tis BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 7 a, House 2. The view is southwest; excavation is beginning in House 3 in background. b, House 4. The view is southwest, toward the entrance. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 8 a, The entrance passage of House J; the view is southeast. b, Collapsed wall poles on the north wall of House 4. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 9 FAA Pottery rim sherds and handles. e-g, 1-7, Talking Crow Straight Rim; 7-, interior of sherd. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE Horizontally incised rim sherds. Note the stab-and-drag treatment to the left of the handle in 7. BUREAU OF AMERICAN ETHNOLOGY Cord-impressed rim sherds. BULLETIN 18 PLATE 11 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 12 Incised and plain rim sherds. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 13 @a | Miscellaneous rim sherds. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 14 Restored pottery vessels. a, Vessel 1, House 2. 6, Vessel from Feature 93, House 4. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 189 River Basin Surveys Papers, No. 35 Archeological Investigations at the Hosterman Site (39PO7), Oahe Reservoir Area, Potter County, South Dakota, 1956 By CARL F. MILLER 139 661—-932—64——_11 re Reese PAs > : ee 7 - is 3 ; re pee aes se Lat ses P YSGlOP HAGA ORG Ge Mest it ' : ABE ntielivd ey é . 1 Y? ey me re i 2 ata al fare «Fe 4 bites 3 Fiva. eth a raatek : “f : a eae : ‘ bie 5 ; ite r ti See ie hae ae ag ne eat ate ' : - } Pi torees #841 q eo See if a a 7 : j it a y i ity t iol ae ; ibs Reenwed te Teese ‘ louse 2, +, Vesved Pram Fente rons: i ? : i na , u 4 ; ‘ = | CONTENTS PAGE Infroductions=- O22 FO Casey Oi) POT! eet MRVIOAE HANS TS ey XE 145 PEOAV EO RBs FA SEROE YY OL SO IU Ey Od a OSE 146 lv OLE. Se 6 CTT ge 2) ep apes tae ec 9c sei a aa pa AR A pT A EI 148 157 28a | Sas eye coy a aebemile qo ep tad fae iy mie oar i. i ee NE AR ES NN 150 PIG UHRR ee a a ak a a carter ae are 151 DUNG eee ine ce eee ee ee Se ee ee eee od ae eS 154 BGO MT CUA TH ee sia asa. A OR MR oer nee te ee ee 154 1 DENTS 2is Fac an fig me a yg ate LE athena es TR VR 55 ROBE Pile acon en Ae Bee ON BOLI LAY oo ee 156 plauphienog of Dutchering areas. 20 2s 157 Jap ENGST ie OT: Sia es Sable poly) ft eager. ith A ntl popes, Nene faye Bet So 159 PRUNE UR ASST CE 2 3) C2 hae et lla le ma A: A lain AA fee pia enh ip renS a pe Dh S 160 PSUDOL Ye te See ee oe me ele ete eens ee ete tee See 160 Classification of body sherds as to exterior treatment___________ 162 Common.Graita Ol Wares=~. scans Ue ce cae te oe 163 Detailed study ‘of body siterds. 27.52" --2222+—5 8 toe aeee toe 165 PAD MECApIARN G2 a Seeds ee ne oe eee le ee ee ane eee 167 Rim forms and their treatment->_ 2228.4 Se NAS 171 Shauider wrentiiento---8 2. n-- oct aa ne eee ae ee ee 182 PACER een Ree ck wen na tee eee ee eee ee eee 183 Wessel siates.s elite. So. s Boles eee see ee eee ere 186 Painted motterye 3122 28s poe See eee ae eek eee ae 188 Miumnburecviessels: ee Seo te kee ee 189 Exotic vessel of possible Caddoan extraction__________________ 190 Ciny Gawiles.\ 2252022 ih ee eee ee ee ee, eee 191 General statementiro-S. cscs ee fee eee eee ee 192 Artifacts other than pottery = 2. ¢.. 2-2... 2202 Vee eee ee ee 193 Chipped-stone artifacts. ..2 2... -.222....2.. 28, 0OW 2a ee Seen 193 ETOJCCUME POMS ores 2 seas aes See tees GO 193 Mirrobladés:: = sc224c2.02u eget oe ee ee 194 DEFAREIS 22s bk a eee eee ee eg 195 OVGIdyGithees scat aoe es. ae ae eee as Aaa ered 203 DifAes Cropperss soe ee es re ee ee Oe eee te a 203 Knives 252056222 c See eee ee ee es es A 203 Burmbkoumiplementss & 225522532052 eos 205 Stone perforators or gravers..2...........-.--.-.-.-.---- 206 Tamellarfigke (OGis =: 502 so 00 Nk ee ee 207 Ground And pecked stone toolss2 25-22 °2>_ 222 2 ee eee 208 Hematite and other paint substances_________________-___-__- 211 BOuG Behan sa ce ee ee ee ee eC a eee oe 212 Petist Grctropny Sk@Ns no. oe Ae ok ee eee 220 SUOLER Hird sternum c= 2. cen ee to ten ele ER OTIOSO RL eee SIUS 220 Winn kedpiesti rete ere ey ore etn AORN ee SIve 220 PU ek Ap Spt las ee eh eel Mele pa aa Re tates Reale hehe Ada Ree ce dieu 221 CGP partie ere ae ne et ee ee ee oe 222 PuMmiary Had COneiibiONsee= oe = ace e et eek AE TS NTN 222 142 CONTENTS PAGE Pret Nidtocc sees eee see Bee coc eek ase eae 226 Appendix 1. Vegetal remains: =: 2253. Lose 5 eek ea eee 231 Appendix 2, Vaumal semaine. . oo. 2 5 Soa see eae eee 233 Appendts. a; Insect remiains- = = 22. ./> See eee eee eee ee 238 Appendix 4. Tables of measureméntaZ jUJ 4100) .----- -.-------=-52-=5 239 PpemetWre veiled < 52s ee oe eek See ee 8 257 Appendix 5. Human skeletal material from the vicinity of the Hosterman site (39P07), Oahe Reservoir, South Dakota, by William M. Bass__._. 259 ILLUSTRATIONS PLATES (All plates follow page 266) 15. Vertical and S-shaped rim types, Hosterman site. 16. Rim types, showing typical forms and treatments. 17. Typical rims with and without handles. 18. Rim types, Hosterman site. 19. Cord-impressed rim treatments, Hosterman site. 20. Incised rim sherds, Hosterman site. 21. Double spouted miniature vessel, unusual to the Plains. 22. Group 1 and Group 2 projectile types, Hosterman site. 23. Group 8, Group 4, and notched projectile types. 24. Microblades and lamellar flake knives. 25. Scraper types, Hosterman site. 26. Scraper types, Hosterman site. 27. End and side scraper types. 28. Side scrapers and burin form, Hosterman site. 29. Knife forms and biface ovates. 30. Flake knives and side scrapers; grooved ax and maul. 31. Graver forms, Hosterman site. 82. Lamellar flake tools. 83. Cut bone sections, worked antler, bone tubes, and scapula hoes. 34. Scapula tools; fleshers and sickles. 35. Scapula knives and fleshing tools. 36. Scapula cleaver, pentagonal-shaped tool, hafted knives, and antler tools. 87. Bone awl types and miscellaneous worked-bone objects. 88. Fetish or trophy skull of Vulpes velo. TEXT FIGURES PAGE Snnthe Hosterman site, B9PO7 2... 220 ee ee (facing) 147 Zo. ,APesture 22, PLOSteTMIah SifC. at ee 148 27. Cross section of the fortification ditch, Hosterman site_____----.---- 149 28. Detail of stockade post arrangement, Hosterman site__________----- 150 Jo qheature 34, EMOSherMan sive. 2 522-2522 2 ~ Ses oe ee 155 S0-qesture 32, Hosterman site. .-2 2. 6 ee 155 31. Origin of various shaped pits, Hosterman site___-_-------------+--=-- 157 32.. Landmarks on pottery vessel, Hosterman site_______--------------- 162 3a0cPottery trend within Hosterman site_ ....-—-=--~-- 0 otennn Sone oo 164 34. Horizgntal parallel incised rim sherd__-_- ~~. -_-_--_-.-=--+=----==2- 166 ILLUSTRATIONS » Vercn rim, profiles, Hosterman-site.. =~. -.2-22.= -2-S2---2_5---4.- . S-shaped rim profiles, Hosterman site_......_.....-..--=.--..--==--- a5 Shape soli MeeOrations.._10 50 30 mes Fee A’ LA 8). d . Barrel-shaped rim, simple-stamped rim sherd_-__------------------- Pipe -pldck Zoval DeCOMAliONB 3M. ge Ses ee oe ee ee ie ) hin treatment and decorative zones...) 25 +.=-) = 2.is2-.-24.--23.24 . Restoration of vertical rimmed, simple-stamped vessel-_--_--___-__--- Shoulder decoration on large jar sherd, Hosterman site_-____--_-_--- . Rim and shoulder area of S-shaped rimmed vessel___-----__-------- . Line-block design on shoulder area, Hosterman site__-_.________-_-_- . Line-block design on shoulder area, Hosterman site_______._-------- . Line diamond design on shoulder area, Hosterman site__------------ » ands types showile pointes Origine .24 4-5 Lo eek Various miniature vessel types, Hosterman site___--_-------------- . Two-spouted vessel, Hosterman site, showing decorative design--_-__-- Ee ICC SC TOPE VCS aoe ee ee i ee ee ea . Burin with arrows indicating position and direction of flaking_--_---~- . Craniostat drawings of adult male (Individual No. 6) from the Hos- terman site, Potter County, South Dakota_.----_-_--_---------- 4 ge ss ; thar >? Res ee 6 sO il ed ag “~ Shao a5 Pogeiehtoaions gi Been te ro lancer ae tA | ee nr ie aon And os bs Fie Beatie Ange i= } ver Disine RARRY gtauidi : Hey ROLE : nr gee aoataiord (ncaa spear , sen etor Lamia gate? Yo aaa blunt “—s- ‘onl RRO See alfa Ra earshot te . «+4 =ttpieeb cs ihe ovis cdemidinol osay baiuede eV ev tien PS cS aan Gane ae Bag om Fer hs VB GNoe. . .eenpyt Tagasoa BUS GL Depa. wuldatteeh throm jwitbng gubieoibat rors pene 45. YrpweoRh ins moe Wo bwivibad stem dlehe to egeiwarh de Emu wyies linstermean leatodaG ti08 mld et rotiod oti aa RA Ny Tae. KEG inmutonedin Towser 2 Se, Rnbiam) vis maerie, Acetecman my x3 Phe es Ble Gat woe vetant ae 5 = afoen 5 ead fom A heal Po oM als ee ero My CoPeis: & KU ee e. ile gi. TAxmeil ar ; 8 is f et tame > (a, Pa ote ren, «ha $i, Thin geravete wid! dracln goiter), Mokbesiuney olde Ry AS ie: Co WIM Petii-+ th eT AT Van Hoe? Bites err atonte =e yuy 2k Ra Ri. Orstaw turtat, U3 wend | Ae vache as One teh a, " mee Ger Lone ee, 2 ? Woe yt, 2 ees a eyo RY. CAN, ral S atal te Reto kel ven dad Aeeing tacee. a ‘ ee ee tl ee mel ft Meee ; LOL th ee Bee SE ions wel Croes bot wiiieeiienowns wo ted bund a\jorre, ae Vay? tn be Tee ts to yelp Cees ry ai ‘% ia OT rere ; 0 fe j pn ee He Sento Bae a ; i , «0 se Lice: wiowhlols.of Wado’ e tiga deb, Homer ote wh Teetes| 4 aveokedm Jpah Mirae Seine Frei a ke We Powter U4, Aakurmiee Bien). BS Pe at gn ir i, eat Oo, Elena ate a i » ora yam ath amped Sei tedete: ot vac ain sepa 2404, Bigot Mii ene aap eee sdoie (2 le wae on poltory nema, Baetermiga dy. |. sine -* yak ; itu e Oyun Hi Gh: Pies oo. Shes ” serene prancee ney h esiasueed Puede feat ire AE: 46d es +e ene | ' ARCHEOLOGICAL INVESTIGATIONS AT THE HOSTERMAN SITE (39P07) IN THE OAHE RES- ERVOIR AREA, POTTER COUNTY, SOUTH DAKOTA, 1956? By Cart F. Mitter INTRODUCTION The Hosterman site, named for John B. Hosterman, owner of the property, is located in sec. 36, T. 119 N., R. 79 W., Potter County, S. Dak., on a high bluff on the east bank overlooking the Missouri River about 214 miles north of Whitlocks Crossing. It is on the western margins of the Coteau du Missouri, “that part of the Missouri Plateau section of the Great Plains province which lies east of the Missouri River.” The name of the Coteau dates back to the days of the French fur traders. The bluff slopes gently toward the Missouri River, then pitches steeply into the river valley about a mile from the present stream. The former stockaded village overlooked the gently sloping plain with a broad view of the valley and the high tableland extending to the east, north, and south. The elevation of the site, 50 or more feet above the flood plain, was advantageous in the defense of the village from attack from the west because of the sharp rise of the bluff on that side. The village site was first surveyed by Dorothy E. Fraser on August 7, 1949. She described it as residing on a high bench 75 feet above river bottom on the north side of a small creek and now a half mile or more from the present stream bed. It isa beautiful specimen of what appears to be a really old site. River probably swept in at the foot of bluff at time of occupation. Owner extremely interested and cooperative. It was he who directed attention to this site, hitherto unknown. [*] Paul L. Cooper revisited the site in September 1949, and he noted that there were 12-15 depressions within (the) ditch which extends from river terrace border to tributary gully at south of site. Area within ditch ca. 320’ by 290’ (paced). 1 Submitted February 1960. 2 An excellent study of the Pleistocene geology of eastern South Dakota, including data on the physiographic environment, climate, soils, and biogeography, was issued by the U.S. Geological Survey in 1955. (See Flint, 1955.) 8 Field notes in files of Missouri Basin Project. 145 146 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 House pits occur outside of the ditch to the east and northeast. Depressions rather shallow and poorly defined. Ditch is also relatively shallow throughout most of its length; varies from ca. 0.5’ to ca. 2.5’ in depth. [*] Early in October 1951, Richard P. Wheeler revisited the site and put down a series of test holes within the palisaded area. He noted: The total area (inside and outside the ditch) measures about 500’ north-south and 400’ east-west, and cover a little over 4% acres. Elevation, 1,570’ to 1,580’. Test pits produced evidence similar to that reported by Dorothy Fraser in 1949; a blanket of sterile silt underlain by two strata of occupation debris separated by a layer of sterile sand .... Windblown silt has all but obscured the house depressions and defense ditch (on the east side of the site): artifact material does not appear on the surface. A small amount of pothunting has occurred and the specimens found in an anthill in 1953 were discarded by relic hunters. [*] I began a partial investigation of the site on June 23, 1956, at which time my crew consisted of: John Anderson, Lincoln, Nebr.; Norman Barka, Chicago, Il].; Tyler Bastian, Stockton, Wis.; Gordon Dentry, Baltimore, Md.; Edwin Floyd, Canyon City, Colo.; Hugh Carl Jones, Provo, Utah; and Dale Osterholt, Platte, S. Dak. Mrs. Ruth Miller, my wife, acted as housekeeper and cook for the crew. EXCAVATIONS Prior to the beginning of the 1956 work, the major part of the site was covered with a heavy sod that had sealed in almost all of the cultural debris deposited by the former occupants, and the area to the north and east of the stockade trench was planted to watermelons. The area under cultivation showed islands of heavy cultural debris and shell material presumably marking refuse dump areas outside of dwellings. Shell deposits were heavier outside the stockaded area than inside. The surface inside of the stockade ditch was marked with a series of depressions of various shapes and sizes. The most prominent depressions were circular in outline and of various sizes, ranging from 20 to 40 feet across; the others were irregular in out- line and of various depths, but none were very distinct or deep. Surrounding the main portion of the village were the remains of a stockade ditch in the form of a crescent, the open ends being to the west and southwest and terminating at the edge of the western bluff. During excavation, it was found that the normal stratigraphy within the site consisted of a relatively distinct sod layer with its attendant root zone that extended to a depth of 0.5 foot. This zone was culturally sterile, and below it for another 1.0 to 2.0 feet was a zone of sterile windblown loess of silt and fine sand. Below the latter was the uppermost layer of cultural debris accumulated during 4 Field notes in files of Missouri Basin Project. “t ROINGIRE RD a Ar ove ‘ 4 ‘gigs i nate yiel he nib . ly" Pa, Aer’? iy Ms ea flv dé Tra bergen pr aeeay: i) iG i, ey me | RAD oT ed | T Pw arin O0Ll-Jos UO - 64 (Pace p. 147) Riv. Bas. Sur. en Pap. No. 35] HOSTERMAN SITE—MILLER 147 the occupation of the site. This layer was of uneven thickness and varied from 0.5 foot to 2.5 feet in thickness, depending upon the con- tour of the former surface and the presence of shallow pits, and below this were patches of a former humus zone that was not entirely de- stroyed by the former occupants of the site. Below this zone was another thin layer of cultural material that was rather spotty in deposition. The area within the protective moat measured 269 feet from the inside, or eastern side, of the stockade trench to the brink of the bluff on the west, and 340 feet north and south. Before starting the excavation a north-south line, y coordinate, was established 9 degrees east of north, and an east-west line, x co- ordinate, was run at right angles to it, bisecting the north-south line at the 160-foot stake dividing the site into four quarters of almost equal size. The site was then staked off in 10-foot squares. Square designation was determined by the number of the stake in the upper right-hand corner of each square. Thus, square 170 is determined by having this number in the upper right-hand corner of the square. The coordinate position of each archeological feature was obtained by scaling # and y coordinates of the approximate center of the site. Each feature was then plotted to scale on a coordinate base chart according to the square in which it appeared. An exploratory trench, 10 feet wide, was started at square 170 and extended for a distance of 100 feet so as to intercept four of the de- pressed areas in order to test their validity as house sites. At the same time other men were placed in squares 180 and 190 (fig. 25). Along the east faces of squares 170, 180, and 190, we found the first indication of cultural material at a depth of 1.0 foot below the root system of the present sterile windblown grass-covered loess. Pro- gressing westward this layer slowly decreased and the overburden of sterile material became noticeably greater. The upper portion of this cultural debris consisted mainly of small fragments of bone. Sherds were derived from the top 0.2 foot of this deposit whereas the number and size of the bone fragments increased as the distance down- ward was increased. Entire bones were found at the base of this level. Artifacts were few. A small unnotched isosceles triangular projectile point and a badly decayed bear’s jaw were recovered from this bone layer. The main purpose of cutting this exploratory trench across the area was to test the four depressions that were thought to be the remains of dwellings. None of them proved to be such. They were either cattle or bison wallows. The house structures uncovered were not manifested in any way by surface indications. Other areas tested where depressions occurred proved this contention. 160 4 é aoStake here 140 120 Test Area Stake here Stockade F1cuRE 25,—The Hosterman site (39P07)-. Measurements for Profile 30 40 50 yc Stake here Village Plan 120 /40 160 180 ao (0 20 ——S Scale in Ft Contour Interval — 5.0° 39 Po 7 148 BUREAU OF AMERICAN ETHNOLOGY [Bull, 189 Feet e Post hole (F) Fire pit (c) Cache pit Ash pit Loess BR Zn Refuse WMA Native soil Hearth Ficure 26.—Feature 22, Hosterman site. Near the west end of the exploratory trench we located and com- pletely uncovered the remains of a circular structure, Feature 22 (fig. 26). Once the house pit was outlined, all of the fill was passed through screens down to within 0.5 foot of the floor. This portion was carefully removed by trowels; the material was screened and placed in separate containers. Even with all this care we recovered very few artifacts from the zone. After the floor had been cleared, the various features, such as post- holes, cache pits, and firepits, were cleaned of their contents. Any- thing found within any of these features was kept separate from that found on the floor. Later the surrounding area was explored in order to locate the entranceway or any other feature that might have been associated with the structure. We not only failed to find any evidence of a passageway, but we found nothing that we could positively asso- ciate with the house structure. Cache pits within the house floor ranged from small shallow depres- sions to fairly large bell-shaped pits of a size sufficient for an average- sized man to squeeze into easily. Most pits contained loose soil; some had an occasional artifact and the larger bell-shaped pit was com- pletely filled with disarticulated bones and very little soil. FORTIFICATION DITCH The fortification ditch began to the northwest of the site and pro- ceeded eastward along the north side, gently curving to the south. Along the east side it continued around to the south to a place where Riv. Bas. Sur. Pap. No. 35]. HOSTERMAN SITE—MILLER 149 it gradually blended into the steep sides of a gully, thus surrounding the major portion of the occupation area. On the surface it appeared as a Shallow trench almost 20 feet wide and of various depths. The greater depths were found in the northern section; the eastern section displayed the shallowest. Selecting a place where the ditch was the deepest, we cut a 10-foot trench not only to cross-section the ditch but to determine if any remains of a stockade still persisted. Six-inch levels were main- tained at all times. The trench walls exposed four distinct humus zones; the uppermost averaged 0.4 foot in thickness and incorporated the grass roots of the present ground cover. Underlying this was a layer of sterile loess 1.2 feet in thickness. Beneath this was a second humus layer 0.5 foot in thickness. This in turn rested upon 1.6 feet of sterile loess, and beneath this was the third humus zone, which averaged 0.3 foot in thickness. Underneath was a very thin zone, 0.2 foot in thickness, of very fine silt, which rested upon a deposit of midden 0.4 foot in thickness. Beneath the midden layer was the earliest humus zone, which measured 0.3 foot in thickness. This humus rested upon a 0.3 foot zone of sterile loess, which in turn covered 1.2 feet of laminated silts deposited in the original bottom of the ditch as constructed by the inhabitants of the site (fig. 27). 10- ze 4 6 & [e) ¢ d hc FEET Figure 27.—Cross section of the fortification ditch, Hosterman site. 150 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 INTERVAL BETWEEN POST NOS. land 2-0,8' 2and3- 1.2 Sand4-1.1' 4and5-0.9' Sond6-I.1' 6and7-/./' 7and6-/.0' 8and9-1.0' 9andio-1.J' l0and 1!-0.8' Mand 12-0.9' !2and 3-0.9' AVERAGE parry MPHUESS 40' W- WOOD IN R- ROCK IN Hie B- BONE IN HOLE UNEXCAVATED UNIT 3 39P07, HOSTERMAN SITE DETAIL OF STOCKADE POST ARRANGEMENT Ficure 28.—Detail of stockade post arrangement, Hosterman site. PALISADE The stockade (fig. 28) appeared on the village side of the ditch alined in two rows; remnants of several of the original posts were still in place. Contact was made with the tops of the postholes at a depth of 2.5 feet beneath the present surface. The average interval between posts was about a foot. The average diameter of the holes was 0.8 foot, but the larger holes measured slightly over 2.0 feet in diameter. In a number were found sections of bones and several stones that were used to wedge the post firmly in place. Whether the Riv. Bas. Sur. Pap. No. 35], HOSTERMAN SITE—MILLER 151 double stockade was erected at the same time or represented an in- terval of time could not be determined, for both lines of posts originated on the same level. It is suspected that the two struc- tures were contemporaneous, since the site appeared to have been occupied for only a short time. HOUSES Houses were of two general shapes: round or circular and rectangu- lar. Apparently the rectangular house, which was partially uncov- ered, was of an earlier period than the circular houses. One circular house was completely excavated, and only a part of another was exposed. Circular structures, Features 22 and 34, are characterized by a cir- cular pit that was excavated at various depths, depending upon the inclination of the builder, into the sterile native soil. We found no signs of any plaster being used on the walls of the pit, and the floor was firmed by trampling. There were four centrally placed roof sup- ports set in a rough square surrounding the centrally placed firepit. Such pits were basin shaped. When these were uncovered we found them completely and firmly packed with a whitish ash that could have resulted from the burning of cottonwood or buffalo chips. Cotton- wood, which does not burn with a high heat, was likely used, since not only the basin itself but the area around the basin was burned a brick red. Scattered over the floor between the central four supports and the edge of the floor were a number of smaller postholes and a number of pits, some of which were used as cache areas and others for midden disposal. Placement of the wall posts showed some varia- tion, which also must have been true for the superstructure. In Feature 22 there were 10 large peripheral posts incorporated within the wall that were more or less evenly spaced around the cir- cumference of the structure. The intervals between these were filled with fairly evenly spaced smaller posts. Precedent indicates that the round houses of the Plains were entered by way of a roofed and walled entranceway constructed at right angles to the line of the two pri- mary roof supports. Such a structure was not found connected with this house. Postholes appeared as either soft spots or darkened circular areas in the house floor. Their diameters varied from 0.25 foot to 2.1 feet and in depth from 0.3 foot to 3.5 feet. In most cases the holes were filled in with a light, fluffy soil, and sometimes the butt end of the former posts still rested therein. In several of the larger postholes were sections of bison bones and ribs or waterworn cobbles that were inserted as wedges to stabilize the posts. Field examination of the wooden fragments indicated that most of the posts were of cotton- 152 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 wood; only an occasional piece of juniper was used. Probably the use of cottonwood was due to the fact that it grew readily at hand in the nearby bottom lands, whereas juniper had to be transported from a greater distance. The general practice probably was to choose the material closer at hand. . Within the floor area we found the same variation as to the place- ment of pits. In Feature 22 there were a number of pits. Some were undercut and bell shaped; others were straight sided, and still others were what we have called pocket caches, for they were small and shal- low. All appeared to have been dug while the house was occupied, for they all originated at floor level. In Feature 34, which was only partially exposed, we found no floor pits outside the centrally placed firepit. Lehmer (1954, p. 31) suggests that other holes in the floor may have contained wooden mortars after the fashion of the historic Arikara and Mandan. These holes were cylindrical, somewhat larger than the largest postholes, and were always located more or less on a line between the firepit and the entrance. They were generally slightly offset toward the entrance from a line connecting the eastern pair of primary roof posts. Of all the structures investigated, none was burned. It would ap- pear that they were abandoned and allowed to decay gradually, leav- ing no tangible evidence as to the arrangement of the elements of their superstructure. On the other hand, Feature 32 was characterized by a rectangular pit of considerable size with rounded corners and outlined with small individual posts more or less uniformly spaced. The west wall was completely excavated, and also parts of the north and south walls. Again, there was no evidence of the use of plaster on the walls of the pit. The floor of native material was firmly packed. Entrance into the structure was gained by a walled and roofed passageway from the west-northwest, overlooking the river and the bottom land. There was no evidence of any antechamber. Several hearth areas were un- covered within the area of the structure. No arrangement of central roof supports could be found. Whether all of these hearths were coeval with the structure could not be determined, for there was some indication that the whole area had been worked rather intensively by the Indians after all traces of the structure had disappeared. There were several large deep refuse and cache pits present in the area. The wall posts, when first found, appeared as small soft circular spots along the periphery of the house floor. When cleaned of their contents they were found to be cylindrical with a rounded base aver- aging 0.3 foot in diameter and 0.9 foot in depth. A number of them did contain some midden material. Be Oar HOSTERMAN SITE—MILLER 153 The arrangement of the posts inside the house and the entranceway would suggest that some sort of baffle or screen must have been put up to shield the occupants from the chilly north winds of winter. While working at the site we noticed that the prevailing winds were out of the north and south rather than from either the east or west. Whether there is a change during the winter months is not known, but there must be some correlation between the wind direction and the placement of the entranceways. Then, too, the Missouri River is to the west of the site. Perhaps these two factors may have determined the placement of entranceways. Circular houses were characterized by being more or less round in outline and sunk into pits that now occur from 1.8 feet to 2.8 feet below the present ground level. There was no sign of plastering on any of the walls, and the floors were firmed simply by trampling the bottom of the pit. In cross-sectioning the floors we found that no surfacing material was brought in to cover the floors. Despite ex- tensive search we failed to find the entranceways into the circular houses, especially in Feature 22. Houses in comparable sites normally have definite passageways leading into them. They were constructed at right angles to the line of wall posts and were walled and roofed with the posts set into well-defined holes. Entrance floors usually rose somewhat at the end away from the house. There was some variation in the number of central roof supports. One of the circular houses had four, which appears to be the conven- tional number in the Plains, and another had five. There was also considerable variation in the construction of the superstructure of the individual house. The number and size of the poles that went into the superstructure could vary considerably in their placement. The main firepits were always located at the center of the house floor and were basin shaped. The floor area around these fire basins nearly always was burned as much as the walls of the firepit. From this con- dition it would appear that no one ever took the time to clean out the pit once a fire was started. Ashes were allowed to accumulate, be- coming firmly packed as time went on. As the basin became filled and fires kept burning over it, the heat of the fire spread over the floor area surrounding the pit, changing the physical structure of the clay floor. Such evidence throws insight into the character of the Indian housewife. She was either too busy with her other chores to take the time to carry out the ashes or it was customary to leave them. Small hearth areas were found scattered around the perimeter of the floor, probably representing individual cooking fires that were built to supplement the heat given off by the central firepit. It would appear that each house furnished shelter for more than one family. 154 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Floor pits were not excessively numerous. Some were bell-shaped, others were cylindrical, while the smaller examples were shallow basin-shaped pits probably dug to hide certain objects. All appear to have been constructed during the life of the house. Some were definite cache pits and others were midden disposal areas. FEATURE 22 Shape: Circular in outline. Dimensions: Diameter, 26.0 feet east-west; 27.0 feet north-south. Depth of pit from present surface, 3.0 feet. Pit walls: Vertical unfaced midden and native soils. Floor: Trampled bottom of the pit. Roof supports: Five primary central posts arranged around the central fire basin. Other large posts were spaced at irregular intervals around the perimeter of the house floor and interspaced with a ring of smaller posts, some of which may have supported a sort of bench or platform as well as afforded storage racks. Entranceway: No such feature could be determined. It was probably thought to have joined the structure from the west or northwest if such a feature ever existed. Firepits: Large centrally located basin filled with compact whitish ash. Slightly to the east-northeast was a smaller firepit containing firmly packed whitish ash. Hearths: Several were scattered around the perimeter of the floor. Floor pits: Five such pits occupied positions within the floor of the structure. Three were bell shaped and filled with some ash, midden material, and a few artifacts. One such pit was filled completely with the disarticulated bones of an antelope. The smaller cache pits con- tained a few projectile points or scrapers. FEATURE 34 Shape: Circular in outline. Dimensions: Diameter, ca. 40.0 feet east-west; 30.0 feet north-south. Depth of pit from present surface, 3.5 feet. Pit walls: Unfaced refuse and native soil. Floor: Trampled bottom of pit. Roof supports: Four primary posts and a row of posts outlining the floor area. Some of the smaller floor posts may have served as supports for benches or platforms as well as storage racks. Entrance passageway: Opening to the west. There were some well-defined post- holes on each side of the entranceway and this sloped slightly upward. Floor pits: One vertical-sided pit just to the north of the entranceway that might have contained a wooden mortar. To the north of this was a small cache pit slightly bell shaped in cross section. To the east of the mortar pit was a small basin-shaped pit containing a few sherds. Firepit: Centrally located basin filled with compact whitish ash that was super- imposed upon an earlier hearth area. Associations: None. Comments: This structure was found to the east of a surface depression and showed no relationship to the depression. Riy. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 155 oO 2 4 6 8 10 Feet e Post hole (F) Fire pit (4) Hearth PRE] «Refuse Mortar hole e o 2 4 6 10 e Feet | e@ @ Post hole / 1 e e Ficure 30,—Feature 32, Hosterman site. FEATURE 32 Shape: Rectangular with rounded corners. Dimensions: 32.0 feet by ? feet. Depth of pit from present surface, 2.0 feet to 2.3 feet. Pit walls: Unfaced refuse and native soil. Postholes: Holes indicate wall posts were rather small, averaging between 0.25 foot to 0.3 foot in diameter. These were very definite in outline but rather shallow, which would seem to indicate that posts leaned: against some sort of support and were not firmly planted into the ground. 661—-932--64——__12 156 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Floor: Trampled bottom of the pit. Roof supports: Several large posts were arranged in the walls which were prob- ably crotched and upon which stringers were placed. No post was found to support the central portion of the roof. Entranceway: A concentration of postholes occurred in the vicinity of the in- side entrance to the house, a fact which may indicate that some sort of screen or baffle was constructed here as a windbreak. . There was an absence of any step down into the house from the entranceway, which opened to the northwest with a slight curve at the opening. Firepit: It is not certain that the hearth uncovered actually belonged with this structure. Floor pits: A number of large refuse and cache pits were found within the confines of the structure; they appeared to be secondary. Associations: None. Comments: The fill was practically sterile within the structure. About 0.15 foot above the floor there occurred a change in color and texture of soil. It appeared to be of humic origin. A few bits of midden were scattered throughout this deposit. Whenever a large cache or midden pit was found there occurred a corresponding break in this stratum, indicating a secondary origin for these features. CACHE PITS In general, cache pits were of two general shapes: bell shaped with narrow mouths and flaring out toward the bottom, and comparatively shallow saucer-shaped ones. Bell-shaped pits varied considerably as to size. Most of these were fairly large and deep but there were a few that were much smaller and relatively shallow. The latter occurred mostly inside structures and the larger ones were scattered throughout the village area and outside of houses. Pits of this shape were found to originate at depths rang- ing from 2.0 feet to 2.6 feet from the present surface. On the other hand, saucer-shaped pits showed greater variation as to overall size and depth. Pits of this shape originated at greater depths, from the present surface, ranging from 3.2 feet to 4.5 feet. Some ap- parently were borrow pits, where the soil was obtained to cover the outsides of their houses. Later these were used as midden disposal areas. Whether these points of origin designated their relative age within the site is suspected as an index but when the contents of both types were compared there was practically very little difference in the gen- eral overall contents. The bell-shaped pits were always richer in cul- tural remains, whereas the saucer-shaped pits were more or less on the lean side. From all appearances the earlier ceramic forms would seem to be confined to the saucer-shaped pits. Plotting the depths of the origins of both types of pits on a chart shows, more clearly, the placement and separation of the types as to depths (fig. 31). Riv. Bas. Sur. = Pap. No. 35] HOSTERMAN SITE—MILLER 157 PRESENT GROUND SURFACE 7777 10 2.0 2.0 wK wy BELL~SHAPED ly 2.5 ie 2.6 2.8 3.0 2 = 32 3.3 SAUCER-SHAPED 3.5 3.5 a0 BASIN= SHAPED ly = 4.5 50 ORIGIN OF VARIOUS SHAPED PITS Ficure 31.—Origin of various shaped pits, Hosterman site. SLAUGHTERING OR BUTCHERING AREAS Four unique features, 12, 13, 14, and 19, distinctive for the Hoster- man site, were uncovered. These consisted of large basin-shaped pits with parts of articulated bison remains resting within. Not all por- tions were articulated. The appearance of the remains would sug- gest either slaughtering or butchering areas; possibly both functions were represented. Being basin shaped, the top of Feature 12 was first picked up at a depth of 3.5 feet from the present surface. Three-tenths of a foot deeper the outline of the pit was distinct. At this depth it measured 8.0 feet in maximum length and 7.0 feet in maximum width. Later the base of the pit was found to be comparatively flat and at a depth of 6.7 feet from the present surface. Inside were not only the articu- lated sections of vertebrae, and individual bones, but the skulls of two bisons. From the arrangement of the bones, it was surmised that this was a slaughtering or butchering area inside the village proper and that animals were brought in, butchered, and the unwanted sections discarded. 158 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The larger and heavier of the two skulls, apparently that of a bull, with complete mandibles still in proper articulative position, had both horns missing. ‘These apparently were severed from the skull by means of heavy blows with a stone maul or some similar object. The fact that the lower jaws were complete and still in proper place would exclude the idea that the tongue, a prized delicacy, was removed by - the people of the Hosterman village site. The other, a badly decayed light skull of either a female or a calf, still retained both of its horns. From the general appearances, it likewise had not been mutilated to secure the brains, tongue, or muzzle. The presence of these two comparatively whole skulls, the articu- lated sections of vertebrae, scapula, as well as individual bones, all within a small confined and constructed area, attest to the fact that these animals were transported into the village from some unknown distance and dressed there. Since there were no artifacts within the pit or closely associated with these remains, we could not determine whether these animals were killed by means of bows and arrows, spears, bison kills, or were drowned in the nearby river and dragged to the village to be butchered. Feature 13 resembled Feature 12 in that the basic characteristics were the same: the comparatively round, shallow, basin-shaped pit that contained not only individual bones but whole sections of animals still in articulative condition. This specially constructed pit orig- inated at a depth of 4.5 feet from the present surface, measured 7.8 feet in maximum length, 6.4 feet in maximum width, with a depth of 2.9 feet. Resting therein were the articulative hindquarters of a single bison together with its caudal vertebrae. The impression was that the flesh of this particular animal was mostly wasted, for most choice cuts are derived from this part of the animal. There is no telling whether the Hostermanites cut away the heavy flesh from the bones, took the kidneys and kidney fat, and left the undisturbed bones to be covered over with clean wind-blown material as we found them. Feature 14 has been labeled as “slaughtering area number 3,” in that it, too, conformed to the generalized pattern established by both Fea- tures 12 and 13. Like the others, the pit was found at a depth of 4.5 feet from the present surface; it was saucer shaped; it measured 6.8 feet in maximum length, 4.8 feet in maximum width, and had a vertical depth of 1.2 feet. Resting therein were sections of vertebral columns, all articulated, and portions of leg bones. Like the former areas this feature was barren of any midden material. Clean, windblown sand surrounded the bones. This was the smallest of the slaughtering areas. Feature 19, the largest of the slaughtering areas, had its origin 4.5 feet below the present ground surface. It measured 10.6 feet in maximum length, 10.0 feet in maximum width, and had a vertical Riy. Bas. Sur. Ba Pap. No. 35] HOSTERMAN SITE—MILLER 159 depth of 2.5 feet. This pit, like Feature 12, contained the skulls of two adult animals, together with individual bones, articulated sections of vertebral columns, and other portions all in proper alinement. Horn cores were intact and the muzzle portions were not complete, a fact which may indicate that portions of the nasal cartilages and upper lips were removed. White (1954, p. 167) surmised that these portions were probably considered a delicacy similar to those of the moose, which the northern Indians converted into a rich stew. From the appearance of the skulls these portions were removed while in the butchering area. Caudal vertebrae of one individual were in place, indicating that the tail was not always removed along with the hide. Superimposed above this feature were a number of random post molds. A sterile layer of loess and sand, 1.3 feet in thickness, sepa- rated the two features. None of the postholes were deep enough to penetrate into the mouth of the slaughtering area. There is a definite time differential between the two. All four of the butchering areas had their origin at a greater depth than any of the other features located. This would indicate that they represented the primary occupancy of the site. Whether this was the nucleus out of which the village was established or was just a hunting campsite could not be determined, since only a very small portion of the site was investigated through excavation. I lean toward the theory that these were probably the remains of an early hunting camp near a place that bison frequented, since there were no-cultural re- mains found in direct association with these features. MIDDEN PITS After a time lapse and at a higher level in the site, we found that the midden pits, which were probably early borrow pits, tended to be saucer shaped and fairly shallow, with the exception of Features 10 and 20. Feature 10 had a vertical depth of 4.1 feet; Feature 20 had a vertical depth of 5.3 feet. Other saucer-shaped pits were: Features 2, 3, 15, 16, 23, 24, 27, 33, and 35. Feature 35 was the shallowest of the series, being only 0.3 foot in depth, and Feature 16 was the deepest, being 2.2 feet in depth. Some were circular in outline, another was quasi-rectangular with rounded base, and one had an irregular outline with walls sloping in toward the center. All contained some midden mixed with loess. Depths from the present surface ranged from 3.2 feet to 3.8 feet (see chart of depths for the various types of pits (fig. 31) ). ' Deviating from the norm were Features 10 and 12. Instead, their contents appeared to have been richly mixed with humus and very loosely inserted into the pits. In other features the pit fill leaned 160 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 toward the compact and had to be troweled out. In Features 10 and 12 the fill could be lifted out by the unaided hand. The rest of the pits originated between 2.0 feet and 2.8 feet. These were bell shaped and probably were initially intended for storage purposes. When no longer needed or used for storage, they were con- verted into midden disposal areas. Their overall sizes varied. Some ~ contained more midden material than others, but in general they fol- lowed a pattern. (See table 1.) TABLE 1.—Pit measurements (feet) and general shapes Depth from | Maximum Minimum Depth of Feature No. surface dene of lene of feature General shape ase ase Bell shape. Saucer shape. ell shape. Vertical wall. Bell shape. Dee a a a a are 0. recor, a, Saucer shape. UND OOO 69 02 69 ND OD RO C9 ATO C0 “TONED EN ° Bell shape. Saucer shape. — oo or ' ' 1 H ' ' i \ ’ \ \ 1 1 ' \ ' ' ' ' ' ym mm 09 C9 G9 CO CD CO CO COC CI CH WO DI NO NNN AAAMWOAAMWWNNNNNWOAROCO SP AV QO GO NI G9 BS 00 09 G9 G2 G9 Cr OV GO NTS OO NI AWWOWMSOCHWNARONAMWDNHRADOARS SRS NOV DOO NOON OWS 09 Ct CO NID NN SCOROSRDOROHODONAMUWHHWRAHO NPN Nose eh, ba _ —_ ARTIFACT MATERIAL POTTERY During the course of the excavations at the Hosterman site numer- ous fragments of pottery, of various sizes, were found. They occurred most abundantly in cache pits, in some quantity in the shallow pits under house floors, in postholes, and in small isolated midden heaps. Only a few were recovered while we were sectioning the fortification ditch that surrounded the site proper. Pottery was practically absent from the surface, since the cultural deposit was rather deeply buried. Occasionally, small bits were found around a few of the gopher holes or where some pot hunter had dug in the past. Practically all of the pottery remains were those from vessels. They exhibited a paste made by mixing the natural local clays with a tempering material of crushed igneous rock; the resulting paste is quite uniform insofar as hardness, color, and firing practices are concerned. Color ranges from buffs and tans through browns to various shades of grays to soot black. Fire clouds are fairly common. Texture is me- Riy. Bas. Sur. oe Pap. No. 38] HOSTERMAN SITE—MILLER 161 dium to coarse, and there is a common tendency to develop internal cleavage planes parallel to the vessel walls that is mainly due to the method of manufacture. All vessels were either modeled or made by using the paddle and anvil method. Insofar as we can determine, all of the vessels were basically jar shaped, globular in outline with ver- tical or S-shaped necks, and medium to large in size. Vertical necks are medium to high; some with slightly everted rims. There are various modifications in the S-shaped variety. Vertical types pre- dominated over the S-shaped variety. A few miniature and unusual forms are present but they constitute only a fractional part of the ceramic complex. Handles, although of various shapes, are not nu- merous. Most of the sherds—since this is a study based solely on sherds rather than on whole vessels—have smoothed exterior surfaces, but there is strong evidence that a grooved or thong-wrapped paddle was used during the last stages of manufacture and later this exterior surface treatment was partially or entirely smoothed over. The paddle stria- tions covering the major part of the exterior surface are vertical or nearly so. Once the exterior surface has been smoothed it may be given a number of different treatments. Necks may be vertically brushed and the upper parts incised; the shoulders may be incised, covered with the original simple stamp, brushed, or even smoothed over and left plain. Lips, upper rims, and shoulders were the favorite portions of vessels to be treated. They may be incised, punctated, impressed, pinched, or left plain. As in other archeological areas of the United States, entirely too much stress has been placed on the ceramics in the Plains. In truth, this trait was only one of the many that made up the economy of these people; however, there are many other traits, though of less perma- nence, that were of greater importance to the makers and users of these forms. Only because of its permanence and durability does pottery overshadow, in quantity, the rest of the cultural remains. Then, too, it does reflect considerable responsiveness to cultural changes, through time and space, in the kinds of decoration used, the shapes of the ves- sels themselves, methods of manufacture, and other features. Wheeler’s (1954, pp. 8-4) definitions for morphological parts of the vessel have been standardized as follows: “The lip is the juncture of the exterior and interior surfaces of the rim; the rim is the section between the orifice and the neck; the neck is the point of maximum constriction, or the point of marked change in trend or direction of the vessel; the shoulder area is the space between the neck and the point of maximum diameter of the vessel.” (See fig. 32.) Before going into the morphological details of the pottery remains, I must say that I am in close agreement with Lehmer (1951) and 162 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Lip Rim collar aa ~——_— Neck Shoulder Body Base Figure 32.—Landmarks on pottery vessel, Hosterman site. Smith (1951) in regard to the methodology used in typing Plains pottery, so ideas similar to theirs need not be repeated here. Like Smith, I, too, was confronted with the problem of associating not only the exterior surface treatment with rim decoration but with lip deco- ration and treatment as well. The Hosterman site pottery description is based on a total of 75,814 body sherds, 3,155 rim sherds, and 5 possibly restorable vessels that were never restored. The pottery was first washed in the field laboratory at Lincoln, Nebr., and the rim sherds were sorted and cataloged. ‘The body sherds were not cataloged individually, as were the rim sherds, but were grouped according to the square, depth, and feature from which they were derived. In making the ceramic analysis, the body sherds were first studied en masse in order to acquire a reasonable working know]- edge of the exterior surface treatments so as to avoid making certain snap judgments based upon misconceptions as to what should be found. Once the body sherds were studied and analyzed they were resacked according to the lot bags. CLASSIFICATION OF BODY SHERDS AS TO EXTERIOR TREATMENT ' All of the body sherds were classified, according to the exterior sur- face treatment, into eight main categories: plain, 42.16 percent; simple-stamped, 40.09 percent; éncised, 13.96 percent; brushed, 2.19 percent ; curvilinear-stamped, 0.84 percent; cord-impressed, 0.22 per- cent; mat-impressed, 0.18 percent; fabric-impressed, 0.12 percent. Riv. Bas. Sur. Bagh Naot HOSTERMAN SITE—MILLER 163 The minor types were: painted, 0.06 percent; punctated, 0.05 percent ; pinched ridged, 0.01 percent; fingernail pinched, 0.01 percent; and corncob-impressed, 0.01 percent. Out of the various types present there are four main wares: plain, simple-stamped, incised, and brushed. Percentages of these were plotted, according to our arbitrary levels within the site, on a com- mon graph to determine their trends. This superimposition of wares reveals that the plain, simple-stamped, and incised follow almost iden- tical advancements and declines throughout their existence, but brushed described a different curve for the first five levels within the site. On the next to the top level (sixth from the bottom) all were in fair agreement, proceeding along similar curves up to the time that the site was abandoned. Brushed had advanced steadily, even though it had a late start, and had gained prominence over the other three, percentagewise, before the sixth level was reached. All of this is shown on the graph (fig. 33). TABLE 2.—Composite data on body sherds, Hosterman site (89P07) Body sherds at indicated level (feet) Exterior treatment 0.0-1.0 | 1.0-1.5 | 1.5-2.0 | 2.0-2.5 | 2.5-3.0 | 3.0-3.5 | 3.5-4.0 | 4.0-4.5 | Total ncisedieeta er sees. ee 113 506 229 408 261 51 3 Sse eee Plaire tse shee Ts Se 293 | 1,987 871] 1,110 378 S01 wos. |e ee ee IBTusNp eens seat cree ae ed 14 104 46 ie eS ae ae ena ee SE (ES S Simple stamped____________ 166 | 1,033 829} 1,225 387 59 10 16))| ete INDI bersee es. see 586 | 3,630] 1,975] 2,770] 1,062 311 13 24) 10,351 Percents soe eee 5.6 35.0 19.0 26.8 10.3 3.0 0.1 0.2 100 Plain INGTIMperee eee ee se a 293 | 1,987 871 | 1,110 378 P01 Ke) ie ee | a 4, 840 IPOnCen basn22=—=5-26 55555 6.1 41.1 18.0 7.8 A eof sSclecescuns 1 Brushed INGUIN DOP 2222s oe es 14 104 46 27 [ial oe eR a ID) | 227 Percenti2as Foe Si 6.1 48.8 20.3 11.9 15.9) |cee ee Soe eee wee ere 100 Incised Nm bers es ee 113 506 229 408 261 51 3 8 1, 649 Pereanitesss == so eee 6.9 30.8 18.1 24.7 15.8 3.0 0.2 0.5 Simple stamped: LTT ee ee 166 1, 033 829 1, 225 387 59 10 16 3, 725 IPerOeH Giese it 2 a he 4.7 2203 32.9 10.3 1.5 0.2 0.4 1 While examining and studying the body sherds, it was determined that manufacture was by modeling rather than by using the coil sys- tem. A microscopical examination of the paste was made at the same time but we could not differentiate the physical characteristics. Once these were determined and described, we felt that there would be no need to repeat them for each of the wares. COMMON TRAITS OF WARES PASTE: Method of manufacture: Probably made by lump modeling using a paddle and anvil instead of coiling. Temper: Grit tempered, ranging from particles that are barely visible up ito those around 3 mm. in diameter; probably crushed granite. Sometimes 164. BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 ARBITRARY LEVELS USED IN THE EXCAVATION OF THE HOSTERMAN SITE 00-10 LO-LS 1.5-2.0 2.0-2.5 2.5-3.0 3.0-3.5 3.5-4.0 4.0-4.5 4.5-5.0 ome =P! A/N eaceme BRUSHED mewomem SIMPLE -STAMPED ecoeees INCISED a_——— COMPOSITE Ficure 33.—Pottery trend within Hosterman site. the larger particles protrude on the surface; this occurs mostly on the interior surfaces where the vessels have been roughly smoothed or brushed. Freshly broken edges have a granular appearance, depending somewhat on the quantity and the aplastic coarseness. Texture: Medium to coarse and at times flaky. Cohesive tendency: There is a tendency for the walls to split into unequal sections owing to the method of manufacture. Structure: Tendency toward lamination. Easily splits parallel to the sur- face, and edges are crumbly. Riv. Bas. Sur. Pap. No. 85]. HOSTERMAN SITE—MILLER 165 Color: Color varies from a light tan, buff, and gray, and often a sooty black. Fire clouds are common on the lighter colored sherds. Cross section may be the same as the surface or it may have a darker center. SURFACE FINISH : In most cases the paddle was used on the exterior surface during manufacture. It was wrapped with some sort of vegetal fiber or gut or was roughly carved. The overall design of parallel grooves was repeated, the application of the paddle imparting grooves and lands. Grooves measure roughly 3 mm. in width by 15 or more mm. in length. The lands are round to round pointed and the grooves are U-shaped in cross section and somewhat rounded in general shape. Usually the stamping is vertical to the lip but there are rare cases when the stamping occurs on a slight diagonal to the lip. The initial stamped design apparently covered the complete exterior surface of the vessel but this was either completely eradicated through sub- sequent smoothing or partially obliterated by brushing. Interior surfaces may be roughly smoothed or brushed. Occasionally one appears to have been semiburnished by having been rubbed with a pebble or some other hard sub- stance. When brushing occurred interiorly it was always applied horizontally to the main axis of the vessel and parallel to the lip. Decoration varies with the component types. Decoration of lips is frequent. Decoration on shoulders consists of series of parallel lines arranged in opposition to each other, in contiguous plats forming a broad continuous band around the vessel. Bases were untreated except for the initial surface treatment. DETAILED STUDY OF BODY SHERDS During the process of manufacture some sort of paddle was applied to the outside of the vessel. It was either carved with a series of shal- low parallel ridges or wrapped with narrow strips of leather, the results of which show up as a series of shallow parallel lands and troughs that cover practically the entire outside of the vessel. Usually these impressions run vertically, but over the base of the vessel they may overlap and crisscross. Such treatment has been described as “simple-stamping.” Wares impressed with a carved paddle had more regular impressions and show a uniformity of depth, whereas the leather wrappings were not as regularly placed and there was some variation in the overall depth and width of the impressions. Plain wares are those whose exterior surfaces were smoothed over completely so as to eradicate any previous surface treatment. Some were subsequently rubbed, sometimes with a small stone, or a similar object, to impart a pseudopolish or burnish. Vessels treated in this way were much smoother than those that were simply smoothed by hand. Decorations of incised elements were confined to various portions of the vessel: the lip, rim, or shoulder area. They were confined to zones around these areas. Decorations that had repetitious elements within the zone have been called “line block” in this report. There is wide variation as to the method of rendition, the width of the ele- ments themselves, and the placement of the designs. Some decorations 166 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Figure 34,—Horizontal parallel incised rim sherd. show that a very sharply pointed instrument was used, resulting in a very narrow line of moderate depth. Others, of a trailing nature, were made by using a wide blunt-bladed implement, resulting in a wide, coarse, shallow, troughlike line showing a poor sense of symmetry and touch. The design was not only carelessly drawn but it was not always complete (fig. 84). This kind has been called Nordvold Hori- zontal Incised. 'These are the extremes; there are some that range in between. Brushing was confined mostly to the exterior neck and upper shoul- der areas. It was also associated with various other exterior surface treatments. Some brushing, in which parallel strokes were used, oc- curred in the upper portion of throat areas, but this was not a prev- alent practice. The areas that were brushed carried a series of sharply defined parallel vertical or horizontal scratches either at right angle to the lip or parallel to the lip and appeared to be the result of using a bundle of coarse grass or small twigs. Several sherds were noticed bearing red pigment mostly on the in- terior. These showed that some sort of slurry made of powdered ocher was applied. None of the vessels appeared to have been fired after this application of the pigment, which, at the present time, could easily be brushed off onto the finger as it was rubbed across the sur- face. This color can truly be called a “fugitive red” even though some of the pigment had worked its way into the natural pores of the vessel walls and in so doing had given the surface a faint pinkish or reddish hue. This trait was checked under a microscope a number of times and it was seen that the porosity of the vessel walls was just a “natural” for absorbing this material. Whenever cord impressions occurred, they were confined to the rim, the lip, or to the upper portions of the shoulder area. Cord impres- sions on the lip were rare, but they were numerous enough so that various types of treatment could be determined. They occurred either as a series of parallel impressions running straight across the lip, or at a diagonal, or as a series of parallel impressions covering most of the lip as they encircled the vessel. Practically all of the cord was Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 167 of the right-hand twist. Diameters varied somewhat, even on the same vessel; but on the average it measured around 1 mm. In the group of body sherds there were several whose exterior sur- faces were extra rough. When a positive impression was made of this treatment, with the aid of plastic clay, it was found that some sort of coarse fabric had been used. This fabric was not applied flat but was apparently wadded up and then applied, hence the term “fabric- impressed.” On a few sherds we found that a coarse mat impression covered very limited portions of the exterior surface. These sherds were small and we could not determine how extensively this technique had been used. We noted that in several instances corncobs had been used to roughen the exterior surfaces of vessels. This practice was apparently not common, for we found only a limited number of sherds treated in this fashion. Among the rare forms were a small number of plain sherds that were from scattered positions within the site. They were of a curious green color, more of a bile green, on the exterior, which seemed to go through the entire thickness of the wall. They were made of a hard and homogeneous paste, and their exteriors were very carefully smoothed. The thickness of each is uniform, ranging from 1.5 mm. to 3.0 mm. We could not determine their source or where they would fit into the general picture. LIP TREATMENT Lips were of many forms: they were simple-rounded, flattened, rounded and flattened, inward beveled, outward beveled, reinforced and beveled, reinforced and rounded, and even folded (see lip profiles, figs. 85 and 36). Lips had been given several types of treatments. They were smoothed or left plain; ticked or “tooled” in various ways with dif- ferent size sticks or other tools or ticked by having sections of cords impressed across them; incised by having herringbone designs drawn across them; punctated in various manners; indented by pinching the unfired clay into undulations or indentations; impressed with a number of twisted cord elements running parallel with the lip; given a wavy effect by pressure of the finger on either side of the lip; or given the stab-and-drag treatment with a rounded, pointed, or squared pointed implement resulting in a line of shallow punctations. Lips that were punctated, in some instances in the literature were referred to as “tool impressed,” impressed with a rounded, squared, or pyramidal pointed stick; the size and the shape of the impression depending on the actual outline of the tool and the amount of pressure used to bring about this effect. Some impressions were narrow, others Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER Ficure 36.—S-shaped rim profiles, Hosterman site. 170 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 were placed squarely across the lip, and others were at a diagonal; some impressions appeared to have been made by a single unit and others appeared to have been made by a double unit. Whether a stick or a sliver of bone was actually used to bring about this effect could not be ascertained. Cord elements of various diameters were used in this same manner with the impressions running either parallel with the lip, at right angles to it, or at a diagonal across the lip. Lip punctations varied exceedingly from small to large: those that were comma-shaped, teardrop, or oval in outline. There were a few whose punctations were tantalizingly irregular both as to outline and depth. From plaster casts taken to secure a positive of the element used, it appears that bits of rough textiles were pressed into the pli- able lip, creating these irregular, uneven, and unequal indentations. Then, too, there were impressions that were usually made by apply- ing the tip or one side of a blunt implement. Some impressions were shallow, circular, and saucer shaped in outline and spaced fairly close to one another. In a few instances a squared or very roughly pointed implement was used in place of the rounded-tip tool, but the technique of rendition and the results were the same. Whenever the side of the implement was used, the resulting impressions were elongated and trough-shaped in that both ends were open. They were either at right angles to the sides of the lip or drawn at a diagonal. This leads to the next technique in that these same two implements, but of smaller sizes, were used in a “stab-and-drag” technique in which the implement was thrust or pushed into the moist clay. Without withdrawing the tool it was dragged or pulled a short distance; this was then repeated over and over again, resulting in a line of intercon- nected impressions. In so doing, the terminal part of the former drag was pushed in, somewhat marring the sharpness. Herringbone designs on lips were of the widest variation. Some were rendered with very finely pointed tools and carefully drawn or incised, whereas at the other extreme are those whose elements were drawn with a very blunt instrument, at times deeply inserted into the lip. Usually those of the thin line elements were regular and care- fully drawn, whereas the coarser the rendition the more careless it was. If, by chance, the vessel should have had a handle or handles, this same treatment was usually carried over onto these features. Finger indentations of the lips were of two kinds. They could be pressed in from the sides to give the rim a sort of undulating effect or pressed down directly upon the lip to give it a scalloped or a wavy vertical effect. In some, where the latter method was used, the intervals were fairly regular, while in others the intervals were irregu- lar and the amount of pressure used varied from point of contact to point of contact; in other words, the “trough” varied in depth and was Riv. Bas. Sur. ss Pap. No. 35] HOSTERMAN SITE—MILLER 17 1 not at all consistent. This same scalloped effect was brought about by pressing some rodlike tool onto the plastic lip. The intervals between pressure points were never uniform. Whenever pressure was brought to bear vertically upon the lip it caused a certain amount of thicken- ing to that feature. Whenever pressure was applied at right angles to the lip the areas affected were thinned so that the intervening portion of the lip was thicker than the area worked on (fig. 40, a-d). TYPE NAME: “Le Beau Finger Indented” (pl. 19, A, h). TYPE MATERIAL: 142 sherds. Form: Rims: Vertical to gently flaring. Lips: Lips were rounded in preparation to being treated with the finger. After this the lips were broadened in the area by the downward pressure, bringing about a certain amount of thickening. Whenever finger pressure was applied alternately to the interior and exterior a wavy effect was created and the thickness of the walls was reduced. Thickness: Vessel walls varied from 3 to 5 mm. in thickness. Vessel form: Probably globular in shape, rounded shoulders, and straight to slightly flaring neck. Decoration: Variant A: Like the type description of Hurt’s (1957 a, p. 39) these rims were indented by applying the finger from the interior and ex- terior in an alternating pattern giving a wavy effect. Variant B; In this case the pressure was brought directly downward from the top of the lip giving a scalloped effect. The difference be- tween pressure points may vary. Some may be a continuous group; others were spaced with an alternate plain area. In every case the lips were rounded. STRATIGRAPHIC POSITION: Hurt has found that this type of pottery was as- sociated with the Le Beau Focus at the Swan Creek site (839WW7) and that percentagewise there was a tendency to increase with time. Ticked lines across the lips, frequently referred to in the Plains literature as “tool impressed,” have the widest variation as to width, length, and the amount of pressure used to bring about this effect. On some of the specimens the lines were very thin, short, and exceedingly shallow. On others the lines were long enough to reach completely across the lip and were both wide and deep. Still others were made at an angle. Inside measurements of the orifices of jars showed that there was not a great variation. Openings ranged from 10.3 cm. to 21.7 cm. (See Appendix 4.) RIM FORMS AND THEIR TREATMENT Rims of the plain, incised, brushed, and cord-marked wares were separated. It was found that all shared common basic characteristics in temper, method of manufacture, texture, hardness, color, surface finish, and general vessel form. All types are primarily groups of similar rim forms decorated in more or less the same general way. 661-932-6418 172 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 | NSS / YUL Ee a. / YD. OUD, «FL l Y >). Ble TERR) FRR » ps3 Vy y seas 5 \ Ficure 37,—S-shaped rim decorations. Riv. Bas. Sur. 7 AR Pap. No. 35] HOSTERMAN SITE—MILLER 173 Rims are of two main shapes, depending upon the configuration of the neck: vertical and S-shaped. Vertical rims ranged from short to tall and a few had the tendency to flare slightly at the top. There are several minor rim types present, but these are of limited number. After the rims were separated according to the two rim forms, they were next separated according to lip treatment, for I found this com- bination to be most diagnostic and one that was fairly sensitive. Each rim type was further subdivided according to the final exterior surface treatment of the rim. These include such treatments as: incised with a series of horizontal parallel lines; plain after smoothing; incised with a series of either short vertical and/or di- agonal lines; herringbone incised; a zone of either conjoining parallel lines or chevrons arranged as a band around the tip of the neck or rim, called line block; cord-impressed; simple-stamped,; and brushed. In separating the rim sherds into the various types it was necessary to take into consideration three main factors: (1) the general configura- tion of the rim, (2) the treatment of the lip area, and (3) the final exterior surface treatment of the rim or neck area. After all separation and counts had been made and tallied, it was found that 1,592 were vertical rims, 1,039 were S-shaped, and 524 were too fragmentary or incomplete to make positive classification possible. There are four types of vertical necks: (1) those that are almost vertical with just the slightest degree of outward flaring near the top (pl. 15, A, a) ; (2) those that are almost vertical with the exception of being slightly barrel shaped (pl. 19, B, e, fig. 38); (3) those that have vertical necks surmounted with a certain amount of rim thick- ening (pl. 15, B, a); and (4) those that were short and flared out- wardly at the top (pl. 16, B, a). Ficure 38.—Barrel-shaped rim, simple stamped rim sherd. 174 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 S-shaped rims were curved, flattened, or short with an acute angle. In either type the zone of decoration was confined mostly to the upper portion of the figure “S” (pls. 17 and 18; fig. 36). TyprE NAME: MHosterman S-shaped rims (resembling typical Le Beau series). TYPE MATERIAL: 1,039 rims or 32.9 percent of all rims. Exterior surface treatments: : (1) Horizontal parallel line incised—3872 rims, 35.8 percent (pl. 17, B; fig. 37, 1-6). (2) Line block—277 rims, 26.7 percent (fig. 37, 19-31). (3) Plain—138 rims, 13.8 percent. (4) Herringbone incised—124 rims, 11.9 percent (pl. 15, B, a; fig. 37, 14-18). (5) Vertical and/or diagonal line incised—66 rims, 6.4 percent (fig. 37, 7-13). (6) Cord-impressed—62 rims, 5.9 percent (pl. 20, A, a—d; fig. 37, 32-36). PASTE: All, with the exception of a very few, showed the general traits listed and described for the Hosterman pottery. ForMS Rims: There was quite a variation in the S-shaped and so-called collared rims. Lips: Since this was one of the variables, considerable variation was noted: (1) Ticked or tooled lips—546 lips, 52.6 percent of all lips. (2) Plain—309 lips, 28.8 percent. (3) Herringbone incised—138 lips, 13.3 percent. (4) Punctated—45 lips, 4.3 percent. (5) Finger indented—1 lip. Vessel form: Globular bodies with constricted necks with rather high rims. In some the curve is pronounced; in others the upper portion of the curve tends to be flattened somewhat. STRATIGRAPHIC POSITION: S-shaped rims were found in all levels of occupation at the Hosterman site and fragments of these same vessels were widely scattered. TABLE 3.—Types and numbers of S-shaped rim sherds, Hosterman site (89P07) Lip treatment Total S-shaped rims Herring-}| Punc- | Finger | Number | Percent- Plain | Ticked| bone tated | indented in age incised group Exterior surface treatment: Parallel-horizontal line incised___- 80 252 33 (| Bere a 372 35.8 ir) See Se ee eee ae oes Sac 27 32 57 21 1 138 13.3 Vertical and/or diagonal line in- CIROG = m8 ee oe Beate 52 Al (he ane Fip\ (ees Bee a 66 6.4 Herringbone incised_..._.-.------ 30 58 31 | ee 124 11.9 ineiplecks 24.2020 3 5 eee 77 180 16 4 ee Se 277 26.7 Cord-impressed.----..-..-------- 43 17 1 tf | PRRRCRE set 62 5.9 9] 1 | ea ee cae ee RIE ST do a 309 546 138 45 1 1, 039 100.0 INCISED S-SHAPED RIMS: 839 rims or 80.8 percent of rims. Form: Rims: Varied in height and in curvature. Some were rounded, others were flattened, still others had a distinct ridge or collared effect (fig. 35). ee HOSTERMAN SITE—MILLER 175 Lips: Several lip treatments were represented in this group: 1. Plain and rounded—239 lips, 28.5 percent of lips. 2. Ticked or tooled—497 lips, 59.2 percent of lips. 3. Herringbone incised—S80 lips, 9.5 percent of lips. 4. Punctated—23 lips, 2.8 percent of lips. Decoration: A zone of incised decoration was confined to the upper portion of the S-shaped rims, excluding lip treatment. Design techniques consisted of in- cising, including broad trailing, sometimes combined with punctations (fig. 39, a, b,c). There is a wide latitude in the design form from the simplest, which is a series of horizontal parallel lines crudely drawn across the upper portion of the neck. In certain vessels these lines were interspersed with an equal number of herringbone elements, or were bordered either at the top or bottom with punctations, or short diagonal lines. Then, there were rows of vertical or diagonal incised lines with their combinations; herringbone designs; opposed diagonals; lineated bands framed by diagonals: broad incised diagonals; broad-trailed diagonals; and curvilinear “rainbows” on a lineate field. PLAIN S-SHAPED RIMS: 138 rims; 18.2 percent of rims. Toru : Rims: Same as for incised. Lips: Same as for incised with the addition of finger indented—1. DEcoRATION: Surface has been smoothed after being paddled. No decoration applied. CORD-IMPRESSED S-SHAPED RIMS: 62 rims, 6.0 percent of rims. ForRM : Rims: Same as for incised. Lips: The four basic lip treatments still apply with the exception that four additional sherds were present whose lips were crossed with cord impres- sions. DeEcorRATION : Type A (Le Beau Horizontal Cord Impressed) has a series of hori- zontal cord impressions running parallel in the series starting just below the outer edge of the lip and continued down over the major part of the upper curve of the S-shaped rim. Type B (Rygh Rainbow Corded) has its series of horizontal parallel impressions interrupted by the insertion of a rainbow element made by impressing short segments of cords in this fashion to create this effect. CORD-IMPRESSED S-SHAPED RIMS: 62 rims, 6.0 percent of rims. These resemble the Le Beau Horizontal Corded of Hurt’s (1957 b, p. 41, fig. 21,5) type descrip- tion and a few are identical with the Rygh Rainbow Corded (ibid., p. 42, fig. 21, 1). Both of these types are well described and illustrated by Hurt so there will be no need to repeat these here. Rim decoration (fig. 87, 7-12), on vertical necked vessels with rounded lips, resembles the “unnamed” ware of Hewes (1949, pp. 59- 60, pl. VI, a, 6) of the Heart Butte Campsite, 32GT1, North Dakota. As indicated by Hewes this ware resembles one from the Burgois or Double Ditch site on the Missouri River, North Dakota (Will and Spinden, 1906, fig. 40, @) and from sites in the Upper Missouri Valley, North Dakota (Will and Hecker, 1944, pl. 6) all of which date in the late 17th and early 18th centuries. Rim decoration (fig. 37, 32 and 33), which is cord impressed on S- shaped rims with rounded lips, resembles part of the “unnamed” ware 176 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Ficure 39,—Line-block zonal decorations. ae Sr HOSTERMAN SITE—MILLER ‘77 of Hewes (1949, p. 60, pl. VI, c) from the Heart Butte Campsite, North Dakota, 32GT1, which is similar to “pottery in many Upper Missouri River village sites of the Mandan-Hidatsa culture, specifi- cally Sperry, Larson, Fort Lincoln, 32SI4 near Fort Yates, Hagen site (Mulloy, 1942). Pottery of this general type covers a wide area in the North Plains and probably several centuries of time.” (Ibid., pp. 59-61.) Rim decoration (fig. 37, 35 and 36) on S-shaped rims resembles a ware illustrated by Hewes (1949, pl. VI, d, 7) with the exception that the pottery from the Heart Butte Rockshelter, 32GT5, North Dakota, is cord-impressed and the sherds from the Hosterman site, 839PO7, are for the most part incised (fig. 40, 2,2). There is a certain amount of this same cord-impressed ware at the Hosterman site, but it is not abundant (pl. 20). Rim decoration (fig. 37, 34) resembles Hurt’s (1957 b, fig. 25, 2a, 26) Le Beau Cord-wrapped rod designs as well as Rygh Rainbow corded. The difference between the Le Beau Cord-wrapped rod and the Hosterman cord-impressed is only in the element used. Those from Le Beau used cord-wrapped rods, and those from the Hosterman used sections of cords (fig. 40,4). The placement of the elements on S- shaped rims of the Hosterman site and Rygh Rainbow Corded is identi- cal. Whether there is a time differential is not known at the present time, but it is suspected that the treatment of the Hosterman group pre- cedes that of the Le Beau and the Rygh forms. TYPE NAME: Hosterman Vertical rims (fig. 35). TYPE MATERIAL: 1,592 rims; 50.5 percent of rims. Exterior surface treatments: 1. Plain—579 rims, 36.4 percent of rims. 2. Horizontal parallel incised lines—397 rims, 24.9 percent of rims. . Brushed—245 rims, 15.4 percent of rims. . Simple-stamped—232 rims, 14.6 percent of rims. . Line block incised—53 rims, 3.3 percent of rims. . Herringbone incised—42 rims, 2.6 percent of rims. . Cord-impressed—27 rims, 1.7 percent of rims. . Vertical and/or diagonal lined incised—17 rims, 1.1 percent of rims. PASTE: All showed the general traits described for the Hosterman pottery. ForM: Rims: There was some variation in the overall height and some that had a slight tendency to be everted at the very top, but on the whole they are fairly uniform. Lips: Since this is one of the variables in determining types we found that there is quite a range in lip treatment, thus: . Herringbone incised—494 lips, 31.0 percent of lips. . Ticked or tooled—462 lips, 29.0 percent of lips. . Punctated—394 lips, 24.8 percent of lips. . Finger indented—142 lips, 8.9 percent of lips. . Plain—96 lips, 6.0 percent of lips. . Cord-impressed—4 lips, 0.3 percent of lips. Vessel Pie Globular-bodied vessels with vertical necks. OND AR Aor WD H 178 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 STRATIGRAPHIC POSITION: Vertical-necked vessels were found in all levels of oc- cupation within the areas explored at the Hosterman site, and often frag- ments of the same vessel would appear widely scattered. TABLE 4.—Types and numbers of vertical rim sherds (Hosterman site, 89P07) Lip treatment Vertical rims Herring- | Punc- | Finger Cord- Plain | Ticked| bone tated | indented |impressed incised Exterior surface treatment: arallel-horizontal line incised_-__-----__--- 17 79 255 45 lo scccnens|poeeeeeuar Plain co 526. he fee ae 2 OTe 27 165 150 162 75 1 Vertical and/or diagonal line incised_________ 3 Iso} Sat 25-62 gy (Se Aree Es see Hermrinebone incised29. 2.25. 93-281 6 21 DU)! | ae os Sed ees Bey aes SD MIME IOC eo ot eae Lats ie 8 39 2 yh eee |e Cord impressed....83>. Joes yee a 14) eee 9 S220. ees 1 Simple stampedes-— eae ee eee 6 66 43 93 24) | cseaeeees Brushed (seco see ee ae 15 79 20 86 43 2 Total Mere. fe pee +S ereg see 96 462 494 394 142 4 If we group all the recognizable rim sherds on the basis of ex- terior surface treatment, irrespective of whether they are vertical or S-shaped in form, as apparently has been the practice in the past, we get the following: Horizontal parallel incised lines—769 rims, 29.2 percent. Plain—717 rims, 27.3 percent. Incised line block of triangles, etc.—330 rims, 12.5 percent. Brushed—245 rims, 9.3 percent. Simple-stamped—232 rims, 8.8 percent. Herringbone incised—166 rims, 6.3 percent. Cord-impressed—89 rims, 3.4 percent. Vertical and/or diagonal line incised—83 rims, 3.2 percent. This grouping at best reduces the bulk to eight wares. It is not only misleading but is all too inclusive. For example, if we choose to place all cord-impressed into a single category we have lumped two distinct wares—the Le Beau Horizontal Corded and the Rygh Rainbow Corded—as a single ware. As to the exterior surface treatment of the rims themselves, some retained the marks of the paddle used during the formative stage of manufacture, better known as simple stamping. These marks formed a permanent exterior surface and as such can be considered a form of decoration. In others the simple-stamped effect was either completely dissipated or partially obliterated, resulting in a smoothed plain surface. The process did not stop there, for this plain surface acted as the background upon which various linear geometric designs were incised and—in a few instances—impressed. The simplest form of incised design consisted of a series of encircling horizontal parallel lines starting just below the juncture of the neck and exterior margin of the lip and extending downward sometimes for a short distance and sometimes covering the major part of the neck of the vessel (fig. 40, ¢). Riv. Bas. Sur. en Pap. No. 35] HOSTERMAN SITE—MILLER 179 Similar treatment occurred on Wheeler Horizontal Incised from the Scalp Creek site and the La Roche site (Hurt, 1952, p. 18; fig. 14, 7-3), and Nordvold Horizontal Incised (Wilmeth, 1958, fig. 17). A variation of this design was brought about by breaking the lines and filling in the gaps with interconnecting herringbone designs, thus: a —~AA~=S|} The intervals between the lines varied considerably on some speci- mens and on others were very carefully drawn. Those that were carelessly drawn had the lines wandering either upward or downward or even inserted small segments that started anywhere and ended nowhere. Other variations of the horizontal parallel line patterns consist of the series of parallel lines delimited by a row of short diagonal dashes, delimited by a row of punctates, surmounted by a row of punctates, or separated by concentric rainbow designs. Next in simplicity are those vessels decorated by drawing a row of diagonal or vertical lines starting just below the juncture of the rim and the outer edge of the lip and carrying them down a short dis- tance on the rim (fig. 40, 7»). Variations of this motif consist of an encircling incised line above the series of lines, an encircling incised line below the series of lines, a row of punctates delimiting the lines at their base, or a series of horizontal dashes just below the base of the lines. Those incised with herringbone designs may have one or more lines of this design used as the decorative motif. Variations consist of a single horizontal incised line drawn above the herringbone design and separating it from the top of the rim area; delimited by a row of vertical punctates at the base, or delimited by a row of round or oval punctates at the base. The most complex are those decorated with linear geometric de- signs. ‘These consist of a series of interlocking triangles filled with horizontal lines, diagonal lines, and herringbone designs. These in- terlocking triangles may be separated from one another by one to six parallel lines. ‘Then, too, a row of punctates can appear at the top or the bottom of the zone of interlocking triangles. In this paper when- ever any reference is made to this type of decoration the term “line block” will be used. Vertical brushing covers not only the rim and neck sections of a great many vessels but is also found on the upper portions of the shoulder area. Some of the Talking Crow Brushed types described 180 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Og \ Oe “theta wt \ (\ Mt 7 an Nfl WN Yd i optidl Eat Uj I py | t Ficure 40.—Rim treatment and decorative zones. by Carlyle S. Smith (Smith, 1951, pp. 34-37) are almost identical with some of those found at the Hosterman site. The majority do not fit into this category. TICKED LINE, INCISEi-X—Type “A”: This type consists of a series of horizontal parallel lines which start just below the lip and cover the entire neck area. Necks are vertical Riy. Bas. Sur. Pap. No. 38], HOSTERMAN SITE—MILLER 181 to slightly flaring or excurvate. Lines are not carefully drawn or joined and many are just short segments; some join while others start and stop without any unity with the rest of the group (fig. 40, e). On the whole it portrays a lack of fineness of design and a distinct indica- tion of a lackadaisical inclination toward the satisfaction of creating something that has been carefully done. Lines are drawn either with sharply pointed tools, a form of incising, or with dull, flat-bladed tools, a form of trailing. Some of the lines are drawn as close together as 2mm.; others as far apart as 6 mm. On the whole, the surface of the neck was smoothed before the lines were incised upon it, but there are any number that show that the surface was first brushed and then the incised or trailed lines were drawn upon this rough surface. Several appear with wavy lines because the surface was poorly prepared rather than because of the shakiness of the potter herself. Perhaps this group can be divided into two subgroups: A, and A. A: has a row of diagonal incisions directly beneath the lip and above the series of horizontally drawn lines. A; has a row of punctations below the series of horizontal parallel lines to set the lines apart from the shoulder area. Wall thickness varies from slightly under 4 mm. to 8 mm. Color variation still persists. Soot incrustations are still present. Interiors are smoothed without any evidence of brushing. TICKED LINE, INcIsED—Type “B”: The same holds true for type “B” with the exception that there are no subtypes. 182 BUREAU. OF AMERICAN ETHNOLOGY [Bull. 189 Color, soot incrustation, brushing on exterior, wavy lines, the way the lines were incised or trailed, distance apart, type of tool used, and general appearances are all alike. TICKED LINE, INCISED—Type “C”: The same holds true for type “C.” Plain lip, vertical neck that is covered with a series of horizontal parallel lines. Al the characteristics as described for the ticked lip group apply here. There is one exception. Two specimens have their lines drawn rather deeply penetrating about half the thickness of the wall itself. Punctated.—All. of the characteristics common to the others are present here. The only difference is in the treatment of the lip itself. In this case the lip is punctated. Punctations appear to have been made by: (1) Fabric impressions (2) Round blunt tipped tool (3) Comma-shaped tool (4) Teardrop-shaped tool (5) Small grass stem (6) Pyramidal-tipped tool SHOULDER TREATMENT Shoulders of the various vessels were either simple-stamped (figs. 41-43) plain-smoothed, or incised (figs. 44, 45), and from the pres- ence of a few black-on-gray painted sherds, it would appear that the shoulders were occasionally painted with crude geometric designs. Parts of three medium-sized simple-stamped vessels are unique in that all of the shoulder areas have been punctated with a series of large punctates, ca. 12 mm. in diameter. The punctates on two of the vessels are spaced about equally apart and appear on the upper part of the shoulder. All are fairly deep. Consequently, nodes ap- pear on the interiors of the vessels. The punctates on the third vessel have been arranged in groups of vertical series, two rows to a series and four or more punctates toa row. The punctates start just at the base of the neck constriction or the top of the shoulder and run down- ward midway onto the body of the vessel (see pl. 20, B, 6). The punctates are not placed in a straight line and appear to have been made by tapping the vessel quite firmly, while it was still plastic, with the tip of the finger or some implement of comparable size. Shoulder incising appears to be quite popular at the Hosterman site. It consists of various combinations of line-filled triangles (fig. 46), triangles filled with zigzags, or crude herringbone designs. None can be said to be very carefully drawn. Apparently, the potters were not after carefully executed work but rather the effect of the design Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 183 INCHES cM, Figure 41,—Restoration of vertical rimmed, simple-stamped vessel. itself, otherwise, it seems, they would have been more careful in draw- ing the lines of the design (figs. 45 and 46). HANDLES Handles on Hosterman vessels are not common (pls. 16, A, a; 17, B; 18, A; 19, A, a). In the main they are of two varieties: strap (19 specimens) and horizontal lug (16 specimens). Strap handles, as a rule, are either welded at the lip or are tonguelike elongations of the lip that curve downward a short distance to be welded or riveted to the body of the vessel. On the other hand, there are rare instances when the strap handle may originate just a short distance below the lip, on the upper portion of the rim, and follow the same pattern as the others in their basal attachment to the walls of the vessel. Most of the strap handles project horizontally before curving downward to make their 184 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 I § u U Uy / 7 og ‘XN Pe he si ba 7 / \ / \ / \ t \ t \ I \ I 1 t ! ( i { | \ | \ \ x \ \ \ Xe \ 4 XN 7 % 4 > 7 \Q 4 x 7 SS pie pies ne ~ -? 0 2 oO (NGHES 6 OO CMS. Ficure 42.—Shoulder decoration on large jar sherd, Hosterman site. juncture. Usually the upper section of all strap handles is much wider at the top than at the bottom, giving the handles a triangular appearance. In cross section most are oval near the base, others are flat, while a few are concavo-convex. Handles of this type may appear as pairs, opposite each other on opposite sides of the vessel, or they may be separated by small lugs appearing in the opposite quadrants. The same design used on the adjoining lip decorates most strap handles, being simply carried over onto the handle. The design may be herringbone incised (16 specimens), horizontal cord impressed (1 specimen), horizontal line incised (1 specimen), or horizontal stab and drag (1 specimen) (fig. 47). None were left undecorated. Lug handles are somewhat variable as to shape, size, and place- ment. Practically all are tonguelike projections off the lip or side of the vessel. Others may appear lower down on the neck or at the Riy. Bas. Sur. ree: Pap. No. 39] HOSTERMAN SITE—MILLER 185 INCHES cM. Ficure 43.—Rim and shoulder area of S-shaped rimmed vessel. bottom of S-shaped rims. Mostly they are just paired horizontal projections extending a short distance out from the exterior surface of the vessel. Almost all appear to have been fashioned at the same time the vessel was finished off; only a very few appear to have been welded onto the vessel afterward. They are roughly triangular in shape. One rare case is that of an S-shaped rim upon which are five luglike projections on the upper part of the “S” that are but slight downward elongations. Each projection is separated by an equal number of cord-impressed rainbow designs that also appear on the lugs (pl. 19, B, a). These are spaced equidistant around the rim of the vessel. These projections vary in the distance they protrude and whether they droop, sag, or extend straight out from the sides of the vessel. As on the strap handles the incised herringbone design is carried over onto the tops of most of them. Width and length may 186 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 INCHES ——— cM. Ficure 44,—Line-block design on shoulder area, Hosterman site. vary as to the width and distance they may protrude. Luglike projec- tions may vary from 2 to 4 or more to a vessel. They appear to be evenly spaced around the vessel from what we can determine from the sherds. One lug handle, on specimen No. 3372/16, is unique in that it be- came limp shortly after it was fashioned onto the vessel and dropped down onto the neck of the vessel forming a very tenuous joint with it. It appears to have been unintentional but whether it was or not it would appear to function as an evolutionary trend in explain- ing the formation of a type of strap handle from the lug type. VESSEL SHAPES Vessel shapes appear to lean to globular with vertical and slightly everted necks, rounded shoulders and bases, with the shoulders dec- orated with incised motifs and the necks covered with a series of incised horizontal parallel lines. This type of vessel shape is rather widespread for it also occurs at the Swan Creek site (89WW7), Steamboat Creek site (89PO1), and the Four Bears site (39DW1), Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 187 = - = =. - — es ee oe INCHES 2 ——=— a CM. Figure 45,—Line-block design on shoulder area, Hosterman site. all of which are neighboring villages. This does not in any way suggest a contemporaneity but only an indication of the relative occur- rence of similar vessel shapes with similar exterior surface treatments within a temporal and spatial relationship of a very small area. A large number of vessels had a heavy incrustation of soot which had adhered to the outside of them, filling in all depressions made by the various design elements, blotting out all indications that the vessel was decorated in any manner. This heavy soot incrustation may be and could be due to the tendency of the potters of the Plains, as pointed out by Will (1906), and others, to coat the outside of vessels heavily with grease prior to firing. The intense heat used 661-932—64—_14 188 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Ficure 47.—Handle types showing points of origin. would change the hydrocarbon into pure carbon consolidating it into a firm mass surrounding the exterior of the vessel. PAINTED POTTERY Three small potsherds found in the large mass of body sherds had fragmentary designs in black painted on their exteriors. Since none Riv. Bas. Sur. fa, Pap. No. 35] HOSTERMAN SITE—MILLER 189 of these sherds were very large, it was impossible to determine the complete design of the original. The basic surface finish of each dif- fered somewhat. Two of them had their exteriors smoothed before the paint was applied and the third one still retained the initial simple- stamped treatment over which the paint was applied. The very presence of even this small number of painted sherds would render this site unique since painted designs on the exteriors of vessels are rare rather than commonplace. The use of a red pigment on the interiors and some exteriors has been reported from other sites. This trait appears to be common enough in the Plains. Even so, it is not too prevalent here at the Hosterman site; so this, too, can be classed as a rarity. MINIATURE VESSELS Several fragmentary vessels in the form of jars were found in various areas in the site. Four basic forms are represented. All have two traits in common: rounded bases and globular bodies. The smallest of the six miniature vessels, field cat. No. 3428, has the following measurements based on possible reconstruction: diameter of the mouth, 35 mm.; diameter of the neck, 30 mm; diameter of the body, 43 mm.; and height, 45 mm. The neck is slightly constricted and above the constriction is a small rim section (see outline drawings of these vessels, fig. 48). The second vessel, field cat. No. 742, is shghtly larger, and resembles the first with the exception that it is squattier. It has the following measurements: diameter of mouth, 36 mm.; diameter of neck, 32 mm. ; diameter of body, 61 mm.; and height, 50 mm. The next vessel, which also has a constricted neck, is larger and the constriction is not as abrupt; field cat. No. 290/60. It has the follow- ing measurements: diameter of mouth, 46 mm.; diameter of neck, 40 mm.; diameter of body, 59 mm.; and height, 50 mm. The next larger vessel, field cat. No. 290/56, has a unique shape in that the constriction is about midway down from the lip and nearly separates the vessels into two equal parts. Whether the base is sub- conoidal as represented or more rounded appears to be questionable. Its measurements are as follows: diameter of mouth, 63 mm.; diameter of neck constriction, 57 mm.; diameter of body, 68 mm.; and estimated height, 63 mm. Next to the largest vessel is a globular vessel, field cat. No. 3206/1, with a slightly constricted short neck. It measures 61 mm. across the mouth, 61 mm. just above the shoulders, 84 mm. in body diameter, and 84 mm. in height. The last and largest of the series, field cat. No. 290/116, is a varia- tion in shape of the vessel just described. The mouth of the vessel 190 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Ficure 48.—Various miniature vessel types, Hosterman site. measures 85 mm. in diameter, 83 mm. across the slightly constricted neck, 97 mm. in body diameter, and 95 mm. in height. Four of the vessels have smoothed exteriors. Specimen No. 742 has a series of radiating lines starting at the top of the shoulder and running down almost to the base. Between these radiating lines are series of horizontal parallel lines spaced so as to fill in the intervals. Specimen No. 290/116 has the rim and the upper portions of the neck covered over with corncorb impressions that are quite distinct. Ap- parently the entire exterior of the vessel was treated with a corncob and later most of the impressions were obliterated. Small patches retain these impressions. The interior of the vessel has a smoothed surface but is rather uneven as to thickness. The vessel tends to be lopsided, Table 5 gives the measurements of all the miniature vessels. EXOTIC VESSEL OF POSSIBLE CADDOAN EXTRACTION One very noticeable miniature vessel, field cat. No. 1347, possesses two mouths or spouts (pl.21). It was found in a cache pit, Feature 9, Riv. Bas. Sur. Dae Ne Sot HOSTERMAN SITE—MILLER 191 TABLE 5.—Measurements of miniature vessels Field specimen No. Diameter Diameter Diameter Height of mouth of neck of body mm mm. mm mm. 5 Use Ee oe Oe ee BLE Pea ed Bs PS EO 35 30 43 45 Ci Eee ENS Fe | ES gee oe 3 BLE 36 32 61 50 POLUG LEST SIT IN, 8 ey eT 46 40 59 71 OO bie ss ee 8 ee a See seit ees 63 57 68 63 BOUGMOII! 8 IST Hass oye Sis 61 61 84 84 POOFPLGE Men FHSS See ee es TEN 85 83 97 95 at a depth of 5.8 feet from the present surface. This is a most unusual find because its form is one seldom found in the Upper Plains. Its surface color is mottled in that it runs from a tan to gray and black with fire clouds very much in evidence. When the vessel was com- plete it was oval in outline and rather squatty in height. The spouts, or mouths, are atop two short vertical necks that are not decorated. Between the spouts and covering on the shoulders a crude herringbone design was cut only moderately deep into the clay. Over the rest of the body appears a series of vertical and diagonal incised parallel lines, some of which crisscross. The lips of the two spouts are crudely punctated. Walls at the top are rather thin, 3.0 mm., and the basal portion is fairly thick, 8.0 mm. (fig. 49). IN «ee Gi cms. Ficure 49.—Two-spouted vessel, Hosterman site, showing decorative design. CLAY DAWDLES Throughout the site and’in some of the midden pits: were several objects of clay that appear to have been leisurely and probably unin- tentionally fashioned into vaguely recognizable or unrecognizable ob- jects. Some resemble minature clay vessels or toys; others are in the form of spheres or marbles, known during my early years as “Dough babies”; still others are nondescript in form. All have been fired to 192 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 insure permanency. It is probable that such objects were made while members of the community were seated around campfires on some cold winter night and, to pass the time away, took up clumps of moist clay that were handy and unconsciously molded these objects into shape. When the interest in the object no longer existed it was cast into the open fire and soon forgotten by the maker. Later, when the fire area was cleared away, these objects were tossed into a midden pit along with the ashes or scattered onto midden piles within the confines of the village area. Furthermore, a lump of compacted ashy clay was found in Feature 2 at a depth of between 1.0 foot and 2.0 feet. It is roughly triangular in outline and has a maximum length of 4.7 cm., a breadth of 3.8 cm., and a maximum thickness of 2.4cm. Near the apex of the piece there is a truncated conical hole that is widest at its point of entry (1.9 cm.) and tapers down to a diameter of 1.0 cm., continuing through the clay object. Under microscopic examination there are no vertical striations inside the bore, a fact indicating that the object was never used as a form of abrader. The physical composition of the object is rather soft; a fine powder comes off on the finger when rubbed across the surface. The purpose and use of these objects is problematical. GENERAL STATEMENT I found Strong’s (1940, p. 864) summary of Mandan pottery note- worthy in that he found many traits that are similar to those of the Hosterman series. He stated: The pottery is rather granular and appears to have been made by the paddle- and-anvil method, no evidences of coiling being observed. ...is tempered with medium to fine grit obtained from granitic rocks broken down in the fires. Colors are predominantly dark but run from almost black, through browns and gray, to very rare buff or even orange tones. The surfaces are often superficially blackened with grease. ... Red hematite in a powdery form occurs inside some sherds but appears to be accidental, probably as a result of use in mixing paint. The surface of most pottery is marked by vertical grooves and ridges, apparently the result of paddling with grooved or thong-wrapped paddles [simple stamping]. ... The shoulders and necks of vessels show vertical scratches as though grass-rubbed. All have been smoothed down, and the small vessels often have a plain surface. ... Rims and necks that are S-shaped are most common [which is not true of the Hosterman site, as shown earlier], flaring profiles with lips somewhat thickened next [which are a minority form at the Hosterman site], and intermediate forms [7 percent]. Castellations, lugs, spouts, and strap handles are rare. ... The S-shaped rims are usually deco- rated with horizontal or diagonal twisted single-cord impressions, closely spaced. A curvilinear design suggesting a rainbow is very common. . . . Incised designs are common on the shoulders of vessels, usually consisting of opposed diagonals. About one-fourth of the rims are plain, and finger-pinched rims are common. On the other hand, Cooper (1949, p. 303) noted that the pottery re- mains from a number of sites between Pierre and the mouth of the Riy. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 193 Cheyenne River were apparently identifiable with Arikara ceramics as described by Strong (1940, pp. 868-369, 381). The Hosterman site, situated about 40 miles upstream from the mouth of the Cheyenne River, could fall well within the sphere of the Arikara settlement pattern and hence one would expect to find to a certain extent com- parable pottery forms and types of this group at the site. There is close resemblance between the two. ARTIFACTS OTHER THAN POTTERY CHIPPED-STONE ARTIFACTS PROJECTILE POINTS Six hundred fifty-nine projectile points, either whole or frag- mentary, were recovered. Several varieties of stones were used in their manufacture: quartzite, chalcedony, and chert. Two specimens were made of obsidian. Nearly all are thin, and a number show a primary scar on one face, whereas most were made by secondary chip- ping of a small detached flake. The final shaping of the edges was accomplished by flaking from both sides. Two hundred sixty-nine were sufficiently complete to allow for classification. Flaking is neat and rather delicate in most specimens, and since it proceeded from both sides the resulting point is very symmetrical and slightly lenticular in cross section. Most of the stone used in the manufacture of projectile points oc- curs in the vicinity of the Hosterman site, in the riverbed or on the terraces around the site. Some, like obsidian, had to be brought into the area from other vicinities. The arrow points of the Hosterman site are small, light, and pres- sure flaked. Four basic styles are represented as well as the inevitable scattering of deviants. GrouP 1 (pl. 22, A) ; 62 specimens.—Triangular in shape with straight or slightly convex edges; sides are slightly longer than the base; straight bases as a rule, but there are some slightly concave or convex bases that occur rarely. Length: Mean of 14 mm., ranges from 138 mm. to 19 mm. Width: Mean of 138 mm., ranges from 9 mm. to 18 mm. Thickness: Mean of 3 mm., ranges from 2 mm. to 4 mm. Group 2 (pl. 22, B) ; 112 specimens.—Isosceles triangular in shape with straight or slightly concave edges; straight or concave bases. Maximum width at base. Length: Mean of 25 mm., ranges from 20 mm. to 29 mm. Width: Mean of 14 mm., ranges from 11 mm. to 17 mm. Thickness: Mean of 3 mm., ranges from 2 mm. to6 mm. Group 3 (pl. 23 A); 21 specimens.—Triangular with straight or convex edges, straight or convex bases. Maximum width at base. Length: Mean of 34 mm., ranges from 24 mm. to 43 mm. Width: Mean of 24 mm., ranges from 19 mm. to 29 mm. Thickness: Mean of 7 mm., ranges from 5 mm. to 9 mm. 194 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Group 4 (pl. 28, B); 66 specimens.—Side-notched, straight-sided triangular, maximum width at base, straight base, slightly concave or convex bases occur rarely. The notches are at right angles to the long axis and are often very narrow or shallow. Length: Mean of 38 mm., ranges from 18 mm. to 58 mm. Width: Mean of 17 mm., ranges from 12 mm. to 22 mm. Thickness: Mean of 4mm., ranges from 2 mm. to6 mm. MISCELLANEOUS (pl. 23, B) ; 8 specimens.—These eight specimens were complete and none of them fit into the four basic groups. One small point is stemmed; another is corner notched; several have expanded bases; and the remainder are side-notched but of a different pattern than those of group 4. MICROBLADES Microblades (pl. 24) are made of small thin flakes of chalcedony, quartzite, chert, jasper, and obsidian. Unlike the microblades from the areas to the Far North, the blades from the Hosterman site were not struck from specially prepared cores, for not one really good core was found throughout our excavations. Out of the aggregate we separated two groups. Group I are much longer, ranging in length from 25 mm. to 57 mm. Group II are much finer and shorter and range in length from 18 mm. to 38 mm. Both groups are very finely retouched along the edges. Evidently, such small blades must have been inserted into handles of wood, for we found no bone handles that would fit such small flake blades. They were probably used as knives or sickles. Two of the lots are bifaced in that they were retouched on both faces. Whether these are a form of drill, punch, or perforator of some type cannot be determined. Others are forms of side scrapers or knives since one or both edges have been very carefully retouched and the very small thin flaking scars indicate that work was performed on these small thin flakes to convert them into workable tools. The initial flakes were not struck from specially prepared cores, for none were found within the area investigated. Apparently, the workmen were skillful chippers, for they were capable of striking off thin flakes from any one core without too much trouble. Not only were they capable of securing thin flakes but they were qualified in casting off flakes of sufficient lengths that could be converted into workable functional tools. If the bulbar end, or the dorsal or underside surface, is placed downward and toward the workers and the worked edge is noted, it will be found that only 30 percent of these small tools were retouched in the left edge. Many of the flakes, both large and small, show signs of use about one or more edges as though they had been picked up from the scrap heap and used temporarily for scraping or cutting. Once the job was completed or they were too dull to be of further use they were thrown away. er. Bes jSor. HOSTERMAN SITE—MILLER 195 SCRAPERS Scrapers, since they occur over long periods of time, are much too ubiquitous to be considered as good horizon markers as they first appear in Early Man sites, continuing through the Archaic and into pottery-making horizons. They have become so well adapted that they have continued to be manufactured over long periods of time; thus, their usefulness as “index fossils” has been destroyed. Scrapers are flake tools. They vary bothastosizeandform. They are characterized by a plane or slightly curved undersurface (ventral face) surmounted by a dorsal keel or flatness displaying one or more abruptly retouched edges. The working edge is somewhat convex, except in the case of a very thin variety. The convexity is clearly the essential feature. Maximum thickness is not confined to any par- ticular portion of the scraper. They are unifaced and ovate, ellip- tical, lunate, triangular, subtriangular, trapeziform, circular, or ir- regular in outline. The very presence of scrapers portends certain technical knowledge of the preparation of pelts into leather goods for the manufacture of clothing, etc. I believe Stewart’s (1946, p. 45) definition of a scraper is an ex- cellent one. He describes it as: A primitive thing called a scraper is crude and not at all eloquent until you realize that it points to much else. It means not only a scraper, but a thing to be scraped, most likely a hide; therefore it means a growing ability to kill, to take the hide and cure it. That is just the beginning, for a scraper also shows a knowledge of how to scrape, and a desire for scraping, and enough leisure (beyond the struggle to get food) to allow time for scraping. All this means self restraint and thought for the future, and it implies a certain confi- dence in the ways of life, because no one would be liable to go to all the trouble of scraping if he did not have the reasonable hope of enjoying the results of his work. Scrapers can roughly be divided into two classes: end and side scrapers. End scrapers have the working edge or edges on the ends of flakes and are roughly convex, and where there is a central keel it frequently rises up fanwise to meet it. Sometimes much fluting occurs along the side of the flake to trim it into the desired form. End scrapers can be subdivided into: keeled, circular, large and smail teardrop, small planoconvex, large and small flake, flat thumbnail, beveled, triangular, subtriangular, peaked planoconvew, and scraper- graver combination. Side scrapers are subdivided into: large flake single edged, large flake triple edged, large planoconvex triple edged, double edged, prismatic flake, thin pointed double edged, and combination side scraper—spokeshave. 196 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Practically all of the scrapers have been neatly made. The ma- jority of the end scrapers occur on especially prepared flakes. The bulb of percussion of the ventral surface indicates that flakes were utilized and purposefully struck off in the manufacture of scrapers. The scraper end, opposite the bulb end, is neatly trimmed and one or more sides are retouched forming a tool with either a rounded, straight, or ogival double working surface. Asa rule, retouching oc- curs on all edges, but it can occur on only one edge. Scrapers are characterized by retouching along one or more edges of the dorsal sur- face, and the ventral surface is unworked and is either flat or concave as the result of the initial flaking in which the piece was cast off from a core. MacNeish (1954, p. 246), while observing Slave practices, found that end scrapers were used primarily in fleshing and scraping the skins of the animals they killed. On the other hand, side scrapers could also function equally as well as skinning knives and were used in separating the skin from the animal and in the preliminary removal of muscle tissue from the skin. Both functions appear plausible from the form of the many scrapers recovered from the Hosterman site. The total collection of scrapers from the Hosterman site includes 777 specimens. Materials used in the manufacture of scrapers appear to represent the entire range of silicious minerals present at the site or in the immediate vicinity of it. They are described as chalcedonies, quartzites, jaspers, cherts, and flints. By far the more common are the chalcedonies. As described many times, the basic scraper form is the result of striking a core a blow aimed slightly inward so as to produce a short flake. In a number of instances it breaks with a hinged fracture. The flake when viewed from the side is usually considerably thicker at the end opposite to the striking platform and the bulb of per- cussion. In some there is a pronounced downward curve near the thicker end. When viewed from the top, many are trapezoidal or pointed oval in outline with the greatest breadth at or near the thick- ened end. “As is true at so many of the later Plains sites, end scrapers made on flakes are a common type of chipped stone artifact. . . . The only retouching considered to be indispensable was that on the broader end, that produced a more or less straight or smoothly curved working edge and asteep face that intersects the flake scar at an angle approach- ing 90 degrees. Retouching on the sides was presumably intended to produce the proper shape only; it may be absent if the flake already had a suitable form, or it may be present on only one side. Very thick flakes sometimes show a steep and bold retouch with flake scars from both sides meeting in or near the center to form a longitudinal keel. Riv. Bas. Sur. 2 ae Pan. No. 35] HOSTERMAN SITE—MILLER 197 ee oO INCHES 8 METERS Ficure 50,—Keeled scraper types. In others the back was simply left in its original form, often showing longitudinal facets where previous flakes had been struck off the core. It would appear that the makers were not overly concerned with the method by which the result was obtained so long as a serviceable implement resulted” (Spaulding, 1956, pp. 48-44). Nor were they concerned as to the final shape taken by the implement whether it was trianguloid, rhomboidal, or rectangular. There is quite a range in size as to length, width, and thickness. Lehmer (1954, pp. 57-58), in studying the Dodd site, separated the end scrapers into five distinct groups depending somewhat upon their outline and method of chipping to determine the shape. These same forms, as well as others, occur at the Hosterman site in quantity. Keeled end scrapers (fig. 50).—Scrapers of this class were made from flakes having fiat to concave ventral surfaces and a single ridged dorsal surface. The working edge has been rounded and steeply re- touched. Sides taper somewhat to a blunt squared end. Scrapers of this type have been subdivided into type 1 and type 2 (pl. 25). Type 1 keeled end scrapers.—This type of scraper, represented by 75 speci- mens, has all sides retouched with the small fluted scars meeting at a central ridge or keel. Overall length varies from 20 mm. to 37 mm.; overall width from 15 mm. to 27 mm.; and overall thickness from 7 mm. to12mm. Mean read- ings would be lengths, 25 mm.; widths, 22 mm.; and thickness, 9 mm. The retouching along the sides rises up fanwise to a single rounded keel on the dorsal surface varying the thickness of the specimen so that no two are exactly of a uniform thickness. The presence of this type of chipping indicates that the person creating such an artifact was most capable in his ability to cast off these long thin chips uniformly and regularly. 198 BUREAU OF AMERICAN BTHNOLOGY [Bull. 189 Type 2 keeled end scrapers—tThis type of scraper, represented by 57 speci- mens, differs from Type 1 by having a natural triangular keel without being shaped by retouching. The keel is sharp and well defined. Overall lengths vary from 21 mm. to 40 mm.; widths from 17 mm. to 27 mm.; and thickness from 6 mm. to 12 mm. Mean readings would be lengths, 29 mm.; widths, 21 inm.; and thickness, 8 mm. ; In outline, these scrapers are roughly triangular with slightly excurvate edges. Type 1 has its dorsal surface retouched all over rising to a rounded keellike prominence. Small round scrapers (pl. 25, A, p-z).—Twenty-one scrapers have been placed into this category. They are not absolutely round, but tend to be somewhat oval in outline. They vary in maximum length from 19 mm. to 28 mm., in maximum width from 18 mm. to 25 mm., and in maximum thickness from 7 mm. to 10 mm. Scrapers of this form are not too plentiful and are, perhaps, the end result of many retouchings. Large and small teardrop end scrapers (pl. 25, B, a~-d). —Teardrop end scrapers are ovoid in outline and planoconvex in form. All are flake tools. Large teardrop scrapers range in length from 387 mm. to 51 mm., in width from 21 mm. to 30 mm., and in thickness from 6 mm. to 11 mm. Small teardrop scrapers range in length from 23 mm. to 33 mm., in width from 16 mm. to 19 mm., and in thickness from 5 mm. to 8 mm. The ventral sides are always concave while the dorsal surfaces have the edges retouched gradually sloping to a flat keel. The lateral edges are slightly excurvate in outline. Each of these forms is represented by 6 specimens or 12 in all. Small planoconvex end scrapers (pl. 25, B, e-m).—End scrapers of this type do not conform to any special group but are established solely upon form. Seven specimens have been assigned to this type. In length they vary from 23 mm. to 32 mm., in width from 16 mm. to 22 mm., and in thickness from 7 mm. to 10 mm. One scraper (field cat. No. 1887) has a distinct graver tip near the butt end of the tool. Large flake end scrapers (pi. 25, B, n).—There are 20 specimens in this group. Flake end scrapers are mostly irregular fragments of flint with a flat ventral surface and retouching on oneend. The rest of the artifact was never retouched. Their lengths range from 27 mm. to 56 mm., widths from 14 mm. to 29 mm., thickness from 2 mm. to 11mm. One specimen (field cat. No. 3049) has a wide notch worked out near the base. The specimen measures 17 mm. across. Apparently this was used to smooth down wooden tools, such as arrow shafts, etc. Small flake end scrapers (pl. 25, B, o-w).—There are 43 specimens in this group. These are much smaller flakes and, like the larger, are worked on only one end. The difference between the two is a matter of length. These scrapers vary in length from 17 mm. to 29 mm., Riy. Bas. Sur. ae Pap. No. 35] HOSTERMAN SITE—MILLER 199 in width from 13 mm. to 23 mm., and in thickness from 3 mm. to 10mm. One specimen (field cat. No. 2407) has a burinlike base. Flat end scrapers (pl. 26, A, a-f).—Thirty-three specimens are in this group. All were made of comparatively thin flakes. Measure- ments show that they vary in length from 15 mm. to 38 mm., in widths from 16 mm. to 30 mm., and in thickness from 4 mm. to9 mm. The retouched surface may be straight or curved. Thumbnail end scrapers (pl. 26, A, g-z) Twenty-two specimens are in this group. Most are very small and very well made. They vary in outline from trianguloid to rectanguloid. In length they range from 14 mm. to 23 mm., in width from 12 mm. to 19 mm., and in thickness from 3 mm. to 7 mm. The average length runs from 17 mm. to 19 mm. The average widths can range between 14 mm. and 17 mm.; and the average thickness is about 5 mm. Beveled retouched edge end scrapers (pl. 26, B, a-7).—There are two types of these scrapers: those that slope upward to the right and those that slope to the left.. Nine specimens have retouched edges that slope to the left whereas only six slope to the right. Whether these are valid types can only be verified by examination of other nearby collections. Most of the ventral surfaces are flat, but there is an occasional one that is planoconvex in outline. Those scrapers that slope upward to the right come very near to being uniform in length. Three are 25 mm. long; two are 24 mm. long; and only one is 21 mm. in length. Widths run about the same: three are 17 mm. in width; two are 18 mm. in width; and the third is 20 mm. in width. Thickness ranges from 4mm.to 8mm. Five are excurvate trianguloid in outline and the other is roughly rectangular. The scrapers that slope upward and to the left are not quite as uniform as to size. They range in length from 23 mm. to 29 mm., in width from 15 mm. to 21 mm., and in thickness from 6 mm. to 9 mm. Triangular end scrapers (pl. 26, B, j-r).—Nine specimens. These are excurvate edged triangles and are made from comparatively thin flakes. They range in length from 22 mm. to 29 mm., in width from 18 mm. to 25 mm., and in thickness from 4mm. to 8mm. Ventral sur- faces are uniformly flat and the dorsal surfaces have retouching on all three sides rising to a rounded keellike prominence. Slightly modified nodular end scrapers (pl. 27, A, a-1).—End scrapers of this type were made from pieces of jasper that still re- tained part of the natural unmodified outer surface of the rock. The shape of these flakes seemed immaterial to the one manufacturing end scrapers of this type. Some of the flakes are roughly rectangular in outline, others are triangular, and the rest are somewhat irregular in shape. From the table of measurements in Appendix 4 one will note that most of the flakes are about 6 mm. thick, but the range is from 200 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 4mm. to 10mm. Lengths of the flakes are very consistent, and the greatest variation occurs in the widths. Cutting ends have been sharply retouched either in a slight arc or straight across. Trapezoidal end scrapers (pl. 27, A, m, 0).—Scrapers of this type, 15 specimens, are roughly four sided; two of which are parallel. Of the total number, 9 have all four edges retouched while the remaining 6 have retouching only along three sides. Cutting edges can range in shape from a pronounced arc to almost a straight line. Size is fairly uniform. End scrapers with left carinate (type 3).—These 37 specimens are a form of keeled scraper in which the keel appears as a distinct flake scar ridge on the left side of the dorsal surface, creating the thickest part of the tool. All are medium sized and several of them have had the positive bulb of percussion chipped away to bring about a flatter ventral surface. Contrariwise there are 32 other specimens, type 4, that possess a dis- tinct ridge along the right side of each scraper. Whether these are distinct and purposeful tools cannot be determined, but there are far too many of them to be accounted as just accidental objects, or end scrapers. What was once a type 8 end scraper ended up by being an eccentric scraper ; not only were the edges retouched but the working edges were retouched on the ventral surface, creating a very pronounced bevel. The cutting edge now resembles similar edges found on gouges or chisels. Whether this particular tool was intended to function as a gouge cannot be determined, but it is well suited for just this purpose (field specimen No. 1008). End scraper-graver combination (pl. 27, A, p-w)—A few end scrapers had short graver tips at one or both sides of the working face. These were made by carefully chipping a small, fine, sharp point along one edge of the tool. The ventral face is flat, as was that of the scraper itself, and the edges are beveled and brought to a point. At the present time these points are not as sharp as they once were but have been rendered dull through apparent use. Evidently this trait is a carryover from an earlier cultural manifestation and was in its terminal phase at the Hosterman site, as represented by the small number of specimens. Modified end scrapers—(two specimens).—These are slightly tri- angular in outline with a pronounced groove down the dorsal surface, creating a moderately high ridge on the left-hand side of the scraper and one less pronounced on the right-hand side. The two side edges have been carefully retouched, terminating in a rounded butt, and the scraping edge was given additional retouching to give the ventral surface a beveled inward curve while the dorsal surface was carefully Riv. . Sur. Be ee OT HOSTERMAN SITE—MILLER 201 beveled to create a chisellike cutting edge. Perhaps artifacts of this type were used as a form of gouge. Twenty-three fragments of end scrapers are too indefinite to be classified into the various types found at this site. Side scrapers.—Side scrapers are flake tools. It appears that their essential characteristic is the presence of a scraping edge formed along one or more sides of a suitable flake as the result of secondary working or retouching. The ventral surface, or under face, is unworked as a rule. The scraping edge is produced by secondary flaking from the ventral surface upon one or more edges. They vary greatly in size and shape depending primarily upon the type of flake chosen to be con- verted into a side scraper. It would seem that any suitable flake, regardless of size or shape, could have had its edges or edge trimmed for use as a scraper. The working edge is somewhat convex in outline. Several of these flakes are very thin, long, with parallel sides. Most of the Hosterman site side scrapers are made from some sort of chalcedony, quartzite, chert, or jasper. The trimmed edge, in most cases, was formed by pressure chipping in which small fine chips were cast off. On the other hand, a minority was just the opposite in that large rough flakes or chips were cast off with no care taken as to placement of the chips, resulting in large, thick, uneven edged flakes converted into side scrapers. Usu- ally the larger scrapers were made of quartzite. Side scrapers from the Hosterman site have been subdivided into the following types: Small flake side scrapers, 111 specimens; medium- sized flake side scrapers, 15 specimens; large flake side scrapers, 3 specimens; lunate side scrapers, 30 specimens; triple-edged side scrap- ers, 6 specimens; prismatic side scrapers, 12 specimens; double-edged side scrapers, 18 specimens; and combination side scraper-spokeshave; 15 specimens. Small flake side scrapers —These were made from the small ubiq- uitous flakes found scattered at knapping stations within the site or on various dump heaps. The side edge was, as a rule, very carefully chipped with very fine retouching, but there is an occasional specimen whose scraper edge was made by casting off much larger and heavier chips. Measuring along the scraper edge, the greatest and smallest lengths are 36 mm. and 16 mm., greatest widths of the flakes utilized ranged from 31 mm. to 15 mm., and the thickness varies from 2 mm. to 11mm. Large flake side scrapers (pl. 27, B, a-c).—Three have the follow- ing measurements: Length Width Thickness 107 mm. 50mm. 9 mm. 63 53 17 122 66 29 202 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 These measurements would indicate that any flake over 60 mm. in length converted into a side scraper would immediately be placed in the large-flake category. Double-edged side scrapers (pl. 28, A, a).—Scrapers of this type have opposing sides retouched to scraper surfaces. Measurements show considerable variations: in length, 22 mm. to 67 mm.; in width, 14 mm. to 41 mm.; and in thickness, 4 mm. to 10 mm. Five of the specimens have a carinated dorsal surface, and the remainder have a fluted dorsal surface. Triple-edged side scrapers (pl.28, A, b).—Scrapers of this type have two sides and the adjoining end of a flake converted into scraping surfaces. The adjoining end can be convex or straight in outline. At the Hosterman site we have two subtypes depending solely upon size. Type 1 scrapers range in length from 67 mm. to 130 mm., in width from 30 mm. to 65 mm., and in thickness from 13 mm. to 16 mm. Type 2 scrapers range in length from 20 mm. to 28 mm., in width from 18 mm. to 29 mm., and in thickness from 4 mm. to 6 mm. Prismatic flake end scrapers (pl. 28, A, d@).—These tend to be tri- angular in cross section and have only one of the edges converted into scraping surfaces. They range in length from 22 mm. to 49 mm., in width from 10 mm. to 19 mm., and in thickness from 5 mm. to 11 mm. Lunate side scrapers —Lunate side scrapers have one convex edge that was retouched to a scraping edge. Most of the flakes when cast off had an edge roughly convex in outline and were converted into this type of scraper. Only the retouched edge was worked; the rest of the surface of the artifact was untouched. The size of the flake varies in length from 21 mm. to 61 mm., in width from 10 mm. to 44 mm., and in thickness from 4 mm. to 14 mm. Side and concave scrapers (pl. 28, A, ¢)—Scrapers of this type are combinations of a side scraper and a spokeshave or a concave scraping surface. Most of these have only one concave surface, but one of these scrapers has two concave surfaces. These are carefully retouched so that they are not the result of a false blow or breakage. The flake may range in length from 27 mm. to 78 mm., in width from 18 mm, to 42 mm., and in thickness from 5 mm. to 14 mm. Another form of side scraper is that which has two sides that join to form a dull point. Both of the sides are scraper surfaces. These have been called pointed side scrapers in that the two sides may vary in outline from straight to slightly convex. If the two sides are projected to a point, the angle formed may range from 13 degrees to 87 degrees. This diversity indicates that there is wide variance at which the two sides may join to form the blunt point. There is quite a latitude in the size of the flake utilized both as to length and thick- ness. The juncture of the two scraping surfaces may vary from a blunt tip to one that is somewhat squared (pl. 28, A, e-h). Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 203 Modified side scrapers (1 specimen).—This is a small chalcedony flake that had one of its edges carefully retouched. The opposing edge was not only chipped but rubbed to form a backing to the cutting surface (pl. 32, A, 0). This is an unusual treatment and one not often encountered at this site. OVOID BIFACES Several small bifacially chipped stone implements (?%) were sepa- rated out of the assemblage of stone artifacts. These are roughly oval shaped and resemble in a very general way the shape of a human patella or knee cap. In some instances the chipping has been care- fully performed and on others the object has been shaped by means of percussion chipping. Materials were either chalcedony or quartz- ite (pl. 29 B). Length: Mean of 34 mm., ranges from 23 mm. to 41 mm. Width: Mean of 25 mm., ranges from 15 mm. to 35 mm. Thickness: Mean of 9 mm., ranges from 6 mm. to 13 mm. They are not numerous and occurred at various levels within the site. MacNeish (1958, p. 117) reports that “ovoid bifaces occur in all cultural phases,” and, “They occur in all horizons so far found in eastern Manitoba and were very numerous at the Larter site.” BIFACE CHOPPERS Choppers are not numerous. Two are roughly circular in outline; one is ovoid; and two are ovoid but notched as for hafting. All bear bifacial chipping and are shaped by means of percussion chipping. Several have some retouching and appeared to be battered, a condition that may have resulted from use. Length: Mean of 98 mm., ranges from 72 mm. to 123 mm. Width: Mean of 77 mm., ranges from 63 mm. to 90 mm. Thickness: Mean of 29 mm., ranges from 21 mm. to 37 mm. These occur in all levels from 1.0 foot to 6.0 feet and are by no means diagnostic of the cultures represented. KNIVES Stylized knives are characterized by the presence of a double- beveled working edge. Other artifacts have been classified as being knives or having served temporarily as knives. Ribbon flakes with or without retouched edges have also been classed as knives in accordance with traditional usage and probable function. Nearly all of the knives from this site can be separated into five categories: (1) large four-edged roughly rectangular (pl. 29, A, d), (2) four-edged slim and with parallel sides and rounded ends (pl. 29, A, a), made from elongated vein chalcedony, (3) four-edged diamond shaped specimens (pl. 29, A, d, g, h), (4) bifaced blades of several 661-932-6415 204 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 forms mostly leaf-shaped (pl. 29, A, f), and (5) ribbon flakes (pl. 32, A). The four-edged roughly rectangular knives are made from elongated fragments of vein chalcedony. Most are fragmentary. One out- standing specimen (pl. 29, A, c) has a curved blade that is not rare but does occur less regularly than the straight-edged forms. The finished form was undoubtedly influenced by the peculiar nature of the material, for this vein chalcedony occurs as flat, thin slabs. A service- able knife could easily be manufactured by simply selecting a frag- ment of the proper size and retouching the desired edges to form a cutting edge. Some knives have only one cutting edge along the long side of the artifact. Others not only have this one cutting edge but the opposing edge sharpened, while others have all four edges chipped to a double bevel. Those with only a single cutting edge may have been hafted, or, when large enough, held in the unaided hand and used without a handle. Also present were several thin-bladed forms in which the chalcedony was chipped so that the original cortex no longer remained (pl. 29, A, e). Cutting edges are fairly sharp. Such artifacts when tapped with a slight blow emit a ring almost metallic in sound. Knives like this are much too thin and delicate to have been held in the unaided hand. They must have been hafted when originally used. The four-edged diamond-shaped variety is present but in no great number. Not only are the cutting edges worked but so are the faces. A form of quartzite and some chalcedony were used in the manu- facture of such knives. In cross section they are lenticular. Bifaced forms are of several shapes but most of them are leaf- shaped. Ribbon flake knives are well represented in the collection (pl. 32,4). Many show that they must have been struck from cores primarily as temporary knives. Not a great deal of work was ever expended on them. When they became too dull through use to function properly they were soon abandoned. The dulled edge or edges bear small nicks with very fine chips that were not the result of intentional chipping but were brought about by using the flake as a cutting edge. Some of the ribbon flake knives were pressure chipped on only one edge; others were chipped on both of the long sides from both faces, resulting in a double beveled effect. Those with only a single chipped edge are hard to distinguish from side scrapers, for both were fashioned by pressure chipping. Ribbon flake knives were made from chalcedony, quartzite, jasper, and chert. Some of the smaller knives are in the range of microblades and probably were hafted so that several were used in a common artifact. Riy. Bas. Sur. pete Pap. No. 38] HOSTERMAN SITE—MILLER 205 INCHES METERS Ficure 51.—Burin with arrows indicating position and direction of flaking. BURINLIKE IMPLEMENTS Burinlike implements have not been recognized as part of the usual cultural assemblage of the Plains. This does not exclude the fact that such implements were manufactured and used by the occupants of the Plains, for, in the past, they were mostly overlooked and very little attention has been given to other than well recognized artifacts of es- tablished categories. More and more attention is being given to spalls and to the so-called broken artifact as well as to the castoff material resulting from the manufacture of stone artifacts. In this mass of material there have been found many heretofore unrecognized artifacts that have enriched the cultural picture of the Plains. True, the source of the various types of stone cannot always be traced since the Plains have been subjected to glacial action; neither can it be told just how far the material used in the manufacture of stone material was car- ried. Unless there are undisturbed deposits nearby or within reason- able distance of the sites under investigation, we cannot be certain that it is of local origin. Therefore, artifacts must be classified according to the type of stone from which they were made without any reference as to source. This is true of an unusual chalcedony artifact recovered from Fea- ture 3, a cache pit, at a depth of between 2.0 feet and 2.5 feet from the present surface. This is a combination tool, scraper—burin * (pl. 28, B, and: fig. 51). One end has been shaped into a well-made end scraper and the opposite end has been altered into what has been classi- fied as resembling a type of burin known as an angle burin with a 5 Both Drs. H. R. Collins and R. Solecki have examined this artifact and have classified it as a true burin. 206 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 truncate concave retouch, or as the European classifier would say, “burin sur troncature retouchée concave.” To be classed as a burin (or graver) it is necessary that the object should have at least one graver facet. Burkitt (1956, p. 63) says: Let us consider what this means. When secondary work is done on a blade or flake as, for example, when an edge is to be trimmed, vertical blows are dealt on the edge, the flake or blade itself being held horizontally. This is in- variably the case, as a moment’s thought will show. In the case of a graver, however, a variable blow is dealt at the point of the blade or flake down the length of the implement which is itself held vertically. The resulting flake scar, which truncated the edge of the blade or flake, is known as the graver facet. Be- fore accepting a graver as such, therefore, it is necessary to be sure that there is at least one graver facet present, the existence of which can be demonstrated or not according to the position on the suspected facet of the negative bulb or per- cussion and the surrounding rings. According to the classifier, angle burins ... have trimming on the other side of the working edge to the graver facet. When the trimmed edge is at right angles or nearly so to the longer axis of the blade or flake, the term “transverse” is given. There its direction is inclined at an angle to this longer axis and the term “oblique” is applied. The trimmed edge itself in both cases may be straight or concave, and in the oblique variety also convex, but not of course in the transverse variety, as no working edge would result. If the student will draw for himself a transverse convex angle burin, he will at once see the truth of this statement. [Ibid, p. 65.] STONE PERFORATORS OR GRAVERS Perforators or gravers (pl. 31) were made by pressure chipping a flake on one end into (1) short needlelike tips by beveling the sides so that they met on the dorsal, or top, surface leaving the ventral, or undersurface untouched, i. e., they were retouched from one plane face only, and (2) larger and heavier tips were made by using this same pressure chipping method on larger and slightly heavier flakes. All graver tips are triangular in cross section. The smaller artifacts with graver tips varied in length from 18 mm. to 31 mm., in breadth from 15 mm. to 22 mm., in thickness from 3 mm. to 8 mm., and the tips themselves from 1 mm. to 4 mm. in length. The larger ‘specimens varied in length from 26 mm. to 48 mm., in breadth from 13 mm. to 24 mm., and in thickness from 4 mm. to 15 mm. Both types are within the limits of those reported by Roberts (1935, pl. 18) ‘from the Lindenmeier site in northern Colorado. Most of the flakes were modified along the sides and some’ onthe dorsal surface but the ventral surfaces were never intentionally altered or modified. | Small perforators or graver tips.also occurred on other tools. All were formed by this same type of retouch from one plane face. One of the lamellar flake side scrapers had a well-preserved graver tip Riv. Bas. Sur. iy. Bas, oor HOSTERMAN SITE—MILLER 207 worked at one corner of the blade (pl. 32, A,”). Several end scrapers show working corners, and all were definite graver tips. Others had this same feature, but on a slightly reduced measure. Whether the size was intentional or the result of prolonged use could not be deter- mined. It appeared to be intentional, but we could not be sure of this. If it was intentional, then it could possibly represent a graver variant. These tips were all worked at the ends of the broadest beveled face of the scraper and their shaping was an integral part of the shaping of the scraper edge and was not a later modification or an unintentional feature. Most of these show, under magnification, that they had received considerable usage and were worn somewhat at the tip. LAMELLAR FLAKE TOOLS A class of tools utilizing thin lamellar flakes of chalcedony or chert was found at the Hosterman site. These were either small knives or side scrapers (pl. 32, 4). Knives are represented by 3, c, d, e, k, l,n, 0, and g. Specimen 0 is unique in that it was pressure-chipped along three edges from both sides with the exception of the wider bottom or basal edge. Later the S-shaped edge was abraded so as to partially obliterate the scarified edge, enabling the artifact to be either hafted into some sort of slotlike aperture in a bone or wooden handle or to facilitate its use without the use of a handle. Specimen ¢ received a double beveled edge along three of its edges. Chipping was very carefully controlled and is very uniform in size. Specimen 7 is a combination tool, a knife, scraper, and graver. All edges have been worked. At one corner is a small graver tip which under magnification displays much finer chipping, attesting to the fact that it was functional. The best of the lamellar flake tools is represented by specimen d. This was struck from a core of quartzite and is roughly rectangular in outline. All edges have been carefully chipped creating a bevel- edged implement. Earlier flake scars are shown on the dorsal surface of the implement. Specimen 6 (pl. 32, B) is the larger of these tool types and is prob- ably a part of a much larger tool. The sides are roughly parallel and the long sides have been given a bevel sloping upward to the top or dorsal surface. The beveled edges are more prominent on this artifact than on others in the collection. With the exception of specimen a they fall within a range whose lengths vary from 28 mm. to 49 mm., breadths from 13 mm. to 22 mm., and thicknesses from 2 mm. to 6 mm. Specimen a measures 56 mm. in length, 22 mm. in breadth, and 4 mm. in thickness. Gravers, as a rule, consist of chance flakes modified in that the needlelike tips were formed on one side or an end. Any flake was 208 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 suitable provided it appealed to the worker. These could be chosen from the flaker’s discard pile or at random from the surface of the site. The small tips did not occur just by chance but were purpose- fully made so that the under side is always flat and the two upper sides are beveled so as to bring the tip to a very fine point. Gravers can be distinguished from drills in that gravers are retouched on only two sides and drills have all sides retouched. On top of this a graver can have one or more such tips. GROUND AND PECKED STONE TOOLS Tools in this category were formed by grinding their surfaces or by pecking. The pecked stone tools include pebble hammerstones, full-grooved mauls, anvil stones, and paint grinders. Ground stone tools include slabs of stone with grooves worn into them, anvil stones, and some rubbing stones, besides a grooved ax. Anvil stones.—These are irregular nodules of granite, quartzite, or other tough crystalline rock. Many of them have been subjected to much abrasion. They were found scattered throughout the site. It has been suggested that anvil stones were used to pulverize dried meat, seeds, and berries by pounding rather than grinding and in so doing the anvil stone was “probably set in a sort of rawhide ‘pannier’ and the food substance placed on it to be crushed with a stone-headed pounder. This was the method followed by the Dakota, who in some instances at least used anvils almost identical to those of the Pawnee” (Wedel, 1936, p. 75). Since this site is neither Dakota nor Pawnee, it has been attributed to a pre-Arikara—Arikara group in whose cultural assemblage there is comparable agreement. Rubbing stones.—These are usually flat, round stones, and range in size from small pebbles to those easily handled in the fist. They were used as a rule without any retouching and were probably used in rubbing down hides as several have acquired a rather high polish and are very soft and smooth to the touch. Several show very shallow parallel scratches resulting from use. Pecking stones.—These are small pebbles of granite, quartzite, dio- rite, or some such hard stone. Usually, these are unaltered pebbles; some are round, others are natural elongated forms. Both forms show much battering or abrading. Similarly, others that are small, flat, and circular are worn about the edges. Paint stones.—These paint stones are usually highly granular stones that were chosen because of this trait. Several show that they had been used to reduce lumps of hematite into powder by grind- ing. Specks of hematite have worked themselves well into the sur- face of these stones and cannot easily be rubbed off. Paint stones are not plentiful and are widely scattered in the site. Riv. Bas. Sur. fake Pap. No. 35] HOSTERMAN SITE—MILLER 209 Grooved mauls—These mauls are represented by two specimens. They are made of granite boulders, roughly shaped, smoothed, and grooved about the center for hafting. The smaller and squattier speci- men is oval in cross section while the larger is somewhat trianguloid in cross section. In actual weight there is a very little difference. The smaller specimen weighs exactly 4 pounds 6 ounces and the larger one weighs 4 pounds 12 ounces. Neitherisbadly battered. Bothshow that they had received considerable usage, probably in the crushing of dried meat, nuts, or berries. Exterior surfaces are finely dimpled (pl. 30, B). Grooved stone ax (pl. 30, B).—Only one stone ax was recovered and this is not a complete specimen, for the butt end is missing. Appar- ently this is an unusual type in that it has a groove completely en- circling the implement separating the blade from the butt. During the life of the implement the butt section had been completely broken away and another groove was being pecked ahead of the older groove which was never finished, for the blade became broken and the tool discarded. The tool was well fashioned and finished. The blade is still fairly sharp. This particular ax was made of a greenish-colored diorite, a stone which does not occur locally. It must have been traded into this region from the general Southwest for it resembles those occurring in that region. In its present form it weighs 2 pounds 4 ounces. Both fragments were found in Feature 9, a cache pit, at a depth of between 7.0 and 7.5 feet. Pumice abraders—Pieces of waterworn pumice or naturally burned lignite of various shapes and sizes, some less than an inch across up to pieces as large as grapefruit or larger, were utilized as abraders. Several have faceted sides, others have narrow shallow grooves, and still others have wide and deep grooves. The shallower grooves were probably made from sharpening pointed bone and wooden imple- ments while the wider and deeper grooves were probably used and made as arrowshaft straighteners and smoothers. Sandstone abraders.—Bits of sandstone, of various degrees of coarseness, show one or more surface grooves resulting from the rub- bing of honed implements across them. Some are irregular bits of sandstone and others appear to have been shaped into rough rectangu- lar blocks and used as hones or specialized whetstones for sharpening bone awls or smoothing arrowshafts. Arrowshaft straighteners.—Implements for this purpose were made of sandstone and display one or more grooves running the length of the implement. Usually they are rectangular in shape, but some are triangular in cross section. Whether these constitute distinct types or just individual likes as to shape must be taken into consideration. Some may have liked the rectangular form, others preferred the tri- 210 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 angular, while still others were not too discriminating and used what- ever shape of sandstone came to hand. Probably too much stress has been brought to bear upon implement shape by both the profes- sional archeologist and the amateur. They both fail to take into con- sideration that they are not only dealing with inanimate objects but with human personalities as well and it is the latter that are the real determining factor. Personal likes and dislikes play a distinct role in the implement form and use according to the type of material to be utilized. Handstones.—Several roundish stones were recovered not only from cache pits but from the general site. These have been classified as handstones because they display a certain amount of wear. Five of these are outstanding in that they were pecked and then smoothed. They are roughly circular in shape and resemble, to a certain extent, stone balls. None of the edges have been battered even though they show considerable wear, which would appear to indicate that such pieces were never used as hammerstones but were probably used for pounding a resilient substance such as softening a dried up animal hide, dried beef during the manufacture of pemmican, or even soft berries. One specimen, in particular (field specimen No. 2014), ap- pears to have had a fatty substance driven into the cortex of the stone giving it a distinct color and feel. Maximum lengths of these hand- stones range from 55 mm. to 81 mm. Minimum lengths range from 51mm. to 75mm. Maximum thickness varies from 34 mm. to 57 mm. Another form of handstone is represented by a single specimen. It is disk shaped and almost circular in outline with opposing faces. One of its faces is almost flat and the other is a flattish dome with a shallow central pit sunk into it. The pitted face shows no evidence of having ever been used as a grinding tool. The pit probably served as a finger hold. Maximum and minimum lengths are 52 mm. and 50 mm. Maxi- mum thickness is 34 mm. Utilized pebbles—Small river hardheads were among the stones present within the site. Most of them are oval in outline and some- what flattish. In many cases the ends of the long axis had been pecked or show that they were utilized as either light hammerstones or used to peck away irregularities on the faces of other stones to be utilized as tools. Then too, there are those that have one of their flat faces rubbed very smooth and somewhat polished. Such tools were probably used in the manufacture of pottery since they could have easily functioned as an anvil opposite the paddle used in shaping clay into vessels. Others are egg shaped with a tendency towards flatness. Both of the flat faces bear a number of parallel scars at right angles to the main axis and cover an area roughly 27 mm. in diameter. These surfaces Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 211 were smoothed prior to scarification. Their purpose and function are not known. Practically all of the rock, of any size, found within the site was put to some use at one time. Not all are recognizable tools. Some of the larger fractured pieces may have served as a form of chopper; others are hammerstones, etc. HEMATITE AND OTHER PAINT SUBSTANCES During the partial exploration of the Hosterman site, several worked nodules of hematite were found with faceted faces. These must have been rubbed to reduce them to a powdered form to be used as a paint. Then, too, a couple of lumps of a whitish material were found along with a lump of a creamy whitish material and a lump of powdered red ocher. These substances were found mostly in midden pits, Features 9 and 22, as well as in small cache pits within the circular house pattern of Feature 22. Since no paint brushes were found during this period it raises some doubt as to how this paint was applied. Various sherds had paint stains on their interior surfaces. That this red paint was put into these sherds as a container or as a crude sort of palette seems very likely but it cannot be proved that such was ever the case. Several of the bone implements bore reddish stains as though they had come in contact, at one time or another, with red paint. Catlinite—Four fragments of catlinite pipes were recovered from Features 10, 16, and 22. The largest of these four pieces is about half of a bulbous pipe bowl. The bowl rim has been brought to a sharp edge rounding somewhat as the wall proceeded downward. ‘The basal section of this fragment is covered with numerous vertical parallel thin scratches and the major part of the exterior has been smoothed and polished. The smaller fragment of a pipe bowl appears to. be tubular in form. It, too, is covered on the exterior with a number of scratches that form a rough hatchure. One of the broken edges had been reworked, and four shallow grooves partially cross the thickness of the wall. The lip of this particular pipe is flat and bevels slightly outward. The other smaller fragments are terminal sections of pipe stems. Some work was started to reshape the larger portion but this was never completed and the section was discarded. A large section of rough catlinite was recovered that has one cut surface. Shallow scratches cover the surface of this piece. A number of its edges have been rubbed but still the piece does not conform to any known object. It is just one of those partially worked pieces which appear to clutter up the site without indicating their purpose. A small well-rubbed piece of catlinite, 21 mm. in length, 9 mm. in width, and 8 mm. in thickness, was found at a depth of between 2.5 212 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 feet and 3.0 feet in the village midden. Whether catlinite was ever reduced to a powder and this product used as a body or facial paint or dust is not known to the writer but this piece resembles similar pieces of rubbed stone which were put to this use. One other small flat nondescript piece of catlinite completes the collection. Belemnite—Sections of belemnite, a conical fossil shell of an ex- tinct cephalopod, were recovered from two of the cache pits, Features 9 and 13, and from various isolated midden areas within the site.. The blunt spicule ends of two specimens were encircled as though some sort of attachment was added so that they could be suspended. Whether these were worn on the person or attached to some sort of garment could not be determined. Three other sections were unworked. BONE ARTIFACTS Several types of bone tools were found in considerable quantity at the Hosterman site. They were particularly abundant in cache pits, midden areas, and to some extent in house fills. Preservation was exceptionally good even though many were broken by the aborigines themselves during the period of occupation. Normally, bones from bison, deer, and pronghorn were converted into major types of tools but the bones from other types of animals were converted into other types of tools whenever the need arose. Cut animal bones.—Several of the articulative ends of deer, prong- horn, turkey, and other bird bones were found that were severed from their shafts. A V-shaped cut was made almost into the medullary canal, and with slight effort this section was separated from the shaft. These ends are the waste from the manufacture of tubes and beads (pl. 33, A). The most outstanding bone tools are those manufactured from the scapulae of bison. In this group are scapula hoes, digging tools, knives, scrapers, hide scrapers, thong stretchers, and even sickles. Scapula hoes—tThis ubiquitous Plains implement was quite preva- lent at the Hosterman site. Scapulae of bison were not greatly modified in the process of manufacturing these implements (pl. 33, B). The articulative ends remained unaltered, and there was no evi- dence that any hafting was ever attempted. The making of a hoe necessitated only the removal of the spine and the posterior border (postscapular process and acromian), from the exterior surface to about the level of the fossae, thus creating a more or less flat imple- ment. Before the scapula could be put into use as a hoe the working edge (suprascapular border) must be trimmed and sharpened. On some the areas from which the postscapular process and acromian had been removed were left relatively rough. On others this was smoothed Riy. Bas. Sur. — Pap. No. 36] HOSTERMAN SITE—MILLER 213 down. Subsequent usage would eradicate any evidence of working on the suprascapular border, for it would quickly take on new char- acter and the area would be given a high polish by soil abrasion, the blade would become beveled and in a number of instances deeply scored. From the evidence at hand, one might say that scapula hoes continued to be used until they became exhausted and quite short; they were then discarded as no longer serviceable. The mortality rate was rather high among scapula hoes. When this occurred the thinner interior sections, as well as other sections of unworked scapulae, were converted into many other types of tools. Knives, hide scrapers, and possibly thong stretchers are among some that can be listed. Scapula spines—tThe spines were removed from the shoulder blade when the blade was converted into hoes, and the spines themselves converted into digging tools. The sharp picklike distal end readily lent itself to this purpose. Wear was never excessive on any of the Hosterman specimens and apparently they were only expedient tools and were quickly discarded (pl. 34, B, b). Metcalf (1956, p. 306) considers implements such as these to be a “type of awl that has not been previously recorded from the Plains.” The Dodd site gave up a number of tools made from these scapula spines. Those being reported from the Hosterman site were re- covered from the diggings by pot hunters in a midden area prior to our work there. Scapula knives.—Knives of various shapes and sizes were made from thin sections of bison scapula (pl. 35, 4). Some are roughly oval in outline with one or more edges sharpened; others are rectanguloid shaped; still others are roughly diamond shaped. Practically all have acquired a very high polish even though there is considerable variation in the overall lengths. The shortest specimen measured 12.5 cm. in length; the longest measured 19.6 cm. in length. Most knives were made from the flat section between the acromian and postscapular process; others were derived from the postscapular section of the bones having one exposed edge of cancellous material opposite the working edge. Constricted sections may have served as handles. Their period of usefulness was evidently short, for many fragmentary knives were recovered from the site. Scapula cleavers—Several knifelike cleavers are in the collection (pl. 86, 4). These were fashioned from bison scapulae by cutting lengthwise through the postscapular fossa so that a long compara- tively thin blade was created. The back edge of the cleaver is con- siderably thicker and a portion of the acromian or postscapular process was left intentionally for the purpose of strengthening the implement. Like the knives, the tapered end served as a handle. 214 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Among the fragmentary forms were those that were spatulate in outline and comparatively thin. These were made from the post- scapular fossae sections with the sides ground to a thin beveled cutting edge and the ends neatly rounded. Scapula scrapers.—These were made from fragments of the post- scapula fossae sections. Most were worked from fragments into rec- tanguloid shapes. Those that were made from sections of the post- scapular process and acromian were much thicker and required much more work to complete. Others were cut so that only a fractional part of the scapula spine formed one edge (pl. 35, B). Scapula hide scrapers.—Like other scapula scrapers, these were cut from sections of bison scapulae. Most of them are either from the posterior border and display scars where the ridge was left intact or partially obliterated or were cut so that only a fractional] part of the scapula spine formed one edge. Basically, these tools are rectanguloid in shape with a constricted section forming a ready handle. The working portion of the blade has parallel sides and the cutting sur- face was given a pronounced bevel over the entire edge. They roughly resemble the blade in a modern carpenter’s woodworking plane in shape (pl. 85, B). On the backs there appear series of shallow parallel scratches which have been acquired during the life of the implement. The specimen shown in plate 35, B, is 13.0 cm. in length, 6.4 cm. in width, and is the best of the type found at the Hos- terman site. Thong stretchers——A single thong stretcher was found. This was made from a small thin section of a.bison scapula (pl. 34, B, 6), whose working edge was given a pronounced U-shaped notch. The edges and sides of the notch acquired a very high polish and the sides were well smoothed. The specimen shown here is 7.5 cm. in overall length and 4.1 cm. in greatest width. The notch itself measures 3.2 cm. across the opening, 0.7 cm. near the base, and is 1.2 cm. in depth. Scapula sickle-—What appears to be a fragmentary scapula sickle was recovered from Feature 25, a midden pit. Metcalf (1956, fig. 103, b), features a complete sickle derived from the floor of a rectangular house at site 39L.M3, Lyman County, S. Dak. This fragment appears to be a section of the handle with its adjoining notch where the blade joins the handle. From the base of the handle up through the notch this object displays considerable wear, accompanied by a high polish. ‘It is 25.0 cm. long. Another formative sickle is illustrated in plate 34, By a. Flakers-—F our flakers were made from bison rib bones. The heads of the tools were rounded and the tips were brought to an abrupt dull point. ‘These are distinctly triangular in cross section and one side appears to have been smoothed, for it has a smoother feel than does Riy. Bas. Sur. Fees: Pap. No. 35] HOSTERMAN SITE—MILLER 21 5 the opposite side. The cancellous tissue is still in evidence. Lengths of these four flakers are 72 mm., 64 mm., 50 mm., and 32 mm. Needlelike tool—aA very thin sliver of bone was converted into what appears to have been a needle. It is incomplete in that the butt, or eye, portion is missing. This object is 75.5 mm. in length and at the broken base it measures less than 2 mm. in diameter. Needles of this sort have not been described from sites in the Plains of South Dakota. Whether this is an intrusive tool from North Dakota or out of Minne- sota cannot be determined. Shaft wrenches——Shaft wrenches of various lengths were made from sections of bison rib bones. Most of them have but a single hole, and these tend to be oval in shape with the sides beveled from use. Ap- parently the wrench was held at a slight angle to the shaft in the straightening process. In one the hole is beveled in three directions. This is not the rule but an exception. Some of the rib bones are highly polished. Very few were complete, for most of them had been broken and apparently discarded by their owners. This fracture tendency is at right angles across the hole dividing it in half. An incomplete wrench was attempted. On one flat surface of a rib fragment there appears a conical pit 3 mm. in depth and 8 mm. across the top that had penetrated the cortical bone and started into the cancellous portion of the core. Drilling was attempted only on one side. There was no evidence that any attempt was made to drill from the opposite side to meet this section of the perforation. Knife handles—Knife handles were made from segments of bison ribs. These were grooved along one edge sufficiently deep to allow for the insertion of a stone knife blade. A complete knife was re- covered from a shallow cache pit beneath the floor of Feature 22, a circular house (pl. 36, B). Other handles and blades were found unjoined. Hide grainers.—Sections of bison humeri and femurs were sectioned so that a wide expanse of cancellous tissue was exposed. Four speci- mens were recovered from the Hosterman site. These were used during the process of dressing down hides into usable pelts. Notched ribs—Two fragmentary ribs have opposing notches at the ends at right angles to the long axis. These were meant to receive either string or gut attachments. Not having heard that any of the Plains Indians ever made use of a bullroarer, it is suggested that these two objects could easily have functioned along this line. This could have been some boy’s toy. Cut ribs—Numerous sections of ribs were recovered that were un- doubtedly either unfinished and broken artifacts or rejects. Some had acquired a high polish. Punches.—In Lehmer’s definition (1954, p. 66) he describes his Group 1 type as: “Irregular plates split from bison or elk ribs, can- 216 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 cellous tissues of split surface sometimes ground down, sometimes unmodified; edges unmodified; butts rounded; tips worked down to a blunt point.” Two similar specimens were found at the Hosterman site. Both of them follow a common configuration, but one has its edges smoothed and polished and the other has untreated edges; one has - a tip that was brought to a modified point, and the other has a spatulate point. The complete specimen, except the extreme tip, measures 168 mm. in length, and 25 mm. in width across the base, and has an average thickness of 6 mm. Scapula hoes with deep U-shaped notch.—This notch which was worn or cut into the bit end may have functioned, under certain con- ditions, as a type of thong stretcher. The inner edge of the longer spur has been beveled while the edge bordering the shorter spur is not quite as sharp (pl. 33, B,a). Worked pronghorn metapodial bones—One complete and one frag- mentary pronghorn metapodial bone showed signs of having been used. The complete specimen was from a young individual, as shown by the incomplete joining of the epiphyses (ankylosis) to the shaft of the bone. The distal end was drilled. The shaft is lightly scarified by a series of parallel scratches running at a diagonal to the main axis of the shaft. Just above the basal articulative surface, on the shaft, is another series of shallow cuts. These cuts are at right angles to the axis of the shaft. The purpose and use of these scratches are unknown. Fishhooks.—Four fragmentary and one unfinished bone fishhook were recovered during the process of uncovering remains at the Hoster- man site. All were made of a very composite bone. With the ex- ception of the unfinished hook, the others were round in cross section. All are of the unbarbed type and display excellent workmanship. The unbarbed arms are pointed and scored with a series of shallow parallel incisions at right angles to the main axis of the arm. The shank arms also are scored near the ends, probably to prevent what- ever tie was used from slipping off. The unfinished hook is from a section of bison long bone, demon- strating the method by which this sort of implement was made. Worked animal scapula—A small section of what appears to be the scapula of Canis sp. has two small holes just 9 mm. below the posterior border that were punched through the thin section of the blade. The two holes are 10 mm. apart and are not clean cut. None of the edges of this bone object were smoothed or worked. There is some question as to whether this particular bone object was perforated for suspension as an ornament or for some other purpose (pl. 37, B). Pentagonal-shaped bone tool—A large pentagonal-shaped bone Riv. Bas. Sur. mae Pap. No. 35] HOSTERMAN SITE—MILLER D7 tool (pl. 36, A), prepared from a bison’s scapula, shows considerable wear in that the surfaces as well as the bordering edges were polished. It measures 9.8 cm. across the base; the lateral sides are 16.4 cm. long, and the tapering sides average around 11.1 cm. in length. Both the tapering sides as well as the lateral edges show that they have received considerable wear. On both faces, bordering the sharp working edge, are series of thin shallow scratches running almost the entire length of the implement. Superimposed over these, on one face, is a roughly ovoid area that is highly polished, a gloss resulting from long-continued use. This ovoid area measures roughly 6.5 cm. in length and 4.0 cm. in width, and is located 3.3 cm. from the base of the implement; it was probably the position of the thumb while the tool was in use. It is definite that this was never used to dig into the soil, for the edges of this thin object could not have with- stood such use. Its actual use cannot be determined. Similar ob- jects have not been reported from this section of South Dakota. Bone tubes——Twenty-eight bone tubes of various lengths were recovered from the Hosterman site. Out of this number only one was fragmentary. Fourteen were made from cylindrical sections of bird bones. Most of them had their cut edges polished and smoothed, but few of them were completely polished. They ranged in length from 28 mm. to 119 mm. (pl. 33,4). The other 14 were from sections of mammal bones, and practically all of them had their cut surfaces smoothed and had developed a good polish. None of the bone tubes were decorated in any way. The lengths ranged from 56 mm. to 128 mm. On the basis of lengths, the bone tubes were separated into three groups: Group 1, 5 specimens: Lengths: Range from 28 mm. to 49 mm., mean of 38 mm. Group 2, 15 specimens: Lengths: Range from 53 mm. to 94 mm., mean of 73 mm. Group 3, 7 specimens: Lengths: Range from 102 mm. to 128 mm., mean of 115 mm. Bone awls.—Various types of bone awls are present at the Hoster- man site (pl. 37,4). One of the prevalent types comprises those made from split deer or pronghorn metapodial bones with split distal ends forming the butt. The spit surfaces on most of the specimens were ground. and smoothed and the cancellous material removed, leaving a U-shaped shaft to the tool. Both the shaft portion and the articu- lative surface had acquired a high polish through heavy use. Lengths varied from 58 mm. to 100 mm. This group easily falls within the limits of similar awls reported from the Dodd site, and Swan Creek site, and is ubiquitous in the Plains. Nineteen specimens are repre- sented. 218 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Rough splinter awls.—Seven specimens. Awls of this group are comparatively few in number. They were made from rough, irregu- lar splinter sections of bison ribs and long bones as well as some of the long bird bones. They are unworked with the exception of the tip, which is ground to a smooth point. Lengths varied from 31 mm. to 118 mm. A combination tool, made from a splinter of a long bone of bison, consists of an awl and a spokeshave. This tool, 19.6 cm. in length, was brought to a sharp point at one end; the other was broken off squarely. Forty-six mm. from the base there is a wide notch that functioned as a spokeshave. On either side of the notch the tool was well smoothed. Under magnification there is a series of parallel grooves within the notch at right angles to the axis of the bone, indi- cating that this was a working surface. Other splinter awls were made from sections of shaft of long bones, mostly deer or antelope. These were well worked over most of the surface. In cross section some are round, others are on the flat side, and still others are shallow U-shaped. One of the specimens has a V-shaped cut on three of its sides near the base as though its owner intended to reduce the overall size of the implement and rid it of its irregular rough butt, but the task was never completed. Lengths range from 55 mm. to 131 mm. Specimens from another group were made from the edges of bison ribs cut so that a portion of the cancellous tissue was still included. The cut sides were carefully smoothed and the butt ends were finished off either by working into a rounded, squared finish or bringing them to an abrupt, dull point. In some, the tips were brought to a gradual point; in others the tips tend to be more abrupt. In cross section the awls of this group tend to be triangular. Lengths range from 57 mm. to 144 mm., the majority being around 90 mm. in length. In the next group, represented by seven specimens, the awls were made from edges of bison ribs with portions of the cancellous tissue included. As in the group immediately preceding, the awls received about. the same treatment to convert the raw material into workable tools. Instead of being triangular in cross section, these tend to be more circular with tips brought to a more abrupt point. Cancellous bone balls—Three small ovoid-shaped cancellous bone balls were recovered. Two of them were found in Feature 9, a cache pit. One was found at a depth of between 6.5 feet and 7.0 feet; the second came from a depth of between 7.0 feet and 7.5 feet. The third ball came from one of the small isolated midden areas at a depth of between 2.0 feet and 3.0 feet from the present surface. Whether these were the remnant portions of bone abraders, or were actual abraders intended for some much finer work, or were a form of toy could not be determined from the evidence at hand. Riy. Bas. Sur. As Pap. No. 35] HOSTERMAN SITE—MILLER 219 Bone picks—A form of pick was made from the shaft of a bison radius by splitting it lengthwise, leaving a part of the proximal articulative end intact and sharpening the other end. The one speci- men that was recovered is badly battered. At the present time it measures 17.1 cm. in length, which is far from its original length. One edge has been smoothed to within 8 cm. of the articulative end. This smoothing was probably not intentional but acquired through use. Split pronghorn metapodial bones.—Pronghorn metapodial bones were split lengthwise and the halves were used as beamers. The bone was sawed lengthwise in order to split it into two equal parts. The distal end displays additional saw marks that were not carried to completion. Awl blanks were similarly constructed from the split- ting of metapodial bones. Perforated rib sections.—Sections of small ribs were perforated at both ends. These were probably used as ornaments of dress since they were attached to objects, or they may have been worn as pendants. In the collection there is one complete specimen and fragments of five others (pl. 37, B). Punches.——Punches were made from the cutoff or broken ends of deer or elk antler. The tip ends of these show definite signs of having been sharpened. Several acquired a fairly high polish. Antler sections Sections of antlers with their prong tips missing definitely show the tips were severed from the main portion. To do this the prongs were sawed or cut almost through and then broken away from the stem. In one instance the prong was sharpened before it was severed from the stem for it must have been easier to do it this way than to do itafterward (pl. 36, B). Split antler section—A small section of deer’s antler was cut and broken off and the ends left unworked. Later it was split down the center and most of the cancellous tissue removed. The long edges were then smoothed and evened up, leaving a sharp outer edge. From the looks of this artifact it would appear that it was intended to be used as some sort of scraper-smoother. Mineralized bone—A small fragment of mineralized bone was found in Feature 22, a cache pit. It apparently was never used by the occupants of the Hosterman site. It was probably either picked up from the surface outside of the village and brought in as a sort of curiosity, and lost or discarded afterward; or, it could have appeared naturally on the surface of the site and later gathered up along with the rest of the trash and dumped into the midden pit. Cut antler fragments.—Two large basal segments of deer antlers were made into percussion instruments or hammers by having one end slightly rounded and the other roughened. One was 17.0 cm. long 661-932-6416 220 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 with a diameter of 3.7 cm., the other was 19.4 cm. long with a diameter of 4.6 cm. FETISH OR TROPHY SKULL The skull of one of the Plains kit foxes (Vulpes velox) was found in the midden of Feature 2 at a depth of 1.8 feet. This fragmentary | skull (pl. 38) is peculiar in that there are, at the present time, four small round perforations with beveled edges occurring in the oc- ciputal region penetrating the brain cavity. The diameter of these openings is 4 mm. on the outside, tapering somewhat until they en- tered the interior of the brain case. They were spaced so that they covered this area rather well. The central perforation just touches to the right of the sagital suture, two others are equidistant on either side of this suture about 1.25 cm. distant, and the fourth open- ing is just above the right auditory meatus. Whether there ever was a comparable opening above the left auditory meatus is not known, since that portion of the skull is missing, but it would seem plausible that such was the case. As we do not know exactly why this particular skull was treated in this manner, we have assumed that there must have been some religious ritual significance attached to it. We do know that the Mandans and other Plains Indian groups utilized the skulls of bisons and humans during certain ceremonies. Then, too, each man had his own medicine bundle usually wrapped in some animal pelt. Whether the skull of this particular fox was especially valued would appear likely since so much work was performed on it. SCORED BIRD STERNUM A fragmentary sternum of Aguila chrysoelus canadensis (golden eagle) was found in Feature 3, a cache pit, at a depth of 3.0 feet from the present surface (pl. 37, B). Across the right keel and along the posterior margins are numerous sharp and shallow incisions. There is no definite pattern these scars take. It would appear that they were the result of cutting away the heavy fleshy parts during the term of a meal. Whether eagles were obtained just for their tail feathers and the flesh eaten afterward is not known; but it would seem very probable that the fleshy parts of the bird were consumed. WORKED TEETH The front half of the post-lingual surface of a split beaver’s incisor, recovered from the floor of the circular house, Feature 22, shows cut- ting marks on the concave surface, the beveled cutting surface, and a shallow cut at the base of the tooth. Similarly prepared beaver in- cisor teeth were reported from the Dodd site in South Dakota, from Riy. Bas. Sur. _— Pap. No. 35] HOSTERMAN SITE—MILLER 291 various sites in Minnesota, Ohio, and North Dakota, and they appear to be common throughout the Northeastern United States as well as in some of the Southeastern sites. This is a specific trait of the Point Peninsula Focus (Ritchie, 1944, p. 117), which appears to relate some- what to the cultural horizon represented at the Hosterman site. A heavily eroded wapiti’s tooth (Cervus canadensis) was found in Feature 34. It had been perforated for suspension and was prob- ably used as a pendant worn around the neck of some individual. It is the only specimen of this sort found at the site. Human remains are represented by 5 incisors and 1 canine tooth. The five incisors were found in a small pit along with other debris, but no human bones were present. The single canine tooth was found in a nearby midden pit. All of the six teeth were greatly worn, having been reduced to about the gum line. All were flat across the top. Adhering to the sides of the roots of the five incisors was an osseous growth, known as cementosis. The amount of this growth or deposit was not constant; some teeth have more of it than others. This is an apparent indication of age, for all five teeth appear to have belonged to a single individual. SHELL Shells were not numerous at the Hosterman site. Several large mussel shells were found thinly scattered throughout the midden, some appeared in midden-cache pits, and a few came from house areas. Dr. J. E. P. Morrison, Department of Zoology, U.S. National Mu- seum, identified four species of shells from this site. They are: Las- migona complanata (Barnes); Anodonta grandis plana (Lea); Oli- vella biplicata (Sowerby) that was introduced from California; and Succinea grosvenori (Lea). Usually these shells are unworked. Sec- tions of Lasmigona complanata were cut into triangular, rectangular, and other simple geometric forms, though none were perforated. Other Lasmigona complanata were converted into spoons or scrapers whereas portions of others were shaped to disk pendants that were perforated for suspension. A small section of a serrated shell frag- ment was found within the midden area at a depth of 2.5 feet to 3.0 feet. The margins on either side of the serrations have been rubbed smooth, while the basal section holds the fractured break. Apparently this was some form of pendant. TABLE 6.—Shell identified Species Number Artifact Source Anodonta grandis plana (Lea) --------------- 6 | Unworked. <._--~2-.----.--- Local. Lasmigona complanata (Barnes) _._---__-__-- 21 | ee GOeea aac e Aate eee coees Do. Oe ee ee ee ee ee JANI WiOLKCUl. onan os ceeoe oe euanee Do. Olivella biplicata (Sowerby) ---.-------------- Dh WBeRGe ses o5 coc ot oe Soe California. ead Succinea grosvenori (Lea)..-----.------------ 2) UMWORKOG > seme coeenoeeeas Local. 992 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 A shell disk, 34 mm. in diameter, is incomplete. Near the outer margin isa small perforation for suspension. Presumably the pendant was made from a section of Lasmigona sp., for this type of shell was more numerous than any other and the texture fits this sort of shell. COPPER A single small copper bead, 16 mm. in length and 3 mm. in diameter, was found near the mouth of Feature 16. It is the only metal found during the period of excavation. The bead was made by rolling a small sheet of copper around a twig or some similar object. At the present time its surface is completely covered with a verdigris. From its location in the site, it is presumed that it was intrusive there either after the site was abandoned or at the terminal phase of occupancy. SUMMARY AND CONCLUSIONS The existence of rectangular house pits suggests that a group of the Upper Republican culture had established at least one permanent resi- dence here at the Hosterman site. True to their nature they spaced and scattered their houses widely apart and without pattern. Their economy consisted of limited agriculture with greater stress being placed on hunting and seasonal gathering. The presence of lim- ited amounts of charred corn and a few beans and squash seeds attests that they had enriched food habits. Pits of wild plums and choke- cherries show that these fruits were then in use. There is evidence that they were supplemented, in season, with the wild berries of the nearby area as well as with seeds from the many wild grasses growing there. Communities were located on bluffs or terraces and, in this case, the main stem of the Missouri River. During this period, house pits were rectangular in form and were sunk to various depths, depending upon the inclination of the builders. This variation in depth probably accounts for the uneveness of the aeolian deposits that accumulated over the remains after the site was finally abandoned. Inside the pit was constructed a house with indi- vidual posts fairly closely spaced, surrounding four centrally located roof supports that in turn surrounded a centrally placed firepit. The walls served as studding to support the wall plates. Rafters were closely spaced, running from the wall to the center of the structure. Over this basic framework were piled layers of brush, grass, and finally a heavy dirt cover. The finished house resembled a low dome-shaped earthen mound which was entered by means of a long, narrow, covered passageway. Smoke escaped from a centrally placed opening in the roof. Riv. Bas. Sur. plas Pap. No. 35] HOSTERMAN SITE—MILLER 223 Hunting played a paramount role in their economy. With limited farming and plenty of meat these people were assured of an abundance of food. There must have been surpluses that required storage. To meet this problem they dug storage or cache pits that were bell-shaped, narrow at the top and widest at the base. Most of these pits were placed outside of the houses, but smaller examples were found beneath the floors of some houses. The large pits were rather deep and were capable of holding vast quantities of supplies. From the bison, and to some extent, the lesser animals, they ob- tained the bone material out of which they fashioned awls, flakers, bodkins, fishhooks, hoes, knives, scrapers, arrow-shaft wrenches, sick- les, cleavers, and many other forms. The essential forms of many of the artifacts had carried over from an earlier Archaic horizon. Flint tools were still common. Snub-nosed scrapers were very much in evidence, and though projectile points varied little in form and size they were never plentiful. They made use of the local stones, which furnished sandstone shaft polishers, hones, and smooth- ers of pumice. Hammerstones were many and were usually natural round boulders that were small enough to fit comfortably into the hand. Some were fashioned deliberately by pecking them into the desired shape; others were used as they were found. Some were employed until they shattered through use. The scattered fragments were found in midden piles and in midden pits. Grooved axes were present, but they were few in number. They were not replaced at the Hosterman site by the broad, flat chisel or gougelike implement called a celt. After these people abandoned the site, another group, presumably from farther north, moved in and built circular houses around four centrally placed poles surrounding a firepit. The floors were sunk to various depths. Attached were covered entranceways. Smoke escaped from centrally placed openings in the roofs. Subsistence apparently was equally divided between limited agri- culture, hunting, and gathering, with fishing playing only a minor role. Charred corn, beans, and squash seeds demonstrate their lim- ited capabilities. Seeds of wild fruits, charred and uncharred, indi- cate that gathering during the proper season was still of prime im- portance economically, while the greatest stress was placed on the plentiful supply of wild animals and the hunt. Byproducts of the hunt consisted of the hides; bones, and sinew out of which were fashioned not only the tools used to sew the tanned or prepared hides into articles of clothing but those for other major purposesas well. ‘The scapulae of the plentiful bison were fashioned into hoes which were used in the cultivation of crops in the rich bot- tom land below their village and for grubbing for roots and tubers 224 BUREAU OF AMERICAN ETHNOLOGY - [Bull. 189 that grew wild in the immediate vicinity. Other bones were made into awls, punches, knives, hide scrapers, and grainers. For some reason projectile points of stone were not plentiful. Those that were used were small, triangular forms. Earlier a side-notched variety was introduced. They all ranged from small to medium in size. ; The crude grooved stone mauls, similar in form to those found at the Payne and Swan Creek sites, apparently were used in the prep- aration of pemmican, for none were badly battered. Both hafted and unhafted flint knives were adequate for the skin- ning of animals or for cutting purposes. The manufacture of pottery was among the chief industries at the Hosterman site, as is demonstrated by the quantity of sherds recovered from the limited exploration. Due to the fragile nature of the pottery, its attrition rate must have been great. Pottery was first modeled and finished by the paddle-anvil method. The paddle was usually carved with simple shallow grooves that imprinted a simple decora- tion of ridges and sunken lands of the simple stamped variety. It was tempered with a fine to medium-sized grit of decomposed granite and fired to a tan, gray, or black color, depending upon the use of a cover- ing of bison fat on the exterior. The use of fat was demonstrated by the heavy encrustation of soot on the exteriors of vessels. Their chief form of decoration was by crudely incising geometric patterns over portions of the lip, rim, upper neck, and shoulder areas. _Cord-im- pressed and brushed designs occurred infrequently. There is considerable range in both size and shape from miniature vessels to those capable of holding 3 to 5 gallons or more. No two pots are exactly alike. This same holds true as to decoration, for it seems that little if any attempt was made to establish a definite style. Bases, as a rule, are rounded, while most mouths are wide. Ornamenta- tion is either incised or cord-impressed. Incised designs are neither precise nor carefully applied. Designs of both types appear on rims, necks, and shoulders. Incised designs are linear geometric elements consisting chiefly of series of parallel lines, chevrons, triangles, or herringbones. They ap- pear to have avoided the use of the curved line in their incised de- signs. When it does occur, which is a rare event, the sherd fragment does not give any indication as to the design element. Punctations, while limited, are used either at the top or bottom of the designs or as a unit. On the whole, incised designs were rather carelessly applied and never developed to any great extent. Contrariwise, the cord-impressed designs show . greater fineness, neatness, and precision in that they were carefully applied either as a band of horizontal impressions around the rim, as short diagonal lines Riy. Bas. Sur. owe Pap. No. 35] HOSTERMAN SITE—MILLER 225 bounded by encircling lines, or as interspersed “rainbow” elements be- tween series of horizontal parallel lines comparable to types illustrated by Will and Spinden (1906, pl. XL, g,2z,and 0). No two patterns are exactly alike. The size of the cords may vary—some are rather small and others are coarse; some have the horizontal elements closely spaced and others are spaced farther apart and not quite so carefully placed ; and still others show greater neatness in covering up the ends of the diagonal lines by the horizontal lines. The cultural pattern, as represented at the Hosterman site, does not seem reasonable. While the main portion of the settlement was surrounded by a palisade and ditch, which would indicate that it was built for defense against external marauding parties, other houses together with their trash areas occurred to the east and north of the palisaded area unsurrounded by any protective device. Why it should be necessary for one portion of the site to be protected by a palisade and ditch while another section of the same village did not require such features poses a problem. Whether the palisade was maintained throughout the short history of the site or only for a brief period, could not be determined. However, since there was such a feature the inhabitants must have felt the need for it; otherwise they would never have expended the time and effort to secure the necessary ma- terial and to erect such a structure. Furthermore, only a compara- tively few projectile points were recovered. These were of insufficient number to arm the inhabitants against any agressive exterior action, to say nothing about the number that the hunters would require to secure sufficient game to feed the hungry. There must be some logical explanation to the situation, but it is not apparent. It may be that other materials were used on projectiles, such as hardened wooden tips, which were not preserved in the site. Contrariwise, stone scrapers far exceeded in number the stone pro- jectile points. This appears to indicate that the need for scraping far exceeded that of a killing nature and that the materials were at hand upon which these tools were used. The scrapers have been used to fashion wooden containers and tools but such would seem highly unlikely. With the great mass of animal bones present, attesting to success in the hunt, hides must have been plentiful and they would have had to be prepared for use. Hence the large number of scrapers. Cultural refuse was nowhere thick on the site except in midden pits. Within the village it was sparse. Even house floors, while firmly packed, were not only barren of refuse but were not as firmly packed as they should have been if the house had been occupied over a long period of time. These two characteristics apparently would indicate a limited occupancy of the site, one of very short duration. No traces of any European trade goods or contact were found. 226 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 On the basis of the correlation of pottery traits, dwelling forms, and village plan with others in the immediate vicinity, it would appear that the Hosterman site, while having certain characteristics com- monly attributed to the pre-Arikara, is more closely associated with those of the pre-Mandan of North Dakota and possibly Minnesota. TRAIT LIST The following list of traits for the Hosterman site (39 PO7) has been prepared from the present excavation. An attempt has been made to arrange these in a functional order (Fairbanks, 1942, pp. 228-229). SUBSISTENCE ACTIVITY: Collecting Complex: Collecting of shellfish (Anodonta grandis (Say) ) Collecting of wild plant seeds (Chenopodium acuminatum) Collecting of wild fruits (Cleome cappariws) Collecting of wild fruits (Prunes virginiana) Collecting of wild fruits (Celtis sp.) Collecting of wild fruits (Rebes sp.) Collecting of wild fruits (Prunus americana) Agricultural Comlpez: Limited growing of maize (Zea mays) Limited growing of beans (Phaseolus vulgaris) Limited growing of squash (Cucurbita pepo) Limited growing of squash (Cucurbita mixta) Hunting and fishing Complex: Bison, deer, pronghorn, fox, rabbit, turkey, waterfowl, beaver, badger, coyote, red fox, gray fox, skunk, black bear, otter, prairie dog, ground squirrel, meadow vole, muskrat, pocket gopher and various kinds of fish. Useof hook andline. Domestic dog. Use of bow and arrow inferential from the presence of small isosceles tri- angular projectile points. COMMUNITY PLAN ACTIVITY: Village location and plan Complez: Village located in close proximity to stream Houses clustered without plan Palisaded village with accompanying ditch Village without benefit of palisade or ditch Midden pits numerous and outside of houses Cache pits numerous and outside of houses Cache pits, small, and inside of houses Hearth areas (fired areas) with broken stones Fire-cracked stones in debris Hearths outside of structures irregular in outline Architectural Complex: Circular house structures Rectangular house structure Bell-shaped midden and cache pits Midden and cache pits beneath floors of houses Centrally located hearths in structures Four central roof poles py. eS HOSTERMAN SITE—MILLER 227 TECHNOLOGICAL AND ARTISTIO ACTIVITY: Pottery Complez: Modeling or molding of clay vessels Use of paddle and anvil in manufacture of clay vessels Pulverized granite tempered Jar shapes only Rims vertical Rims S-shaped Plain Simple stamped Incised Brushed Curvilinear stamped Cord impressed Mat impressed Fabric impressed Painted Punctated Cornecob impressed Horizontal line punctates Random punctates Incising, narrow line Incising, broad line Herringbone Horizontal incised lines Horizontal bands of incised lines Horizontal cord impressions Curvilinear cord impressions Clay dawdles Rough-stone Complex: Hammerstones, natural hardheads Hammerstones, rough discoidals Hammerstones, pitted Chipped-stone Complex: Projectile points: Triangular with straight or convex edges, sides slightly longer than base, straight base as rule, some slightly concave or convex Isosceles triangular with straight or slightly concave bases Triangular with straight or convex edges, straight or convex base. Maximum width at base. Side-notched, straight-sided triangular, maximum width at base, straight bases, slightly concave or convex bases occur rarely. Notches are at right angles to long axis and are often narrow or shallow. Microblades. Scrapers: End scrapers: Keeled end scrapers Type 1 Type 2 Small round scrapers Large and small teardrop end scrapers 228 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TECHNOLOGICAL AND ARTISTIC ACTIVITY—Continued Chipped-stone Complex—Continued Scrapers—Continued End scrapers—Continued Small planoconvex end scrapers Large flake end scrapers Small flake end scrapers Flat end scrapers Thumbnail end scrapers Beveled retouched edge end scrapers Type 1 Type 2 Triangular end scrapers Slightly modified nodular end scrapers Trapezoidal end scrapers End scrapers with left carinate, type 3 End scrapers with right carinate, type 4 End scraper-graver combinations Modified end scrapers Side scrapers: Small flake side scrapers Medium flake side scrapers Large flake side scrapers Double edged side scrapers Triple edged side scrapers Prismatic flake side scrapers Lunate side scrapers Side and concave side scrapers Pointed side scrapers Modified side scrapers Ovoid bifaces Biface choppers Knives Burinlike implements Burin Perforators or gravers Lamellar fiake tools Ground and pecked stone tools Complez: Anvil stones Rubbing stones Paint stones Grooved mauls Catlinite Belemnite Pumice abraders Sandstone abraders Arrowshaft straighteners Grooved ax Handstones Utilized pebbles Hematite and other paint substances Riv. Bas. Sur. ’ ome Pap. No. 35] HOSTERMAN SITE—MILLER 999 TECHNOLOGICAL AND ARTISTIC ACTIVITY—Continued Bone artifacts: Cut animal bones Scapula hoes Scapula spines Seapula knives Scapula cleavers Scapula hide scrapers Scapula thong stretchers Scapula sickle Flakers Needlelike tool Shaft wrenches Knife handles Hide grainers Notched ribs Cut ribs Punches Worked pronghorn metapodial bones Fishhooks Worked animal scapula Problematical bone tool Bone tubes Awls: Rough splinter awls Type 1 (group 1) Type 2 (group 2) group 3 group 4 Cancellous bone balls Bone picks Split pronghorn metapodial bones Perforated rib fragments Punches Worked antler fragments Mineralized bone Antler hammers Worked teeth : Beaver incisor chisels Wapiti’s tooth pendant Human teeth Fetish or trophy skull Shell: Disk Unworked Metal: copper tubular bead er, n na ani nevi gers Bovelat tatonched edge apd 777 1 Types Srinagrnila> oud wore pate Giightly widifed meotiat vist sucked’ 4 a Ch a) sitet ae eS Gnd. ayvapera wit!) iat cartewk. tige R it sere pers with rirknt rartnabea, topes é Wee Kereper eave «grapes on Motiies eth REE tad lutboqntons wanton bono am Bi eC. 50 veal a sing eon fain boxxoW Bema (edie ite eit pect 7 Magtlivcs Titi bhke pb eliesag Hoda ydqor 1 dest = t Nota : Comal bediowall Se Beod 16 lodert taqqos toto ‘em » 4 hinNteti abtalonn ferwst Tpehignese Wim (ad Ax Tipe! ney ra PLATIIMEND, soe cibel vy Pirwrsttte cov? ater puieeh syhatiiy: APPENDIX 1 VEGETAL REMAINS A number of plant remains were recovered at the Hosterman site (39PO7). These were identified for the Smithsonian Institution by Hugh Carter, aided by Leonard Blake, John Bower, and Winton Meyer, of the Missouri Botanical Gardens, St. Louis. They state: All of the carbonized cultivated plant remains are of varieties similar to those grown in historic times by the Mandans and Omaha. The corn of this site has few rows, usually 8 but occasionally 10 or 12, and belonged to the race called Northern Flint (Brown and Anderson, 1947). The cupules are not as wide, the kernels thicker and the shanks more slender than in the eastern forms of this race. This difference may have been the result of mixture with corn from the south and west or with the race of corn which was grown at this site in earlier times. In the middle Mississippi Valley the Northern Flint race of corn came later than a race with more rows of grains, smaller grains, and smaller but deeper cupules. Corn similar to this earlier race is found in pre-1000 A.D. Among the cobs from 39PO7 it was possible to distinguish quite readily a few which were very much like the eastern form of Northern Flint. The fact that these were quite distinct makes it likely that they were grown as a special variety. The Mandans grew several kinds of corn and kept their varieties quite distinct. The corn is almost identical to that described by Nickerson and Hou (1954). More collections of corn from dated sites over a wide area must be studied before a reasonable explanation for the distribution of the various kinds of corn can be prepared. The median measurements of the corn follows: Median | Median | Eastern Phillips Median for for all |}extreme}| Dodd | Ranch for all | extreme |Northern of site site 39PO7 | variety | Flints |Northern] 39ST30 | 39ST14 Flints Row number 22222 25-2 8 8 8 8 8 8 Kernal thickness______ 4.0 4. 6 Ay 4.0 4.0 3. 4 Shank diameter_______ DO: He cae ZO) | 2220). |). S02) 11. 4 Cupule width_________ 9. 0 L250 10. 0 12. 0 9. 5 8. 0 Cupule deptire 2.2 2 eee ao as . 75 Zon | eee Number of rows of Crane. ee eae A eS 8 10 12 14 LGpocept tL eee ae Corncobs measured____ 87 Dit Yah as (Se ee eat |e FOP eee aes Several kinds of Cucurbita pepo, the common pumpkin and squash, were grown. Most abundant, nearly 700 seeds, was a very small-seeded (9 X 5.5mm.) variety, probably like the one called “Mandan,” with small green-striped fruit which sug- 231 22 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 gests in many respects, including flavor and texture, some of the wild species of Cucurbita. Most of the other seeds were small (12 X 7 mm.) or medium (15 X 7 mm.) in size and probably came from fruits like the small sugar pumpkin grown in the Plains area. There were very few large (18 .—--.--- 2. 3 11 48 29 9 11 2 cS pees | Seer oe Mandible- 22925... 222s. 29 38 55 91 5 11 11 4 10} 3] 1 4 1S 1(1) (0 See Sp eee (fee | eee 7 7 RP | Ree ene Pee rem Pee eel We esl Fe S| am | | Vertebra, atlas_---------- 15 6 19 5 2. 3 i Us| mee, | apa 3 [222s Vertebra; axis: 2-=-242-— 8 8 21 5 5 7 |--2 SE Sues pS ar 1 Vertebra, cervical __ 31 14} 100 15 11 oo | ee ete ib. | ace 1 Vertebra, dorsal__-__ 51 31 183 64 26 16 3 1 WAS 2) 22 eee Vertebra, lumbar__------] 38 24 | 172 17 13 AN ea 9 2) Shae eee Vertebra, sacral-_--_--_---- 1 Lt 9 10 jy eee bial Petes, Si isabella aden (eS) (CO Vertepra, caudal__-_----_-- 2 18 8 144 | Feees eed 2k pec Ree eee RST. Sf ee 2 Sa Beapnlgess oa = sce 11 12 51 27 3 1 ees 9 AC) Ste ee 3 Humerus, proximal_--__-- 10 20 26 8 1 2 8 5 9) 2) 2h Humerus, distal____--_--- 17 14 63 58 | eee pee Os essa oe Be eee Radius, proximal_____---- 9 19 48 37 3 AO eee 4 7 ala J Ns | 2 Radius, distal__--__-.---- 3 4 40 Pa} earl ees ae Gry pees elt eh egy Sy oe ee (Osta eee se ee 25 19 39 20 7 ik i Sees 4 6 [2h 42 2 Moetacarpal.-...-.-.-22 y i a 58 Ce a Pe AOS PSIE TSA! ee: | eee Pelvis: 2 2.2222. ug Re 18 17 56 13 2 6 13 4/3) 3 1 Menrar, headae 225. <. 2 -2|-ncnasf ane 2 OM resen= ese Pepe) kya oak co gee ar | (RE |b = Femur, proximal_-_- 5 18 26 12 5 6 16 4 10))|222 som 1 Femur, distal--___- 9 6 26 | ee S| 2oce a Ea ence ae ee Tibia, proximal-_--_-_-_-- 5 21 37 14 v 5 12 6 jess5)" [Eee Tibia, distal... 2ssse- 425, 6 vd 22 ON reese ee 10)") 22-5 ake 22] 14 eee AStraraluss 85 oan 7 9 27 Ce) pear ease 4 1) pees 1 a A ee ey) Wee Sacrum. .--=..-.--<.. 9ihts.... 3 7 3 | ee i) eee Tl iy emer = Metatarsal 2.2. 502 2. 122 67 45 34 21 12) |=== = 12 Db [22s gosleoce by the dorsal vertebrae of the antelope. This is not too highly in- dicative, since the animal was small enough to be easily transported from the point of killing to the village without too much discomfort on the part of the hunter. On the other hand, there. are 91 whole mandibles of bison represented in the collection. Of this number 46 are right and 45 are left, indicating, at least, that there are more than 45 individual bison represented in the lot. Fragmentary mandi- bles were not saved due to the large mass of bone encountered. This is true also of the rest of the bone material. Only recognizable whole bone and some with identifiable articulative surfaces were salvaged for identification purposes. An examination of table 7 discloses that certain elements are con- sistent (horn cores, maxillae, hyoids, scapulae, humerus distals, radius proximals, radius distals, ulnae, and tibia distals), which may indicate the maximum number of individuals represented in the bison group- ing. Other elements (occiputs, frontals, atlas vertebrae, axis verte- brae, and sacra) are conspicious by their scarcity. This can be Riv. Bas. Sur. nave 5 Pap. No. 35] HOSTERMAN SITE—MILLER 235 partially explained in that animals that were killed at a distance from the village were dressed in the field and only the most desired portions were transported back to the village. In dressing, the head, along with the atlas and axis, was severed from the vertebral column, probably with a blow of the ax. As has been pointed out numerous times, the head, as a whole, is a heavy unwieldly part of the animal and is covered at the most with a minimum amount of usable meat; hence it was not usually transported from the kill to the village. On the other hand, if the brains and tongue were desired they could just as easily be removed at the time of the kill. White (1954, p. 164) seemed to think that the mandibles at the Dodd site are close to the greatest number of individuals represented and that none were ever used for anything and were probably brought into the village along with the tongue of the animal. He states: “Certainly the easiest way to remove the tongue would be to smash the ascending ramus of the jaw and remove jaws and tongue as a unit for further cutting at a more convenient time.” If the per- centage were low, then he suggests that most hunting was performed at considerable distance from the village and the tongue eaten at the hunting camp and the mandibles discarded there. Practically all of the mandibles in the Hosterman site were complete with the ascending rami entire. This fact would indicate that they were part of the skull when introduced into the village and that we did not find the entire assemblage. On the other hand, we recovered 24 hyoid bones, which is a small bone attached to the tongue. This number approximates the number of maxillae, a fact that may be significant. Vertebrae——Very few animals are represented by the vertebrae, for their distribution covers the entire column. However, the in- teresting thing is the large number of caudal vertebrae, which signify that the entire tail was left intact with the pelt as it was removed from the animal and brought back to the village to be processed. This fact indicates that the tail was an important ornamental feature, since it was left attached to the hide after skinning. Fish—The remains of several fish were found in various of the midden pits. Fishbones were mixed with the midden in Features 1,2,and 8. In Feature 9 they were found at various levels: between 3.0 feet and 3.5 feet, between 4.0 feet and 4.5 feet, and between 5.5 feet and 8.0 feet. Many were just miscellaneous bones, but the complete articulated skeletons would appear to indicate that the complete fish was discarded because it was not considered to be palatable by the Indian. Fish remains were also found in Features 10, 18, 21, 22, 24, and 25. Individual fishbones appeared in many of the small isolated midden heaps. 661-932—6417 236 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Dr. William Taylor, Division of Fishes, U.S. National Museum, rapidly examined the fish remains and has identified not only the families to which they belong but also certain skeletal elements. He states : The fish material from site 389PO7 consists of remains of two families of fishes, Cyprinidae and Ictaluridae. The Cyprinidae remains are all pharyngeal arches, probably of Hybopsis (Platygobio) gracilis (Richardson). This cyprinid in- habits large streams and rivers in the Plains. The Ictaluridae or catfish remains consist chiefly of spines, pectoral girdles, and vertebrae, with relatively few of the heavy skull elements present, suggesting that the skull parts were often disposed of before reaching the midden. The bullheads, which probably live in the region, are not represented. All material seems referable to the channel catfish, Ictalurus punctatus (Rafinesque) and the blue catfish J. furcatus (Le- Sueur). Since these two species are superficially very similar, I do not wish to rest heavily on definite identifications of either species for any of the bones. Where identifications are made to species, they should be regarded as probable and not positive, for that species. A brief list follows of elements from site 39PO7. Catalog No. 2481____ Cleithra (3) of Ictalurus punctatus ‘and 2 vertebrae of catfish. 1248____ Pectoral spines (2) and dorsal spine (1) plus parts, 2 cleithra of I. punctatus ; also miscellaneous vertebrae and skull elements of catfish. 1117____ Spine and parts of cleithrum of J. punctatus. 1535__._._ Vertebrae, opercle, cleithrum, and spine of J. punctatus. 1403___._ Miscellaneous bones, vertebrae, ete., of at least 2 individuals of J. punctatus. 1427__.__ Vertebrae, spine, cleithrum, etc., of J. punctatus. 8465____ Spine of I. punctatus and 2 catfish vertebrae. 1668____ Several bones of catfish, chiefly J. punctatus. 252___._ Spine of catfish, probably I. punctatus. 234__.__ 2 spines of Ictalurus, probably J. punctatus. 3323__.. Vertebrae and vertebral complex of catfish. 1302___._ Vertebrae, opercle, vertebral complex, etc., of catfish; cleithrum of I. punctatus. 3983__._. Spines of I. punctatus (2). 421____ Spine and cleithrum of I. punctatus. 1182____ Miscellaneous catfish bones, including spine and cleithrum of I. punctatus. 1344____ Miscellaneous bone of catfish. 1469____ Spines and cleithra of I. punctatus. 1864____ Miscellaneous bones and spines of J. punctatus. 1962____ Lower jaw of catfish. 2357... Miscellaneous catfish bones. 1609____ Many vertebrae and miscellaneous bones of catfish. 4186___._ Spine and cleithra of I. punctatus; other catfish bones. 976___._ Pectoral spine of I. punctatus. 1173____ 4 catfish vertebrae. 1497____ Catfish vertebrae and bones. 2477____ 2 catfish bones. 1483___._ Miscellaneous catfish bones, some identifiable as I. punctatus. Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER 237 1270____ Catfish bones including supraoccipital of I. punctatus. 1103____ Spine of I. punctatus ; other catfish bones. 1608____ Spines of I. punctatus and 1 vertebra. 4074___._ Many catfish bones, including parts of 5 skulls and ribs, vertebrae, cleithra, spines, etc., of I. punctatus. 1518____ Cleithra of I. punctatus. 1249____ Pharyngeal arch of cyprinid fish, probably Hybopsis (Platygobio). 3463__.__ Catfish bones and pharyngeal of cyprinid, probably Hybopsis (Platygobio). 1428____ 2 pharyngeals of cyprinid, probably genus Hybopsis (Platygobio). 1917____ Catfish bones. 1343____ Part of lower jaw of catfish. 3979__._._ Cyprinid pharyngeal with 1 to 4 teeth. 1862____ Fish rib and part of 2 catfish spines. 2480____ Cyprinid pharyngeal arch. 1271____ Cyprinid pharyngeal arch. 4102____ Catfish spine, J. punctatus. 1363____ Supraoccipital of catfish, J. punctatus. 3536___. Hyoid apparatus, catfish. 4185____ Pharyngeal of cyprinid, probably Hybopsis (Platygobio). 3749____ Pharyngeal of cyprinid, probably Hybopsis (Platygobio). 684____ Pharyngeal of cyprinid, probably Hybopsis (Platygobio) both arches. 2476____ Pharyngeal of cyprinid, probably Hybopsis (Platygobio) one arch. 420__.__ Pharyngeal of cyprinid, probably Hybopsis (Platygobio). 350____ 2 spines of I. punctatus. 1433__._._ Bones of rather large catfish, genus Ictalurus, part of supraoccipital possibly I. furcatus. 1585____ Miscellaneous bones, spines, ete., of catfish, including some identi- fiable as I. punctatus ; some upper skull elements, possibly of I. furcatus. 1282____ Catfish bones, mostly skull, some axial; spines recognizable as JI. punctatus. 1515___._ Miscellaneous bones, some skull of catfish including I. punctatus. 2457__-_ 2 pharyngeal arches of cyprinid: Hybopsis (Platygobio). 1918____ Bones of catfish, mostly identifiable as I. punctatus. Birds.—Outside of the golden eagle sternum, identified by Dr. Herbert Friedmann, Department of Zoology, U.S. National Museum, none of the other bird bones have been identified. Bird bones were not scarce in the village fill and today there is still no scarcity of birds in the region. Today there are many large waterfowl that frequent the area during certain seasons of the year, and small birds number over 150 species according to Over and Thomas (1921). APPENDIX 3 INSECT REMAINS The chitinous remains of several insects were recovered while Fea- ture 3, a cache pit, was being cleaned out. These have been identified by members of the U.S. National Museum staff and the U.S. Depart- ment of Agriculture as belonging to Carabidae Pasimachus sp. and Calliphoridae sp. In most instances, remains of this sort are not saved, but since these were found in undisturbed deposits within the cache pit, the lowest stratum—between 5.0 feet and 5.5 feet—and the highest—between 1.5 feet and 2.0 feet—it was deemed wise to get determinations in case anyone should be interested in this phase of life at the Hosterman site. 238 APPENDIX 4 TABLES OF MEASUREMENTS Projectile Points Catalog | Maximum | Maximum | Maximum No. length width thickness Mm. Mm. Mm. 194 21 14 2 331 28 15 3 332 21 14 3 359 20 13 3 360 21 17 3 361 15 11 3 362 20 13 3 363 23 16 3 364 18 14 3 365 17 13 3 366 20 ll 3 367 18 11 2 433 15 12 2 446 27 15 3 447 17 12 3 457 18 13 3.5 458 24 16 4 459 28 16 5 520 29 18 3 621 21 16 3 522 21 19 2.5 524 28 17 4 575 22 16 3 617 29 15 3 618 18 14 3 619 17 13 4 625 43 29 7 654 17 15 3 704 26 17 6 705 27 8 5 706 32 20 6 743 24 14 3 744 41 23 4 745 48 16 4 746 20 9 3 747 24 15 3 753 34 19 3 783 32 14 4 784 26 19 3 785 23 13 3 786 21 17 3 788 20 16 3 888, 23 19 4.5 895 29 15 3 915 24 18 4 , |) ete Rea 15 4 988 25 16 3 989 21 14 2. 5 1038 24 14 4 1058 24 14 3 1078 18 9 3 1084 17 14 2 1088 38 23 5 1095 19 15 3 1163 23 14 3 1164 25 15 2 1195 30 20 6 —_——$F | | Shape Material Tsoscelesithiangular ho. one Quartzite. Feed side notches triangular, concave | Chalcedony. ase. Tsosvelestrisnemlar 22 os = oe oe Quartzite. Smal side notches triangular, concave Do. ase. Medium side notches triangular, con- | Chalcedony. cave base. Tsoscelestrianemlar= -) ——- == - oe Quartzite. tape Oa Se es oe ns ee ee See | ha epd On: Seon! Ge nen naan Po ne san aa eee ot) Quan Zi bes led tha CO sere one oe ee ree ee ee halcenon ye haga QO neon eon sac eset ee ene ee |) QUanbzitee EE et (CG ater ae ails Aeolian IY Me 5a Do. sabe CO eee es pee ee ee eet DASE Isosceles triangular, with slight side | Quartzite. notches. Tsosceles triangular square side notches, | Chalcedony. square base. Fsosvelestiniangilat=-¢------- os eee Quartzite. Pe (CQ)Ow) nce Pie Se See eae Seah aeaaiey Boe ee Do. git (oc pet PR oti etiely Gee FO) Bp aed Do. Isosceles triangular, concave base___-___ Do. ee LTE ETE D UC ORD om eee ee See On aed Chaleedony. appa oe eS aie eouaiia © a a Lac #0 (oe? EA a ee inte eres S> <-aeeY quaient. Tapetes triangular side notches, con- Do. cave base. Tsoscelesimaneilar].) --.-.----.-4-2--- 5 Do. Pe (6 (0) Ele Seale eee Fe reeie ueoneeey She pats Do. Isosceles triangular, side notched Do. Tsosceles triangular____---__----_- Do. imate ios eens Chalcedony. ps re do= 2 Pan senna =|) QUAREZILG. ip rea (6 Kaya Pete eed SS + ee ie ee Do. a Oe ree he nn eee eee belay se Gs 2.65. b eo eee || Chaleanony. Isosceles triangular, side notched __-._-- Quartzite. Isosceles triangular, side notched, con- | Chalcedony. cave base. Wise & (1h) ED Gee ES + eeigeyereees We hee Do, Isosceles triangular, slightly concave base_ Do. Tsasceleshirianeulan._O... 22 oooh sees ce Do. Presumably isosceles, concave base. -.--- Obsidian. Isosceles triangular, side notched_-__---- Quartzite. dponceles triangular, slightly concave Do. ase Tsosceles triangular.) ss_.......Joce.-. Do. ae ee es ee ee eee Se Do. etal Sy COE Lag coe oa ono a anos epee | Chalcedony. Isosceles triangular, concave base_------ pe ts Tsosceles triangular. 2 ==. == sen5- jee triangular, side notched____-_-_- Chalesdony. eae Oe SEM ee ee Fe Isosceles triangular, slightly concave_-.- Quartets. Isosceles triangular, side notched__-_--.- Do. Tacaee ips triangular, slightly triangular Do. Isosceles triangular, slightly side notched Do. concave base. Tsosceles} triangular. =2-s-n=--=s2=--<=- Do. fete fo (OAS Nel aarpeleicsetelicieled burst ese nteatcena ogied ha tates oe wala e COG SESE ORD, Pistol PAR, Oc Do. Saar OOF ween sone cee ee one seneccectensetel QUArEeies see Oe ee pt cocceceecesees! Chaleadony: 239 YAO BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued Projectile Points—Continued Catalog | Maximum] Maximum | Maximum Shape Material No. length width thickness Teongeles triangular, deep concave base__| Quartzite. SA by) ee es Oe Se See Chalcedony. eae triangular, side notched___-_-_ Do. Tsosceles triangular, concave base_----__ Do. ee triangular, slightly concave base-_ pie ood Do. Isosceles triangular Chalcedony. Isosceles triangular, slightly concave base_ Do. Tsosceles triangular, side notched_-______ Quartzite. ers triangular, slighty concave base_ Do. CEE: (op aR eh EE Ecce PER a Chalcedony. isiepetes trian oiler: 2 ae oes mo Bees Se Quartzite. Betas (6.1 )ER ys pe EE Vs SOE Be a 8 Do. aver OS: £26 E¥ a Fe As ee Ks Do. Bide notched. 5. sab eee oe eee Chert? ei aie triangular, concave base_____.- Quartzite. Se eet ee [Bo (G(r! OTe ee oe 2 See Chalcedony. igual sided triangular sides, slightly Do. concave. Isosceles trianguler, slightly concave | Quartzite. e. Sidemiotcheds =. .224 2-25 - Ses Sees - Chalcedony. Biponles Urian gilar: 2 on ae See pe Bc (OL Ds eae Ra Fo eee ee 0. Side motenad et Seo ee eee Do. Isosceles his ewan 22 eas 2 oe eee Do. pee se XO (Dod: Eee eer “4 IE RE a Quartzite. Side naidhed pee ae ee 2 Do. Tsosceles triangular, side notched_--_---- Chalcedony. legsveles triangular, slightly concave | Jasper. ase. oe ee triangular. see 2 ee Qua sersefoul (6a) ee EE a ne Spee Ooi Wen 0. ispeesles triangular, slightly concave base_| Chalcedony. pee (lip ie Ee See Be See eres Se (OTE hey pein (6 (oa ae a Eee OSH 2 Sees Fo Fe Do. ee (cee a een Bn ee ee: SiGe Do. neat Chie OO ee Se eee Do. ierateles triangular, slightly concave | Chalcedony. ase lsecelys triangular, side notched_-__---- Quastaie- isse-yere(0 (Op mt eee enn Ok. AE SRE Se ES 0. See (6 Liaise ook Bho Se een See Se et Convex-sided triangular, concave base_ Isosceles triangular, side notched_---_--- Onaleedony. Steninicg = 2s 20s ee ee Isoscelesjtriangmlar® 2} 85 .- 2-2 Quextsite. Isosceles triangular, slightly concave base- Do. Isoscelesitrianevlar=_e2-) = Chalcedony. Isosceles triangular, slightly concave base_| Quartzite. was OO n2 toca pecones= ak aoe seae een ee Onaicedonys Convex sided triangular, concave base__| Quartzite. Tsosceles triangular, slightly concave base. oe Convex sided triangular, straight base__ Tsosceles triangular, side notched--_---- Quake. Tsosceles triangular, slightly concave base_| Chaleedony. Hsosceles}trianpular. 4. 22-322. 2. seo Jasper. eee 5 Coe a ae ae ees See Sees Oleh Arie) ee triangular, slightly side notched_ pe Re QUES = eS in te ee a eee 0. FEA a ee EN Ge eB, Do. isboutes trianemiar. ce selene Chaleedony, pop Ss Orns ce eed yee ee eee | Onenurites Pour OS eee oe ee eC halcenonye Isosceles'triangular, slightly concave base_| Quartzite. NSOSCeles inianeiiian. - ee) ee Chalcedony. Tsosceles triangular, slightly sidenotched_| Quartzite. fsoscelesitriang lan. 8-222. 513 ee oo Do. P42 (Sl0) oo 32 OE AES Pie abr Rag? Sea Do. Isosceles triangular, side notched-_-_-__-- Do. Baas REE SU SRONORE TS SF 9Z 3 Do. Fragmentary side notched_______.-.-__- Do. Tsosceles triangular, slightly sidenotched_ Do. Isosceles triangular, unfinished _--.-_-__- Chalcedony. Tsosceles triangular, side notched_- -| Quartzite. Fragmentary side notched_..----------- Do. Very Crude Point? = sane eee Do. Convex sided triangular concave base... Do. Riv. Bas. Sur. Lae Pap. No. 35] HOSTERMAN SITE—MILLER 241 TABLES OF MEASUREMENTS—Continued Projectile Points—Continued Catalog | Maximum | Maximum | Maximum Shape Material No. length width thickness Mm. Mm. Mm. 2277 31 26 3 sidemotched 2 -<.- s-beho- 2 a Quartzite. RE iA |, Sees ap ete | eee ae et ne a he a ee al Jsoscelesitrianeular: ge. ee Jasper. 2326 25 16 3 Isosceles triangular, side notched_______ Obsidian. 2327 19 14 3 Isosceles triangular, slightly side notched_| Chalcedony. 2328 29 14 3 Tsogcelesitrianeulars 32252 Quartzite. 2330 25 13 3 Isosceles triangular, slightly concave base_ Chaleadgny. 2335 22 17 Diet (accu O(N eo Sane eee, he Se eS Se Do. 2346 35 14 4 Tsoscelesitriangular: 222-2 2225 sae ee Quartzite. 2388 24 14 i ee eee (ay ee See ea eo) eee Do. 2390 22 16 + ee (eee Ree See So eee es Bee eee Do. 2391 24 Lig, ee pee oe Ee a Re Bw OE Ty eee Do. 2394 19 13 4 Tsosceles triangular, slightly side notched_ Do. 2464 19 12 3 Hsoscelesitriangular..* J-___.. Be Do. 2469 25 14 3 Isosceles triangular, slightly concave base. Do, 2470 21 10 Ca eee MOL. twee Sao 8 oo ae ca En Re Do. 2484 20 11 3 ae (6 (er Eee aD ees eS ee Do. 2537 16 15 3 Tniangolarss #22 22 Sec ee Do. 2539 35 21 5 Isoscelesitriangular. .. >. 2.22 .- 24 es | Do. 2680 18 18 Da ae CCE Res te 2B eee Do. 2681 21 15 3 Isosceles triangular, slightly concave base_ Do. 2720 21 14 TE | es (0). ah ee ee eee ere eS ee Do. 2741 17 13 4 Convex sided triangular, convex base_-_ Do. 2742 22 12 3 woe COE. Pees esa ey eR Do. 2765 21 12 3 @oncaye)bises. 2-282 Se eh a Do. 2766 13 12 2 (inianemlanes soos «Soe ss) ot eepee ee Do. 2767 23 14 4 Isosceles triangular____......_______-__- Do. 2768 17 15 3 Triangular, side notched fragmentary__ Do. ROOM eee See en 2 eho Ne ee a Fragmentary, sidémotcued 22-2 aes Do. 2770 20 16 3 Tsosceles triangular, slightly concave base. Do. 2831 15 12 2 prianeniiar. 2225 5b ee ee Chalcedony. 2864 15 13 eee eet CYs) See ES es Dn ey beeen Quartzite. 2865 25 17 5 Tsosceles triangular...) o< =... 3-22. Do. 2866 27 12 3 Crude side notched Se ee Chert ? 2867 18 13 3 Isosceles triangular Quartzite. 2882 | sso e ee oe ete eee ee ae ragmentary side notched___ Chalcedony. 2888 20 16 4 Sidemotehed= =o. 252i =~ _- Quartzite. 2889 32 22 5 Tsosceles triangular Do. 2901 15 14 3 Sidemotchedis.s. 2) es ae Do. 2903 22 16 3 Tenserles triangular Do. 2904 25 16 cn | eS Es (0 RE 2 Oe ee See Do. 2907 19 14 3 Side reiied Ae eer ee eee ree 2 ree Chalcedony. 2963 21 13 3 ee Oe Se oe oe ce ee Quartzite. 2964 24 20 4 Convex sided triangular__.__._________- Do. 2994 31 21 6 Tsosceles: triangular. 22-=-s222-22-2e--= Chalcedony. 3058 27 12 4 Isosceles triangular, slightly concave base_| Quartzite. 3059 27 14 Os | See Ot eee eee er i eee Do. BOGON brs cease ee a/R oan ee 4 Fragmentary side notched______________ Do. 3090 16 12 3 Tsnsceles ian culars. ios 2 Sa Chalcedony. 3107 16 12 Si D 22 dO see os ee eee ee 3130 24 14 4 isosceles triangular, side notched__-____ 3153 14 13 Ze )| 2226200 eee aan ae ee Chalcedony. 3208 20 12 Sen (eee o Le eee A eRe Ee RY 3306 16 13 3 Side notehed straight base, abrupt tip__ Quneees 3308 22 17 3 Isosceles fenguler, slightly’ concave base_| Chalcedony. 3334 25 15 3 Sidenotchedi ft - 22 Se he eee ieee. Quartzite. 3335 28 14 4 Tsoseelesttriangulars= 222 aaa ee Do. 3378 22 15 Lit eee G2 ee ee at EE eee cr Do. 3389 15 13 2 eee Concession e Do. 3397 15 12 Bh WE Saee Gos sa 8 eS Do. 3430 22 15 Be) eee ae See See ee eee es SS Do. 3431 17 11 2 Tsgeveles triangular, slightly concave Do. ase 347 16 11 3 Do. 3472 20 10 3 Do. 3473 16 13 3 Convex sided triangular, concave base_- Do. 3489 25 16 4 Comermnotched = as eek Chalcedony. 3490 15 14 3 Tsoscelesitriancular. 2-12) 2 eas Quartzite. 3491 21 13 3 Isosceles triangular, slightly concave base_| Chalcedony. 3500 18 13 4 Stemmed) slightly... 62-5 > ee Quartzite. 3515 18 14 4 Isoscelesitriangmlar 9220 eee Chalcedony. 3521 21 15 6) 8 an eee (0 (0 yaee Wie Pee Re RE See Ee is) ena Do. 3522 28 14 3 Isosceles triangular, side notched_-_-_-___- Do. 3539 17 14 4 Teoeebles triangular, slightly concave | Quartzite. ase. 3540 24 15 Ah ae Co SiS... | on ee AR er ST 3: Sos Ee Do. 3541 23 13 eh. Pa COS Bis: ae Sr ee Chalcedony. 3546 21 14 3 SA ese, (6 (Galea 2 St Re eRe ORR TT SN Tokens Quartzite. 3601 32 19 (aia ee 0G SEES REDS RE aos SONY 0. 3615 58 22 6 Isosceles triangular, side notched | Chalcedony. slightly concave base. 242 Catalog | Maximum | Maximum |} Maximum No. length BUREAU OF AMERICAN ETHNOLOGY TABLES OF MEASUREMENTS—Continued width thickness [Bull. 189 Projectile Points—Continued Shape Material Tsosceles triangular: 2222s 2-2. ee Quartzite. Isosceles triangular, slightly coneave Do. CO CO ee ase. Tsosceles triangular, medium concave qv haleegeny- ase aaees MOM. 3.28 neon tee Quanatiee Isosceles triangular, slightly concave Do. base. Bete Le RAR Se eC ey 2s CE Do. as Goo sab ws nce ee Do. media ly dos se 6) 2 ik Sob Boe ae eS Do. sous (0 Ca eee ese) Sees a. see Chalcedony. Side notched, short and stubby-_-__-_---- Quartzite. a oa triangular, slightly concave | Chalcedony. ase. aD. Ore iO 5 esc et ea ccc oc ews] "Quanizitoes Jes doo 52) 2.22329 e ee Se | Chaleedony: Isosceles triangular, chipped one side___} Quartzite. Side notched, chipped one side only, | Chalcedony. slightly concave base. Isosceles triangular, concave base- - -- -- Quartzite. oes oy’. $. 85-2 - c2a Seen nn ca -ooe 4|) Chaleadony. ee Gos o 2b oo sie cccl ast wscaressaeese ss | QUantatios Isosceles triangular, convex sided, con- | Chalcedony. cave base. Tsosceles |triangular-. ...->-.2=--..-.£2:-- Do. do Quartzite. Chalcedony. uartzite. isosceles triangular, medium concave halcedony. ase Sidémiatched:. 222 t ee Quartzite. i Do. Sat Ge ee ee ee ee ae ae, Do. Isosceles triangular, side notched-_-__-_-_- Do. Side notched, tip missing_____....._____ Chalcedony. Side notched, one ear missing__-.-.._._- Do. ae. Bas. Sar. HOSTERMAN SITE—MILLER 243 TABLES OF MEASUREMENTS—Continued Microblades: Group 1 i eR I ee Catalog No. Length Width ‘Thickness Mm. Mim. Mm. AQQQR Serer ren tara Je LL 31 13 13 UO Lt: ee epee el ee Coe 32 18 4 ai! eS = Se eee tar 34 13 3.5 PAE a ea lari, SO Beh de 35 12 5 aN O (CS RE ED Bie AS ne ah 35 14 2 i eae | SES, SB, See ee ave pee 35 16 ») TUNG SH ee ee ce fae ee 36 16 3 OOS) ET eR hen a ie Oe a 37 13 2 (UST ee Oe ee dw 37 15 4 ZA foie. Re OIE: es LAPSE we ee Fo 38 22 4 Pepe ee ens Whee i Aen meee Fe Sy 38 17 4 LG Se eee Lae che eo 39 14 8 Ib) ee es Se aa ee 40 16 4 [AOS seas ca SO le 40 18 4 NOOSE ee eee Ee 40 20 4 US (|, fe ES Ik a eee ee ae A 2 45 15 5 vAFeA MD) ais Ne a Sak a ae 47 16 6 74015 { 5 ee aaa Sa See Sy arene eh 47 13 25 VOPR Le eae ee ae oe 49 16 5.5 Oe ee ee Berar dR 57 23 4 1 With attached graver tip. Longest, 57 mm.; shortest, 31 mm. Microblades: Group 2 Catalog No. Length Width Thickness Mm. Mm. Mm. GOO! = ih aes see eeee se pees 18 12 2 DARDS 22 see eee een ee 19 12 3 1Sbeterr ss sees eesesssse 20 19 2.5 (SH, (sk es RS ele gee ae 20 15 2 Dae ee sya = pln SF SD A ie 20 16 3 AN eal cee Ow ee 22 is} 4 DAS Gtr coer towet eer eee se 22 21 4 RO DOR cee oe a Si E = ee ke 23 16 6 pigee-= 22 453i ce eac i fe eeeseks 23 16 5 3) a ho ee ae eee 24 12 3 ET a on es a 25 17 3 9, a ee 25 17 5 LL ee eg 25 17 5 TG (De se wie pl ae ee 25 14 3 4) ye Pe ee 25 15 3 DUO GH as Hans on Nee ee ae we 25 14 4 RNa A not = AR ee ees ae AS 26 17 3 AAS sit nl sc fe Rh Ad he ee ne bs 36 U3 4 si 015 (een ean Set Ce 26 16 5 DUB Nes 2s 3s ene aes eos Ste 26 13 3 Saas wr aise we Na IE os Ah i al 36 17 5 LiGS Mors 2. 2 ao Lobe es 2 ew 8 yar 14 2 DOS 2ENs == Ses aes owe. eid es 27 15 5 BOOST AMA Serie ol ay ee 27 15 3 LOAD LS ao = & sete sie & dete ve 28 14 3 BL DAM > <= 2 on siemens 5 Belt es 2 8 29 17 3..5 Sepa LG ep aye 29 19 2 Sy | Cee © Lehane 30 1g 3 244 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued Pointed Side Scrapers Catalog Degree of Maximum Maximum Maximum No. angle length width thickness Mm Mm. Mm “29h re | One 62 31 31 1 Oe 8. Sabres 8 50 30 20 4 Ae tee eS 2 ee ot 65 28 23 5 TLS ak St ap Saree 81 35 22 9 Gira ees eee 2 32 39 28 iy | GARB os bes 3 26 35 25 6 Sate oe Se. | 2 See 26 26 21 3 2 | Sa ek Ye ea 43 50 27 13 Gaeees 8 Ses 32 34 25 8 OU S a 7 ee eS 45 26 20 4 VDOpe 3 eae Ss 13 25 23 6 1G 2 < Mieee” 2 8 74 42 41 9 LE Th Re BP 28 44 29 9 (ies es eee niet | SN eae 69 tf 43 6 PGg0s4" 2 3 eeae tus 59 43 40 5 I ed | a 18 32 26 9 Pie. ee A 30 26 21 7 5! id See aS ee 87 31 22 7 P2511 ie hee | Serena 31 42 BE} 5 2 ee 30 22 24 4 Pee ee Te 84 33 30 @ Ss Se eet LE 71 17 31 8 Ske! pe ces 78 39 34 5 51 AS 9 aR 58 34 27 5 GLU: oe oie FY 67 25 22 5 Due ee eB aS 69 19 32 10 So ae nae — ¢ Rac 57 22 20 6 Keeled End Scrapers, type 1 Field catalog No. Maximum Maximum Maximum length width thickness LS! | SER aaa eg cea 21 17 9 PALS i SRN Ss te ee cee 24 18 9 6 Ld AA AE ie See ee 28 22 9 2 US RA, 2 eee a 25 21 7 gee. Se eats as Se 27 22 11 AU eres oc Sees 2 Se 32 22 11 TE SEIN, (her SA eRe ee 34 26 11 | V7 eis naan in ses 5. Sees A Op 37 22 vi 5 eS, ee Nate oat peeps 32 24 10 iS): SN Aen eae Seapine Poyr dea 30 23 11 GON eos ok Eg ee ad 22 20 8 i pee aera a Aas Ep 1 Sea re ar 30 21 6 Roe cats sh NR eee Be ale 27 19 8 Ba ee BE eg S Ree 26 8 ‘i Ls Ve aA, Sp Repann Re lees 25 18 7 LR RRS, “re Sees oes 30 19 9 74 SYs aa Oe RD 2 a eRe Sie 28 22 a RENO eee res a os WM Ng a at St SS 31 22 11 Dee OLE gets ee ae 24 22 9 ro | VS SENG Raps eee eee met 25 22 10 Riv. Bas. Sur. Pap. No. 351 HOSTERMAN SITE—MILLER 245 TABLES OF MEASUREMENTS—Continued Keeled End Scrapers, type 1—Continued Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. Lee 2 eee E> Saeee ee epee 21 18 9 ee ee es 23 20 6 cy Se ee ee ee ft * Ae eee 31 17 1§ [aes A | eo eeeeieaae 14 17 6 LES SA ae «Ree | UP “eee peeps ee 22 11 i]. 1 ieee RE, Oe. = alia eee reas 40 25 112 Rea nS VS Re en BM nee yan Oa 34 23 18 LSS (1 eS caaaiae SS: & Saas allel nies 34 20 9 UL? ees ae 2 1 see eee 25 23 9 “fot eaten ieee se | aeaareinens elbelee 39 16 8 1S, | pape Se 35 21 9 NA Gees lore 8 24 20 17 LT aS, 2 eee 26 21 10 1) EU he ES es, 2. See eee 24 23 9 Me5 (Corian Fee Neraouce,s 23 21 6 aes SS: : See 26 22 9 TCE ee ee <.) See een 24 21 5 POG4.. Sst RABE O99. ok. 3 28 23 8 122. Ge eee, 2 3 < Baier 25 19 8 inl) ae Soe Sl 1 See 28 22 7 Bios le 2 50 eS a ee 39 21 8 Panne Woe cee IMR a eS 24 We 8 CLE ee Et { ener 24 19 8 ie Gea EL ee eee 31 21 8 gs (ES Se | Se ee 27 18 7 Ly) ee > |S eee eae 39 27 114 “2 aes See 5 ee een 33 17 8 Eee reek RE ee 25 19 7 Goa aey ee es ES SS Os ee 2 25 12 Dele Bey, eae: 26 26 9 1S Ge BARE 7.” SR 28 17 17 SEO a say OB a eS 24 18 if 2,10 ope Reames © JOE pip Cenenenry aoe 3d 21 11 AN Sane = 2 ey ee eS eh 25 20 9 1H eee eae "ee ens 16 20 7 SO seg 22 ee Ce ee 34 15 11 a Ge S| 2a es 32 19 110 1 Slightly planoconvex. 246 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued Keeled End Scrapers, type 2 Field catalog No. Maximum Maximum Maximum length width thickness =, | ee, ae aeaey OS 7 eae ae ee 35 24 10 Bee eR LE 28 18 8 (OL UE Sg ee Net ge ae 48 24 110 5 aE ES >t Gana © eee 32 20 11 ee ee fe. ee eee 28 21 7 BG) (Da) Bee oe ae 27 ibe 17 O22 ee eS 33 24 10 LL aS | eee 29 21 8 a LS Se eee SH 18 9 "sa |i A I) | 7 TERE pee era 35 24 18 "EL Sea ES OF ¢ SE See 30 22 8 Ee): Se eee, ST ae a ee 28 26 6 Le SS NTS 1 5 SS “eed 31 23 11 2.) aes 2S. ceeemen Gemmrees | Sr, 4s Aa eee 25 142 Lat | Se? OS De: Reems 21 18 10 Ui lt Se See ee 33 19 9 LSS See eee ae ON ees Meee 23 16 16 Li eae Se eae 36 24 141 0) EE OF Ce ee 27 24 9 D0 Se 27 19 9 lo, be ol 7 ES See 28 23 11 1 Soe ES a eee 40 27 10 Loi) aS 2 Ses ee 24 21 8 LS See Se hee Sees 23 18 “4 Ae 1 ES & 2 a: See 29 21 12 Uh 1S CSS ©. ees See 26 21 11 “Le RET 3s) SERED AS CARRERE | St 57 Ee OS eeee 18 8 Se Se eee 30 20 111 -= |G EV eeamaneraeensiesel i { Saaieea Raa 30 22 7 Se a. \ Re OR 30 25 9 JE ED S| Se 44 27 16 ce UE ei A Rate 2 Shs ¢ aed 30 24 9 ee 2 Ee | aes ee 24 17 6 AE ROO aR’ DRE Sea 47 26 12 ASS CT |e | eee eee 25 21 6 JE IE UT - alee ence en ei 29 20 10 LA 3k ee "te | RRS eee 26 18 a. 7 (| es 2 | See mee 41 23 8 Lt) Soe. | Earners 29 25 12 hl ae | eeeeeeene 29 16 7 = Ses 1). eee ee oes 32 19 16 =f ees Cee es, 29 19 8 2. eee Oe ° Poe ae 29 22 7 2 es SS eee ee 38 23 9 3 Sees 0: ) See eee 29 19 11 255) ee of Ee eee 22 19 8 £30) 05 eae ce +e ee 26 17 16 ig ee ont eee ee 24 20 8 3) ees i. See ee eee 23 13 5 2 ee os See eee 24 20 v 3h) ) See 2) eee | 2 eee ae 26 11 25 es 2) eee ee 34 20 12 32; ee a: Se eae eae 31 22 12 | ee 2 |) ee 33 at 9 FL ee Re ee eee oe 30 19 10 AA tire ene ee RAIL AS Se a ee 22 7 BO Gs eh er ee 28 23 9 1 Slightly planoconvex. ey Bee or HOSTERMAN SITE—MILLER 247 TABLES OF MEASUREMENTS—Continued Carinated End Scrapers, type 1 (cutting edge advancing to the right) Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. Aer re oe Ae sl Re 23 20 17 fe) el entice ma) Ma ls ca 29 19 6 SOAs ee eee NR 25 19 7 PAT AUG cde RES Ce Ee ane ea 23 19 9 SLUT thee ah Sanpete eae aes 25 21 7 Ry Om ee ee wr les ee 28 20 6 OO ae See & osc ot 29 18 16 OU Ame es ate EE ee ee 25 as 16 AS I este een oS a 2. 1, Hn ee 25 15 8 1 Slightly planoconvex. Carinated End Scrapers, type 2 (cutting edge advancing to the left) Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. Sa es SY ae NM Dee oe i a 25 Le 8 <5 = RN SAG TS hf aa 25 18 7 Eek, SARIS, Oh SEA ae eee 24 18 6 i 2 5 LE RN Ben. ee ies Pee ae 24 17 14 (40) Et Sie ei Ry = EES ee a Os 21 IW 6 a1) 5 7 a a, SL». oll re 25 20 8 1 Slightly planoconvex; rest are trianguloid 248 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued End Scrapers with Left Carinate—type 3 Field catalog No. Maximum Maximum Maximum length width thickness Mm Mm. Mm 5] 5 payee nape lpm oop ap) = te Rectal! 25 22 8 2 Pugin seamed iets 9 1S lesiadiecle Syste ae! 29 16 6 ee ee oe eee a oes eS 23 18 5 Cis as es ha mes ie: See et gaia 31 22 7 5 len pl pty, 5 aiecenea emcee: (RE 2 AE ie eae ea 28 10 7 | nate consi a eal 6 | selec allt 25 19 7 eta eer Le ee Se 43 30 12 OL eeepateoapespllen A, ‘a: «5 aulpelppetedechantinees | fi. : ape oe DAN 6 TEU U eh pti aeaieenpliele Sa? 7 ‘gute peak ete iene 25 19 7 TI PATI be aretha lected each ethe Saks Ae aller sty tira IG AE Ee ee pe lee 22 5 TAG lg et ec rn TER NE AE 26 22 10 E70 el aly ea ge hs. i ee 26 oe 10 1 el Sct ex dh dale tae Sle hye lhe Moen lee eae 32 20 9 ISI) elaewh eNee hee ile y ale pga Deu cecedgl 26 20 5 0587 aa te lonpehge tn panei 24 23 8 1 peepee rpms bay egempnien npemeapeana 28 25 6 Rpg Stee 2 Sd eh ee 22 20 8 CUE ec eaertpar pe nel ipl dapat 20-0 ati 23 18 10 UIC scp pga a, Seliiyh Alerter A eee 27 22 5 erties atria meatier aaa 25 19 7 LTS oy SS elena fas Apel eet 24 19 6 eg eee le ee ee eS 24 23 5 os Ac yan ally [ne soo Aesets alec aie ease 39 20 11 PAs Be elenclem yee on epee ae a 31 21 5 HL OI oe haayagten laa 2 te Deedee, ele tieh 29 22 14 LAB nse yee mental g ty slop dat pelt aS 28 20 5 ea gern eaten certs wee egetee cen tk, 31 21 4 PADS a rete eeelas wy! Leteak sa ete ene aee 21 19 5 SLD) Sun kek SNE ES Lee eee 21 18 a Eee e een ry ene mre Te Ape ae 25 23 6 512144 | cael eet lecae ed oe 97 Beak de 27 22 7 yA) Ee hee aed eb tlle yt le iad 26 22 6 The le aah geld 5 Cty Sl Sed secs 24 21 7 STN ag odin a lh San oan ed 25 16 6 Shoo Nagy later moped pee ec eee 29 on 7 “LEER AAalpdeed lige yee a Share ae eee 39 26 if EN Ie ata Spal ygeh Npeac | oR ae CE 30 pal 1 Planoconvex ventral surface. Riy. Bas. Sur. Pap. No. 35], HOSTERMAN SITE—MILLER 249 TABLES OF MEASUREMENTS—Continued End Scrapers with Right Carinate, type 4 Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. PND e eae Ve eee Meee eee ae F PAS 21 17 ce ae Sa Rae ena BE Te 34 21 16 DOI eee Oe ae 24 19 5 ADO 2 ee ee SO ee 24 18 6 AQ ot needed i 5 0) alleles pate BA 23 20 6 i7(0)) R95 tI at. ea eras BS 24 22 6 O52ha nea Seo TS Oe ed 25 19 6 FS Batra sine ne ns weternee Meh cept dah oe 22 22 4 17 Jc cee geagoane alacant cae narrate pene 32 22 v | CS Se OR ee ki Br cae Da 24 19 16 2) sees ee Sie Ne ET aes 26 21 5 TSUN Ss ES el 7 aS ee 28 eZ 5 TSS OWSES yet Reed Nas ok 23 18 il L5O0 Omri aN 2) arcttep ie 22 22 14 1GZABaNSs ser Od. ap rap 26 19 5 17 OE LS eal eeicoenonemay pra 29 19 16 NG 2S es Seem ey ce CaN 35 28 11 DUSOS teens Se se ee LE. * 26 20 5 es © | eae 30 19 4 7+) (ae aE NS. “eigenen See 28 Zs 17 aA (\ Reais 2 ean Mee Pes 22 23 17 Fd OLDS eet ae. eh.» Sena aa 27 20 6 co OOo} are Sea ay Se! Et «| ele Sata 333 19 19 SUA Aas ee sede 28 | Eien Se 22 17 16 OA Oh a hee eS CERES 25 19 6 0) Qiere a os Ratt eee eee ors Te 30 26 9 BY Geet e Sabet 23 20 7 Do. LS apeeerteaee amar EK Sai 30 26 7 | Left. 1 UL i pei Se "Sea 30 26 10 Do. r/o, aS 32 25 9 Do. > ¢ ae, |, a 25 23 7 | Both. Small Rounded Scrapers Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. eee ree ee ee BD oe ae = Se 26 20 10 5 rah eld tagcat ae fe enya ee 28 26 9 <1 Sa gatpele Selene ood, + lela atelier 22 22 9 Buen ee ee fess ens 22 22 6 hog eh ee ELE) Sp eee 28 25 9 enews 2 or ian Se res ot Se 25 20 8 OLA a Ape tale 6 § Ips te ae e 24 19 7 Ln a eo ae de 22 22 8 >| Eee _ 5 Soledad 22 20 8 eee ewe cena 28 e ms oe 2 24 24 9 eed ese > sie RDS vp Vs = = 43 39 19 ive] aR Reese Od «| ee ca eh ee 25 20 8 eae 258 foe SE Se Heke oe 16 16 9 * 211 Enea altpataas Meal Se laa 21 19 7 a ES Reade pe ke ap poe deere op pete 23 20 8 Fhe RIE ERRNO IE 19 17 4 eens Sid dn nar eile td MAA m8) 19 18 8 SAR Se hic a's pew alin 3 ahah won ree 19 19 6 Gala Aes = 92 ceeds re BOEING 23 21 6 PN eere sap ini vel enmrice atta rent a- —yee vores 20 20 10 daa eI oc ae 28 23 LE Large Teardrop End Scrapers Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. IR Searels SST sagen Oh. Saree eee 40 22 rs Ree uicc.= ce at or 51 25 9 1 | Ege aro aes Uh RRs 46 30 9 1 ae eas ee ene ea aoe 38 26 un Pi) Ae Se EE) Ee ee ee 37 21 19 LUT IER a) oe perme) Wh) SERINE baleen 40 23 6 1 Positive bulb of percussion on ventral surface eradicated by chipping. Riv. Bas. Sur. Pap. No. 35] HOSTERMAN SITE—MILLER TABLES OF MEASUREMENTS—Continued Small Teardrop End Scrapers Field catalog No. Maximum Maximum length width Mm. Mm. CTE a eee aaa are UE | Senne aera 31 19 1ST ERISA I TP Se 25 16 UCR Eee I SIE eae 23 19 LR RSE! TU, 0’ Seg ea i 30 19 UD Fee hee a ee ee ee 33 18 2 (ES ST BE 31 18 Large Flake End Scrapers Field catalog No. Maximum Maximum length width Mm. Mm. iN elle atlas ees Yel arpa anne 34 15 ae teal tala nip So ee alee eae 28 24 eee ne Se ER eke 37 20 TEAC} aol hare arenes bat fue ieee ane 35 ube MOOS ete ess ok ae eee 32 22 115 a a ea, Y i i 32 18 EDL a ad RRR Cas” tel Get ee 45 14 Mibhtes fe 8 ee ee ee 38 25 LE, alan eae! "SA ha ie as 33 17 SALI 5 eli. OS” RNR pe 32 19 lle Sng eects 8° 5 Pe id i ial ee 30 28 LG al ae RS Gir Yel betes cari 27 22 16 1 Sa ees ec an aes feat Ta ee 29 CUAGMS Ee ra mek oh eee eS 39 29 UT, lee moa) Na Taser ae eid 31 16 2 [PAT epepeeatabaee eae. O° Aci ata Sone ma 56 25 BOB eee co Behe oe cae 36 20 StU SS oie es! ao | ees A ee 30 23 PR eae ee 33 24 RS eee Ae eae Se ae Se i 29 29 Small Planoconvexr End Scrapers Field catalog No. Maximum Maximum length width Mm. Mm 1 1 pe apeiapan rar 2 oS at ope le 32 20 115 oY panes Ss BOE) ie aera 23 16 : 2 eo ties Rae UR 7 SS ae Ee 26 18 LUE 5 en: Seon el ee. rs oe Maer 29 20 OE 3 2a oak aoe eee 29 20 BRS ao in ee eee 27 22 BtG0=.20 2555 252L aoe 29 18 1 Has a graver tip attached to side of scraper edge. 661-932—64——_18 251 Maximum thickness Mm. 7 5 8 8 6 8 Maximum thickness Mm. 2 8 9 vi 6 8 7 6 4 o 6 9 9 6 6 ih 4 8 9 9 Maximum thickness Mm. 9 18 6 7 10 9 7 252 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued Small Flake End Scrapers Field catalog No. Maximum Maximum Maximum length width thickness Mm. Mm. Mm. 1 Pe ee tecteeateceeas tes 5} Ae ieee OE 27 18 6 Ao oer ee ee We eee hs oo 21 7 3 a pees cliente et |S ie a eR 25 20 7 EL OC ee ne a ee A se 28 17 6 Soe sete eee 28-2 ee 24 21 5 2 aes ieee anaes | 5 eee eae 20 21 ui Gis re ea fish he ete aps (ok an Si eRe 22 20 5 SOU) ne it a eb te Laas ARLES 2 7 23 9 UALS en A Sed Ss AS fe iene A ey he poem Shey Se 26 23 8 DOW Soe ia a no, ee 29 16 10 500 22ers en ese eee ee 25 19 7p COZ See oe ee ee een ee 26 14 6 GSO ald Se ele al PM eM een hm 26 17 6 9 (aS Skee ae ee eS 26 19 a 100 8 acer eth eee 21 15 6 20) yee eeere ne tn en Saree 24 / 6 2G saat eee ee Bes Oe 2 24 7, 10 119411 0 eS a oe Be ae eee ere 28 PAN 6 11 Te Je a ed. CL 6 args SP ee 23 20 17 2900S ee See re eR Se 23 17 9 5 (eter eas = eee ee ee eS ae 17 20 6 9 Se a a RO oe ey 18 20 6 I 2 Oe ye oe ee 26 13 uf NG GOES Seen ae eee = eee 27 16 7 OOQR ES setae eee no ea 25 18 ci iO ee oe a ee ee ee 22 16 5 EO SO ees ee cease ae ee 27 UUF/ 6 LOY, lta diag as ew olin, 08,” 5 See ee es 25 15 5 Dal Oban set i Oe ae oe 24 16 5 F124 VS) sane cepa bh Hae a amen ee Ree 27 15 8 AU) eee ye UN eho 27 16 5 DAD (i see ee ee ee 25 19 25 715) 0) | Rene ere BS Se Pea es 23 17 fl TATE, SCR ORNS Ph, 4 Iams Ee 29 18 v4 FAL PG) Se Ro pes PORN SEE... pal ae Sag 24 21 6 FO c} 0) ree apr ated age Sy - i 22 21 7 i Os hereiceah a tale eter (as, 1 eee Ee 22 20 8 2 bre) 0 eid ae el aa a arp rae ae eae 15 19 6 AAD ee ere Pete ay ed ie Ren 8! 29 19 6 7210): ee aia eee ee 29 21 9 OO LO ea ae ete ae 25 18 8 Paya he | eau pee rahe aon red aes 28 19 9 ANNA Arete Stem. 2, FONE og ge 22 16 4 71 Ld een mage nn A (ED a reap ee besa 26 ibe 6 1 Possible graver tip. 2 Burinlike tool at base. Riv. Bas. Sur. Pap. No. 351 HOSTERMAN SITE—MILLER 253 TABLES OF MEASUREMENTS—Continued Thin Flat End Scrapers Field catalog No. Maximum Maximum Maximum length width thickness eee eee Sane Ne ee ane 18 26 5 De ee OE ne ee 18 18 4 S102) zs ask a ese ieee Ue «ka ae ae pedis 38 23 9 BOQ Bee on ae ee ce ea 28 22 8 GOO Be erasers ot ees fg Se 24 19 5 PSII: st spt erential thors Slaea em a toal en 22 19 5 peat So ee ra Te 26 18 5 UPTO 5 late a PAs Sy oe apa P43) 23 5 We Stree fee pet se gee ee 18 19 4 LIPO fel seep epee a «eel apenas |S, sae ene 20 5 NGGIpeeee ete e eR ee aS 26 20 8 NGOZ Metre eee men ye elk tl 27 20 4 UZ ISS Ye 2s ps tere Des ee, Sine bs weet eae 23 16 6 LN ae eae lelan. sue Sr eee er are 17 17 6 1 fo Tinie settle Eanieoal eh sein ee eaneaeti 19 18 5 OG iia ee SE AE 23 22 4 DAD Aas ait eye T Gd 24 23 7 ea eee ee ee op 27 21 7 AWA De NRC eee SSS EE ape eee 20 22 5 PADRE SA ON ek 2 ea eat 5 21 19 8 _AT CaS Solis tee 2, Sa ee 25 25 7 “3 a een 8S Ee ee 17 18 6 DOC9RME Danek to bemath! ) 23 23 6 Ue ements ae eer a 24 23 6 DUG AES eS Ve Sa ee eee eee nee Mee eB 19 3 DRS re Re nS oo Se ae 28 24 tf gS ola ee ean reem v-L- SR) | es tA ee. 20 i BO Oa ete ee Be ee 15 18 4 251) ST See IenEn eT ” AE OR 34 30 5 SOO ene a. ae ee ee 24 21 9 AQQGUMI AN, J) 5 Ee 20 20 5 ASN DNE ARE RNE 2 oe EA 24 16 5 719A | 14 On elk SS Pree, oY Ale 28 18 if 254 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued Thumbnail End Scrapers Field catalog No. Maximum Maximum Maximum length width thickness BE teeeeee oto 1h Rie Lee eee ee 21 ily 15 3 en 2S > Seeger ee 17 17 6 ht BA SA an ars Ba, 21 16 14 (2 71 ae pe re Rae St Ee ae ee 19 15 5 EE 5 > oa a ee 18 17 5 i SSS AE ER ae eee e's SUE oe Eee | 18 14 +L | a eee oe Seems: 18 14 6 [CT eee Een eet ae. 17 17 5 [ico nee eens Doc,” ) SORA eneene ee aS 14 12 5 SEs bra See bh Se a ae 16 19 5 SN Was reas a Ars oc nn ye 22 18 17 oj ig ES TE LEE eee 15 16 5 AAS PY i RRS se A Ws! NB oe Seem oe a 5 19 18 if _ ae oN SEO ee eee, 19 14 3 Loc 3 eee S| See: 14 14 6 ES aes 2: epee rr: 21 15 14 Ng he Sepa Ee Se Shenae ik Lz 5 ree) os f Seen 16 17 6 2. es 8 * Sener eee eee 18 14 3 2 eee Se ©. |: Sere. 23 13 16 eS er ee See ees | 19 14 5 NI nts ie SO alread 20 15 1§ 1 Borderline specimens, Probably should have been placed in type 1 keeled end scrapers. Triangular End Scrapers Field catalog No. Maximum Maximum Maximum length width thickness Ls 2 IE ope aa NO Sp eet ea ete 22 18 4 se ee eee 24 22 6 2 RR ae ee sc ede epee 25 18 + ate, ie ig Ea eee Sy Le 23 22 14 eee ee = ee 27 20 6 17/5 SRR es a re ee a 23 25 18 LL ES ET el one Share 29 20 7 2 2 SSR ee ORR Deen ere Pere Tse. 23 24 5 cree Cre Selenite a ou 27 21 7 1 Planoconvex. Riy. Bas. Sur. Pap. No. 85] HOSTERMAN SITE—MILLER 255 TABLES OF MEASUREMENTS—Continued Trapezoidal End Scrapers Field catalog No. Maximum Maximum Maximum length width thickness Bae se eae xe See ea 28 22 i to) 0S as ee ee eee 26 22 16 > ig eae eee Sr) eeay 29 23 6 HE CN Naas feenpialea ill Bo alien apenas 26 18 5 Sete oe IL 23 19 7 NG 4 ie Apa hk perme 23 22 7 21) Er ey ee 8 ee ee eee 19 23 Lig Re cow Ee ae eS, 22 17 4 7ANNG) otek oe ME cde Lo ied: / sh ale bead 30 22 vf PO DVR ORO Hep 24 18 6 S21 the Ae ey ne eel ee of 26 21 6 gy i amen eke es ees Rape = 23 22 7 Pas ck ta a eg A le 24 20 6 See ee hc Senne eee ere. LTA 24 19 5 SOORRE 4 Getinio’ vation) 27 20 6 1 Slightly planoconvex. Lamellar Flake Tools Field catalog Length Breadth Thick- Tool type No. ness 51 | ee eee 6 45 15 5 Possibly hafted knife. 1 SOS eae 37 14 oa Scraper. 278400 et 38 22 4 Do. C1). et 31 13 3 Knife-scraper-graver. oA soos aa 34 13 aL Knife. AIBGO WEA. ase ote 37 14 3 Scraper. O45 Ss 552 oe 28 14 3 Do. $650. 24a 36 16 3 Do. 2056... 22.29_i0 42 13 3 Do. BOGS. 02. eae 35 14 2 Do. =| eee Oo 40 16 4 Do. te a 56 22 + Do. 22 ees tee 48 16 6 Do. 1Oo_ JS... 400 40 18 4 Do. 7} ee So 49 16 5.5 Do. Os = ae 40 19 4 Do. 256 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TABLES OF MEASUREMENTS—Continued Large Gravers Catalog No. Length Breadth Thickness 205) eee ah 48 21 15 ait eo ae Ses 43 24 12 C0 Het heel tel WS, Subecgnn Sao 42 19 6 EG pe ecg SOR 9h. "Spee Se geste 31 19 6 eee Dns 2 eee ee 36 13 8 1S ie eee uel 0. 1, Sabie secre ee; 26 L7. 7 “OE acai eae, iar eee 29 13 6 Daria fe is hie Ee eee 26 16 5 5 EES oS Siete ee 31 24 5 a ee eee ae te 27 16 7 Bena eS 2 eee Oe ee ee 34 18 4 [is sc chygreghn natal cpa a tar en | 33 SOM Solas Yea OT 2: 2 pps 2 acts omg As Puls ala coer ed 37/ 24 Ns ee ee oe ee rot Her se 39 16¢|. 22.2 [bea pepametale aka bysge pin esd eaWnrer gee 39 30 22-2. ee 1 ea a ut pa Re ye a PR en 43 24-| oo ec eee (icipee spat amet vant (ake "Nets kapapeeee, Swag! 39 03) SES = == Small Gravers Catalog No. Length Breadth Thickness EOC hee a ON es eR ke 24 21 7 PAPA | ee IS EES: TA 24 hg 4 31 ee ee: Cee ee eee 19 18 4 BUG te tos ior aOR oe foe d 18 15 4 SONOS a ee SOR eel 21 18 tf Potee sss teo ress asock oad 24 17 5 PN i het or st AOE oo Ae 22 20 3 S30 Oe eee eee See 20 17 3 le coe el ON 2 te ot Bd 25 1y 3 BUS cote sees ewes d 25 18 17 742 ee ee eee eee 31 18 3 PAECEET (SILT SED OG CI LEAN I ES 22 19 6 TLS OA aT aes Se 7s 27 22 8 1 Scraper-graver. Lengths of tips range from 1mm. to 4mm. Ten Ne aor HOSTERMAN SITE—MILLER 257 LITERATURE CITED Brown, W. L., and ANDERSON, EDGAR. 1947. The northern flint corns. Ann. Missouri Bot. Garden, vol. 34, pp. 1-28. BuRKITT, MILES CRAWFORD. 1956. The old Stone Age: A study of Paleolithic times. 3d ed. New York. Cooper, PAUL L. 1949. Recent investigations in Fort Randall and Oahe Reservoirs, South Dakota. Amer. Antiq., vol. 14, No. 4, pp. 8300-310. FAIRBANKS, CHARLES H. 1942. The taxcnomic position of Stalling’s Island, Georgia. Amer. Antigq., vol. 7, No. 3, pp. 223-231. FLINT, RICHARD FOSTER. 1955. Pleistocene geology of Hastern South Dakota. U.S. Dept. Interior, Geol. Surv. Prof. Pap. No. 262. HEwEs, GorDON W. 1949. The 1947 summer field session in archeology, University of North Dakota. Proc. Fifth Plains Conf. for Archeol., Notebook 1, Lab. Anthrop., Univ. Nebraska, pp. 21-24. Hort, WESLEY R., Jr. 1952. Report of the investigation of the Scalp Creek site, 39GR1, and the Hillis Creek site, 39GR2, South Dakota, 1941, 1951. South Dakota Archaeol. Comm. Archaeol. Studies, Cire. No. 4. 1957 a. The Rosa archeological project. Museum News, W. H. Over Mu- seum, vol. 19, No 6. 1957 b. Report of the investigation of the Swan Creek site, 39WW7, Wal- worth County, South Dakota, 1954-1956. South Dakota Archaeol. Comm., Archaeol. Studies, Cire. No. 7. LEHMER, DONALD J. 1951. Pottery types from the Dodd site, Oahe Reservoir, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 2, pp. 3-15. 1954. Archeological investigations in the Oahe Dam area, 1950-1951. Bur. Amer. Ethnol. Bull. 158, Riv. Bas. Surv. Pap. No. 7. MaAcNEISH, RicHARD S. 1954. The Pointed Mountain site near Fort Liard, Northwest Territories, Canada. Amer. Antiq., vol. 19, No. 3, pp. 234-253. 1958. An introduction to the archeology of southeast Manitoba. Nat. Mus. Canada Bull. 157, Anthrop. Ser. No. 44. Department of Northern Affairs and National Resources, Ottawa. METCALF, GEORGE. 1956. Additional data from the Dodd and Phillips Ranch sites, South Dakota. Amer. Antiq., vol. 21, No. 3, pp. 305-309. NICKERSON, N. H., and Hou Dina. 1954. Maize from the Dodd and Phillips Ranch sites. Bur. Amer. Ethnol., Bull, 158, Riv. Bas. Surv. Pap. No. 7, Appen. 6, pp. 180-181. Over, W. H., and THomas, C. 8. 1921. Birds of South Dakota. South Dakota Geol. and Nat. Hist. Surv. Bull. No. 9. RITCHIE, WILLIAM A. 1944, The pre-Iroquoian occupation of New York State. Mem. Rochester Museum of Arts and Science. Roserts, FRANK H. H., JR. 1935. A Folsom Complex. Smithsonian Misc. Coll., vol. 94, No. 4. 258 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 SHANTzZ, HoMER L., and Zon, RAPHAEL. 1924. The natural vegetation of the United States. Atlas of American Agri- culture, U.S. Dept. Agric., Washington. SMITH, CARLYLE §. 1951. Pottery types from the Talking Crow site, Fort Randall Reservoir, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 3, pp. 32-41. SPAULDING, ALBERT C. 1956. The Arzberger site, Hughes County, South Dakota. Univ. Michigan, Mus. Anthrop., Occ. Contr. No. 16. STEWART, GEORGE R. 1946. Man: anautobiography. New York. StTrRonG, WILLIAM D. 1940. From history to prehistory in the northern Great Plains. Smithsonian Misc. Coll., vol. 100, pp. 353-394. WEDEL, WALDO R. 1936. An introduction to Pawnee archeology. Bur. Amer. Ethnol. Bull. 112. WHEELER, RICHARD P. 1954. Check list of Middle Missouri pottery wares, types, and subtypes. Plains Anthrop. No. 2, pp. 3-21. WHITE, THEODORE. 19538. Observations on the butchering technique of some aboriginal people, No. 2. Amer. Antiq., vol. 19, No. 2, pp. 160-164. 1954. Observations on the butchering techniques of some aboriginal people. Amer. Antiq., vol. 19, No. 3, pp. 254-264. WILL, GEORGE F., and Hecker, THAD C. 1944. The Upper Missouri River Valley aboriginal culture in North Dakota. North Dakota Hist. Quart., vol. 11, Nos. 1 and 2. WILL, GEORGE F., and SPINDEN, HERBERT J. 1906. The Mandans; a study of their culture, archeology and language. Pap. Peabody Mus. Amer. Archeol, and Ethnol., Harvard Univ., vol. 3, No. 4. WILMETH, ROSCOE. 1958. Report of the investigation of the Payne site, 389WW302, Walworth County, South Dakota, 1956. Archeol. Stud., Cir. No. 8. APPENDIX 5 HUMAN SKELETAL MATERIAL FROM THE VICINITY OF THE HOSTERMAN SITE (39P07), OAHE RESER- VOIR, SOUTH DAKOTA By Wuu1am M. Bass Department of Sociology and Anthropology, University of Kansas INTRODUCTION The preceding article by Carl F. Miller describes the archeological investigation of the Hosterman site (39PO7) in Potter County, S. Dak. No human skeletal material was recovered during his excava- tions. In September of 1962, while boating on the Oahe Reservoir, Richard Weeks of Pierre, S. Dak., discovered the partial remains of six human burials eroding out of the east bank of the Missouri River and sent the bones to me for identification. Mr. Weeks is an amateur col- lector who, in the past few years, has checked the changing banks caused by the rising waters of the Reservoir. I accompanied Mr. Weeks on two such boat trips in the summer of 1962 and at that time encouraged him to collect carefully all human skeletal material for study. Charles L. Keeler, a student in anthropology, assisted in the preparation of the material herein described. The skeletal material was found just north of Whitlocks Crossing on the east side of the Reservoir at an elevation of 1,565-1,570 feet. The location of the find corresponds to that of the Hosterman site, and the burials therefore may have been associated with that earth-lodge village. In his letter of transmittal, Mr. Weeks states that the skeletons were found in three different concentrations. The first was found in 1 foot of water; the ribs, vertebrae, and mandible were in the original burial position in the mud with the rest of the bones scattered near- by in the water. A few more scattered bones were found on the bank approximately 30 feet south. A complete burial was found in the bank about 10 feet from the first group. The skeleton “was on its back, head toward the west, legs flexed to the right, arms down to the sides with hands on pelvis” (Weeks, 1962, personal communi- cation). In no case were there any associated artifacts, covering, or anything else which might suggest cultural affiliation. 259 260 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 DESCRIPTION OF BURIALS The bones were submitted in groups corresponding to the three areas of concentration in which they were found. We have retained these area designations and have assigned individual numbers to the remains of the six individuals present. Many of the bones were broken and > scattered due to wave action and slumping of the loess. Whenever possible the bones were reconstructed to obtain anthropometric measurements. Group 1 Group one (those found in the water) consists of portions of two skeletons. These have been numbered Individuals 1 and 2. Individual 1.—Male, 35-45 years old; the skeleton material of this individual was fairly complete though fragmentary, but the skull was broken beyond repair. Age at death is based on changes in the pubic symphysis after McKern and Stewart (1957, p. 85). Although not a good criterion for aging, cranial suture closure substantiates the assessed age according to Krogman (1962, pp. 76-91). There is com- plete endocranial closure but ectocranially all of the sutures are visible except for a small area of the sagittal suture between the apex and obe- lion. The determination of sex is based on pelvis, cranial, and long bone morphology. Stature estimations are based on measurements of the left humerus using Trotter and Gleser’s (1958, p. 120) formula for Mongoloids. The three parts of the sternum (manubrium, body, and xiphoid process) have fused completely. Slight arthritic lipping is present on all of the vertabrae with fusion of the second and third cervicals. Stature range: Low 167.02 cm. (5’ 56%’’) Mean 171.18 em. (5’ 7’’) High 175. 34 em. (5’ 9’’) Individual 2.—Child, approximately 6 years old, represented only by a right humerus. Age, based on a maximum diaphyseal length of 162 mm., is estimated from mean measurements given by Johnston (1962, p. 251), whose measurements of sub-adult Indian Knoll skele- tons go only through 5.5 years with a mean humeral length of 154.67 mm. for the age group 4.5 to 5.5. Group 2 The analysis revealed the partial remains of at least three indivi- duals based upon the occurrence of three right femora. Indiwidual 3.—Male, 35-50 years old; represented by a right femur and possibly a right and left humerus, a fragmentary right scapula and a mandible. The femoral head has a maximum diameter of 46 mm. Riv. Bas. Sur. oe Pap. No. 35] HOSTERMAN SITE—MILLER 961 which places it in the male range for this measurement (Krogman 1962, p. 144). The mandible, two humeri, and scapula were assigned to this individual on the basis of coloring of the bones and general mor- phological characteristics. The mandible has a square chin indica- tive of the male sex. The humeri are rugged and neither contains a septal aperture. Age is based on wear of the teeth. The enamel of the masticating surfaces of the eight remaining teeth are worn off completely and correspond to Hrdlitka’s (1952, p. 53) third stage of wear. Age for this stage is suggested as 35-50. Only eight teeth (four incisors, right canine, left first premolar and the right and left first molars) are present. The remaining six teeth were lost after death. The third molars have not erupted. The missing right first premolar was abscessed and may have been lost just prior to death since little or no resorption had taken place. There is a cavity in the first left molar below the gum line and all of the teeth present contain tar- ter. Stature range: Low 162.51 em. (5’ 4’’) Mean 166. 31 em. (5’ 514’’) High 170.11 cm. (5’ 7’’) Individual 4— Adult female (?) represented by a fragmentary and weathered right femur shaft only. Sex, which is questionable, is based on the smallness and gracility of the shaft. Individual 5—Child, approximately 7 years old, represented by a right femur shaft only. Age is based on maximum diaphyseal length (222 mm.) which, when compared with the generalized postnatal growth curve of the femur given by Stewart (1954, pp. 407-450) suggests an age of around 7 years. Group 3 Group 8 consists of the complete flexed burial of an adult interred with the head oriented toward the west. When first discovered only a small portion of the skull was exposed in the face of the bank. No artifacts or covering of the burial were found. Mr. Weeks states, however, that “the soil was very discolored, a yellowish green” (per- sonal communication) and suggests that this may be due to a decayed hide covering. Individual 6.—(fig. 52). Male, 30-40 years old. Age of this indi- vidual is based on changes in the pubic symphysis and corresponds to a mean age of 35.84 according to McKern and Stewart’s (1957, p. 85) classification. Wear on the teeth compares with stage 3 in Hrdlicka’s (1952, p. 53) system. Cranial sutures are beginning to close endocrani- ally and are visible ectocranially. Arthritic lipping is present in the fourth and fifth lumbar vertebrae only. Sex is based on [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 262 eUlION, ‘py “eyoyeq yyNog *sI[eJUOIZ VUIOU ‘9 fsI[BdI}IOA eUIIOU ‘¢ ‘sI[e19}e] ‘Aqunod 193109 ‘4Od6¢ ‘QUIS UPUJOYSOPFT OY} WOIZ (9 “ON [ENPLAIpUy) IeU I[Npe jo ssuIMvIP JeJsOIUvID—"7¢ TUNSIT Riy. Bas. Sur. a3 Pap. No. 35] HOSTERMAN SITE—MILLER 263 pelvic, cranial, and long bone morphology. All are consistent in ex- hibiting characteristics considered to be those of the male sex. Stature was calculated using the formula for Mongoloids on measurements for the femur plus the tibia given by Trotter and Gleser (1958, p. 120). Stature range: Low 168.51 cm. (5’ 6 6/8’’) Mean 171. 74 cm. (5’ 7 5/8’’) High 174.98 cm. (5’ 8 7/8’’) ANALYSIS Table 1 gives the cranial measurements and indices for Individuals 1 and 6 and for the mandible of Individual 3. All are males and represent the only measurable cranial material. Stature calculations could be made on three individuals, all males. It is interesting to note that the mean estimated stature does not vary more than 5.43 em. (214 inches) among the three. Individual 1 has a calculated mean stature of 171.18 cm. (5 feet, 714 inches) ; Individual 3, 166.31 cm. (5 feet, 51% inches) ; and Individual 6, 171.74 em. (5 feet, 75g inches) for a sample mean stature of 169.74 cm. or 5 feet, 6% inches. TaBLE 1.— Measurements and indices of 89PO?7 male crania ! No.1 No. 3 No. 6 Mean Maximunt leénptho. yrebre e eae id Ae eet | eee ee Maximum breadth____ Basion-bregma height- Basion-porion height -__ (20) Cranialimod wile saa) ~— 5-5 ota ce ee che ona b one Lcecoc bu coufecaecseuaman Bisyeematicgreadth:..82 2h 3 aa 8 od a ob occ ccs ce 143 Manimumifrontal/breadtht).. 5.93.8. i sc beleecxcstecccc caessecowe LT gel ets tS Endobasionenasion=: 32.82b 2.2 3 Ohta bec eewccaleccce|aeceecccccce A ee Wasa) heiphts. 2b A et. 2 Le Ser 1. Ge) to Doves atkins stoss 32 57 55 INP SON Sy et G1 ee Se ee he ee ee ee eee 26) ) | [s2eEscsazese 26 26 Orbital height: Right. ene. 2 Wh ae 8 oo: SY aa BO 90\-4-ccccnsccsadal cccc RRO is iio: 2 eee Es Se Re ee ee ee ee eee BY fee te) ass Orbital breadth: jtAg\ig.. 2) SS eere Bese) eee ee eee Bere 47. 21 passceccscesbsoscessecsse LU ie a pe SE EE 2 is ee Se eee: ee eee Ce eee a |e aE LS 5 © Jbengiichei pi os oneal rene sevecubec cums bleh one coe sececceeenae (74. 58) naa Vey eV ohil fei) ee oe eine OF Ee ee ee | ee ee enna (06,-35)|233_ oo. Indices: PVD ELL = 22 conse amupne eons ep ta gen a eas a |= Ea eR Seal | meee oa a |e BUSY a | LEE) FE 0 oa EY Bs OE el (el ee ee ee (84.08) [es Set A Dy geet ear y a} Ce) 229 eee ee Ee A eee | ee ee ee SS Ree Gh Eee i ete Se Orbital: Ce ps a ee VAJATAL Saat S212 BI Cuke BieT BE Oui Ss (uy FTk IN Og Gace od serene oe tee Sooo ceccc cent cen eeund|sscse ek een eeeLeare aee 88. 10 81.28 DPS a EE ea Ee INR 49: 06 |b. so-eescn es 45.61 47.34 RidiHessioneraninliDASO=s-ccennLscencusccesaseceous |acacesosene™ [ueeeeaeeeeee (01 1) Ba Mandible: Bigonial breadth...-.-..------- tui iii 2-Gsscd 102 (106) om Symiphysisneiehpeee nas = 2 ee oo eee te toda) errs eres (37 Bicondylar diameter_.__...._._____ 182 | ce digas astist- os S3~ Ascending Ramus height 72 76 74 Corpal length (Gonion-Gnathion)-._-.......------ 88 93) t=2..955T sce 90.5 1 Tn all tables figures in parentheses represent estimated measurements. 2 Indices of the cranium are based on measurements indicated here, 264 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 DISCUSSION Because there were no cultural artifacts associated with the burials, no positive cultural identification can be assigned. However, a com- parison of the morphological characteristics of the crania may give some indication of possible cultural affiliation. 'Table 2 compares selected cranial measurements and indices for two adult males from 39PO7 with samples of Arikara, Mandan, Pawnee, and Central Plains Phase associated skeletons. The Arikara and Mandan were inhabit- ants of the region around the Hosterman site. Recently, summaries of anthropometric measurements of skeletal material for protohistoric Arikara (Bass, 1961) and Mandan (Bass and Birkby, 1962) have been published. One of the major areas of cranial] difference between the Arikara and Mandan is in the index. The Mandan are dolichocranic, whereas the Arikara are mesocranic. The only individual from 39PO7 upon which a cranial index could be calculated gave an index of 77.40 or mesocranic. On the basis of both the length-height and mean height indices, the single male from the Hosterman site falls closer to the Arikara means. Although cultural associations based on cranial measurements of only one individual are quite speculative, Individual 6 seems to be related most closely morphologically to the Arikara. TABLE 2.—A comparison of selected mean measurements and indices 39PO7 39P07 | Arikara1! Mendsn Pawnee 1) Central! No.1 No. 6 N 1163 Nil N 294 Plains N 375 Mm Mm. Mm, Mm. Mm. Mm. Glabella-occipitalilength —---_......!____]______i 177 174.7 184. 50 175.8 174.7 Maximum (DredGhe 2-2 seen ee 137 140. 7 136. 17 139. 6 143.0 iBasion-brepma heleht--b2-. 9 eb a ee (132) 132.9 131.75 128.7 137.7 Indices: Craniqiee so hs hee OUR ee I 77. 40 78.6 73. 90 79.3 82.5 Meneth-neiwht a net) bok. a (74. 58) 74.5 71. 49 73.2 79.5 WMieanenele nig. 2. bo 202 eo oe eS (84. 08) 83. 59 82. 31 81. 55 86. 35 Wippertaciel ic. 28 25 er au eer ee aes Bs 52. 45 bd: 405 cece ses 51. 80 51.50 psa eee ee eae. 49. 06 45. 61 CA (| | cea 49. 03 51. 96 Orbital: SET ooh Sete es ar et ee RS CC i (Ere | eee ee ee) ee ns ik DBS i RE RR BAe ARE Ge 8 88. 10 SQESL sree, eS 85. 66 78. 65 { Bass, 1961. 2 Bass and Birkby, 1962. 8 The sample size varies with the individual measurement but in no case are there less than 49, 4 In no case is the sample size less than 17. 5 In no case is the sample size less than 14. REFERENCES CITED Bass, WILLIAM M. The variation in physical types of the prehistoric Plains Indians. Ph. D. dissertation. University of Pennsylvania, Philadelphia, 1961. Bass, WILLIAM M., AND BIRKBY, WALTER H. 1962. The first human skeletal material from the Huff site, 32MO11, and a summary of putative Mandan skeletal material. Plains Authrop., vol. 7, No. 17, pp. 164-177. Riv. Bas. Sur. wie: Pap. No. 35] HOSTERMAN SITE—MILLER 265 Hrpii¢Ka, ALES. 1952. Practical anthropometry. T. D. Stewart, ed. The Wistar Institute of Anatomy and Biology. Philadelphia. JOHNSTON, FRANCIS EH. 1962. Growth of the long bones of infants and children at Indian Knoll. Amer. Journ. Phys. Anthrop., n.s., vol. 20, No. 3, pp. 249-254. KroeMan, W. M. 1962. The human skeleton in forensic medicine. Charles Thomas. Spring- field. McKeErn, THoMAS W., and Stewart, T. D. 1957. Skeletal age changes in young American males. Headquarters Quarter- master Research and Development Command. Tech. Rep. No. EP- 45. Natick, Mass. Stew akt, T. D. 1954. Evaluation of the evidence from the skeleton. In Legal Medicine. Mosby, St. Louis. TROTTER, MILDRED, AND GLESER, GOLDINE C. 1958. A re-evaluation of estimation of stature based on measurements of stature taken during life and of long bones after death. Amer. Journ. Phys. Anthrop., n.s., vol. 16, No. 1, pp. 79-123. y r , : hy i pee i cag pelts ae. oponney ai st wants atc Walleags sept adult mo : ROOT w poraty pose eas phage Lat paige woe catod geld Tin Aritérn amet andar HE +a uke Pa ET ransrerg BaRs bbs t ératapusp! oH ‘zalsan asofusian Bay, al sogagiia P oY dost tag , bata 3 ia: sanila rot bite Mot pees ean a | . weit astie ae , cn . eal oh 2 ont aibotd od had “abiaibie ou’ aot ert BLYH Ok) ie dot Faerts 5 sid Y ; i th tae! "Pe ie algoik:: naa, P ‘tae “4 ar Bate, 29 aT lem nat 1 VEX wcsindarthd “Ao alnoiactaRat ‘io hozet sigtde te aoltatetiagtoe sob sinieraone ia f Mpa uijesd itis. aod 2a6fo (bah, ott! paiuib, ida), Orme iat ¥ I joes Re TG 74 Lae a i. lor P Bi fl words, a) TOT, tot ut v2 < i io wi of iia , me) : a ; { 5 . ‘ a a | i , | a i ; vi “7 - VuigTia & Ags th i a> ' e he 2 A _ BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 15 A, B, Vertical and S-shaped rim types, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE A, B, Rim types, showing typical forms and treatments. BUREAU OF AMERICAN ETHNOLOGY BULLETIN es bhi “x A, B, Typical rims with and without handles. 189 PLATE 17 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 18 A, B, Rim types, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 19 B A, B, Cord-impressed rim treatments, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 20 A, B, Incised rim sherds, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 21 icmM. ~~ YON, Double spouted miniature vessel, unusual to the Plains. Aashaee Abbaaada BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 23 GHOUP - 3 A DIN. A, B, Group 3, Group 4, and notched projectile types. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 24 ‘ é $ i é p £ i ‘ Cc WwW A, B, Microblades and lamellar flake knives. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 25 A, B, Scraper types, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 26 syasoemr A, B, Scraper types, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 27 J wee: d e f A, B, End and side scraper types. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 28 — = iom. Vein. A, B, Side scrapers and burin form, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 29 CM. TN, 5 2 A, B, Knife forms and biface ovates. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 30 A. B, Flake knives and side scrapers; grooved ax and maul. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 31 Graver forms, Hosterman site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 32 A, B, Lamellar flake tools. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 33 A, B, Cut bone sections, worked antler, bone tubes, and scapula hoes. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 34 A, B, Scapula tools; fleshers and sickles. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 35 A, B, Scapula knives and fleshing tools. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE A, B, Scapula cleaver, pentagonal-shaped tool, hafted knives, and antler tools. 36 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 37 7 A, B, Bone awl types and miscellaneous worked-bone objects. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 38 i cm. oe Fetish or trophy skull of Vulpes velox. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 189 River Basin Surveys Papers, No. 36 Archeological Investigations at the Hickey Brothers Site (39LM4), Big Bend Reservoir, Lyman County, South Dakota By WARREN W. CALDWELL, LEE G. MADISON, and BERNARD GOLDEN 661-932—64——_19 peesisrh sink oll 8 wit i af Lov! ereqs aera ‘niesti sovish (M108) odie arortors yoaloiHl orf} is enolingiteovnl [ook x sioda@ dive® zinnod nam siovies busi gi€ L VWAGION GHAVAIE bas VORA .O Tas wa WIAD W CONTENTS PAGE NRG RO GLUT CULO Tie ye eek aaron pigment Se Tee ea 271 Seieseickey Brothers, site:(G9LMM4) 2252 =. ee hee eee ese 272 MEGA Ca WOmprocedurestue > .f< 223 ot EAE. I. See AMS! 32 273 “SUDO Os Se aR ag 8 ne ae Sone CE cel Pa At Sek Mares EN SA 274 TREEHOUSE pg Rp ara een el A ee ee Lee 275 ROTH CS GLO MS oo ee Sia era Sg A ete ce eye ee ie 276 EDI UL ULO THAT C SS eee ak 2 2 SUM ee ee a ed 279 EASELS SS Sg SS rr, PV eC AO Ce TRY pe 284 pOLter yes Ban eewndton? . DEUS AUN AMAT 0k be Ce Pee ss ean 284 Hhithienna terials s Tye 5S ee Seeks OOF ace eat Saye ee ee 285 Bone WHatenial sx sex = Fk oh ee | a ey Se he eben ee Uke A ge 286 ESTOS Son a a eee ee eee eee en ea Sree ee, 287 NTGeTAUGULCACKLC Cine ere say een, a ery pegs ae a CRAM ee ee me 289 ILLUSTRATIONS (All plates follow page 290) PLATES 39. a, The Hickey Brothers site, looking across the northwestern bastion, Missouri River flood plain in middle distance; 39LM4-3. 6, Feature 4, the south- eastern bastion, prior to excavation; 39LM4-16. 40. a, Feature 4 cleared of sod and staked in 10-foot squares;39LM4-19. 6, Sur- face contour of ditch in the vicinity of Feature 5, prior to excavation; 39LM4-22. 41. a, The stockade line of Feature 4, excavation completed ; 39LM-—46. b, Fea- ture 18, postholes of the stockade along the inner edge of the defensive ditch, fire area about 5.0 feet to the east; 39L.M4—48. 42. a, Profile of west wall, Feature 5. The dark fill outlines the aboriginal ditch excavation; 39LM4-30. 6, Profile of east wall, Feature 6, showing rodent disturbance and possible postholes; 391L.M4—24. 43. a, Feature 10, excavation completed, 1958; 39LM4—40; b, Feature 25, a burned area with a concentration of sherds, possibly roof fill of a house; 39LM4—42. 44. a, Feature 39, a house of indeterminate form, postholes, firepit on pedestals; 39LM4-51. b, Celt. c, d, Shaft smoothers. e, Chopper. f, g, Fleshers. h, Scapula hoe. 7, Quill flattener. j, Shaft wrench. 45. a—d, Projectile points. e, g, h, j, m, n, Knives. f, 7, k, Scrapers. 1, 0, Pipe fragments. 46. a, d, g, Riggs Punctate. b,c, e, h, Riggs Flared Rim. f, 7, Anderson Low Rim. Jj, Cord-roughened body sherd. k, Simple stamped body sherd. 269 270 ILLUSTRATIONS 53. 54. 55. 56. 57. TEXT FIGURES PAGE The Hickey Brothers fite, Plan... 222) oso se ct ee ee (facing) 273 Man of bastinn: Heaters 4.2.2 oe ee ee. oe hae (facing) 277 Profiles, plan, Features: 5 and 72. 2-20. 2 eek ae 2 278 Profiles, plan, Features’3 and\6 2.4 2044422... 2 bee 280 Penture.o9, DIOn. =< 2.222 e ok ee 282 THE HICKEY BROTHERS SITE (389LM4), BIG BEND RESERVOIR, LYMAN COUNTY, SOUTH DAKOTA * By Warren W. Catpwert, Ler G. Mapison, and Bernarp GoLpEN INTRODUCTION The Hickey Brothers site (39LM4) was excavated during the summer of 1958 as part of the investigations of the Missouri Basin Project, Smithsonian Institution, within the projected Big Bend Reservoir of central South Dakota. The site was approached with every expectation of adding materially to the corpus of data bearing upon the “middle period” of village occupation along the Missouri main stem. The Hickey Brothers site appeared to be particularly important because it was fortified in a distinctive manner and the fortification system was so well preserved. Unfortunately, the site remains an enigma. The results of the intensive excavations of 1958 were disappointing, even shocking. An insignificant number of arti- facts was found, and beyond a detailed coverage of bastion con- struction, there was little evidence of architectural features. Since problems had been created rather than solved by the 1958 season, a brief period was devoted to further work at the site in 1959. Feature 10, a putative house depression tested the previous sum- mer, was reexcavated and a portion of the original trench widened and extended. A corner of a poorly preserved rectangular struc- ture was uncovered, but because of time limitations the work could be carried no further. Beyond a few very small potsherds, there was no addition to the artifact inventory. The excavations of the 1958 season and the initial analysis of site features and artifacts were carried out by Bernard Golden, then attached temporarily to the staff of the Missouri Basin Project. The brief investigations of 1959 were under the direction of Warren W. Caldwell, archeologist, Missouri Basin Project. The final state- ment presented here is the joint effort of Lee G. Madison, museum aide, Missouri Basin Project, and Warren W. Caldwell. Madison is responsible for the description and analysis of artifacts and site features, while Caldwell functioned largely in an editorial capacity. “Submitted December 1959. 271 272 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 THE HICKEY BROTHERS SITE (89LM4) The Hickey Brothers site is a large village area enclosed by an elaborate fortification ditch. The site is situated in Lyman County, S. Dak., on the right or west bank of the Missouri River, approxi- mately 7 miles north of the modern reservation town of Lower Brule and 35 road miles northwest of Chamberlain. This is one of a large group of important sites concentrated in the constricted neck of the great loop that forms the downstream margin of the Big Bend of the Missouri. This site lies on the eastern or riverward edge of MT-1 (Coogan and Irving, 1959), the second terrace above the river. MT-O, the current flood plain, is poorly developed here. It forms a narrow bench just below the site, widening to the south but pinching out abruptly a short distance upstream from the village. Fresh-water springs emerge from the face of MT-1 in the immediate vicinity of the site. Although the springs are scarcely usable today, there is evidence that they have produced a much larger flow in the past. To the west of the site, MT-1 rises gently toward an irregular rem- nant of MT-2, the highest terrace, which forms a jagged backbone within the loop of the bend. Immediately to the north is a deep ravine system, and to the south the village is bounded by site 39LM215, a late village occupation probably related to the Fort Thompson Focus. In the immediate vicinity of the village, the terrace surface is level and sparsely grassed (pl. 39, a). Tree cover is restricted to MT-0. Such surface irregularities as are evident consist of archeological fea- tures or result from recent ranch activities. The Hickey Ranch build- ings formerly stood on the site, and much of the area within the village was used for stock pens and corrals. The foundations of the ranch- house are still visible near the east-central border of the village. The principal feature of the site is a fortification system consisting of a well-conceived and carefully executed ditch or moat with seven projecting bastions (fig. 53). The fortified perimeter is about 2,000 feet in total length, enclosing an area of slightly less than 2 acres. The ditch is present only on three sides of the village. The eastern boundary, fronting on the river, is formed by the terrace edge. Since it is quite steep, the scarp itself would have served as an excellent de- fense, particularly if it were backed by a palisade or curtain wall. The area enclosed by the ditch is best described as “coffin-shaped.” It measures 1,150 feet along the north-south axis and 700 feet in an east-west direction. The long axis is oriented toward magnetic north. The southern arm of the ditch forms a straight line on azimuth 90° to 270°. At its western end, it angles abruptly north, expanding in a gentle arc to the widest point of the village enclosure (about 44 of mua ” a... eer ea ihe pr . fear i - aie pati Ory OF wee Popreitions, » puinber fd. AN a a ie pits tent pnts: n re 4 ou ha JERE: fens 1 ( Sing M ee ; oe a Lk : : paodunten sme + e& 3auTaaza—~-| , : At. : ij ~~ F : : a ) Ne, 7, ee reres br ry . i} \ j i } | | tar jew 4—1 2 BRUTAST ba ee ti i \ !TAVAOX a> ee : a i $8 3AUTASA f i, Reis q OP SRUTAIT a ly ] — a ne HICKEY BROTHERS SITE 39LM4 [OO FEET Yl ~ a = < DEPRESSION Ti = = ROAD) c——1 AREA OF EXCAVATION - O 2&6-T199 (LZg ‘d aoeq) 79 Riv. Bas. Sur, wees Map. No. 36] HICKEY SITE—CALDWELL, MADISON, GOLDEN 273 the distance from the southern end), where it swings to the northeast in a wide curve back toward the terrace edge. The seven bastions are all similar in form and size. Each is roughly semicircular, averaging 40 feet wide and projecting outward about 40 feet from the moat line. Bastions were built at the corners of the village enclosure and in the intervals between. The distances sepa- rating bastions, proceeding in a clockwise direction from the southeast corner, are as follows: Feet No. 1 to No. 2—310 No. 2 to No. 3—310 No. 3 to No. 4—310 No. 4 to No. 5—250 No. 5 to No. 6—240 No. 6 to No. 7—230 The distance between adjacent bastions decreases along the curved portion of the ditch that forms the northern edge of the village. This fact suggests that the location of each defensive feature was carefully planned in relation to the whole. Reducing the distance be- tween the bastions on a curved section of wall was a necessity if the enfilade was not to be lost. Otherwise there would be blind areas along the wall which could not be reached by the defenders’ fire. It is worth noting that the bastions at the Black Partizan site (89LM 218), a short distance to the south of the Hickey Brothers site, average only 150 feet apart. Here the ditch forms a great oval, thus demand- ing a shorter bastion interval. The fortifications at the Hickey Brothers site, from a functional point of view, were efficient and well thought out. Within the area enclosed by the defensive ditch are a number of shallow, circular depressions of various sizes. The largest are about 50 feet in diameter and seldom deeper than 0.5 foot at the center. Such depressions are scattered widely over the site but with no ap- parent pattern of arrangement. Since they appeared to be house depressions, a number were tested, both by area excavation and trench or pit tests. In only three cases was there firm evidence of aboriginal occupation (see below). EXCAVATION PROCEDURE A grid system was utilized in conjunction with a detailed map of the site made by a Missouri Basin Project reconnaissance party in 1956. A north-south base line, paralleling the long axis of the site, was established from a primary datum located at the southwest cor- ner of the ranchhouse foundation mentioned above (fig. 53). An east- west line, perpendicular to the initial base line was extended from the primary datum to a point beyond the western edge of the ditch. eae k= 16 o V4, \Iy aS a 7U{NS \ AWS FEATURE 17 \ly, Zoe / 70\* EDGE OF TERRACE FEATURE 26" 5 _FEATURE 20 2 Witz, es - N —te : N in “iy AN O—FEATURE 21 eee f HOUSE FOUNDATION ae FEATURE I5—O aN Sit Wie: / FEATURE 8 =~ = me XS [ : ji =—— < FEATURE 14 ; “TANS \ 6) FEATURE 13——O a / \__| | —=FEATURE 18 \ DATUM. STAKE NUFEATURE 25 (N1O00, E1000) FEATURE 27——O FORTIFICATION DITCH 7 FEATURE i2-{-o | Feature a FEATURE 28——O yi O—FEATURE 19 / < FEATURE 38 ae Soe 36 os (7 ConINEE 35 Q 9 Oo F ] 4 ki o Qo fy O——FEATURE 33 FEATURE 37 FEATURE 29~ ‘ 7 7 FEATURE ee ie we : Fear 6 & | |\’ FEATURE 10 i ie \ee aes “EXCAVATION, 1959 FEATURE 30—4—_O FEATURE22—— od FEATURE 40—}| /| Sf. / FEATURE 31 Tay FEATURE 41——p dl FEATURE 24——0 —FEATURE 23 HICKEY BROTHERS SITE / O|—FEATURE 32 FEATURE 5 Z49LM4 / \ 100 FEET DEPRESSION sw OAL FEATURE 9——-C——4 ——J] AREA OF EXCAVATION Figure 53.—The Hickey Brothers site, plan. " 1 ig aw LP Baty itY At dil re 4 Ata " ¥ i iw 274. BUREAU OF AMERICAN ETHNOLOGY [Bull, 189 The datum stake was arbitrarily designated as N1000, E1000, thus confining the entire site to the northeast quadrant of the grid, with the zero point 1,000 feet south and 1,000 feet west of the primary datum. All points on the grid are given as coordinates, first north from the east-west zero line, then east from the north-south line, i.e., N700, E800. : Stakes were set at convenient intervals for the purpose of laying out excavation units. The only such unit not planned in this way was Feature 3, since work on it had begun before the grid was established. Several excavation units consisted of 5.0-foot and 10.0-foot squares. These were designated by the position of the grid point in the south- east corner of the square. For vertical recording, the top of the primary datum stake was ar- bitrarily assigned an elevation of 100.0 feet and all other elevations were related to this point. Elevations were systematically recorded at the top of each grid stake and at the ground surface next to it via stadia rod and transit. The provenience of artifacts and other material of special significance was recorded by means of two horizon- tal coordinates on the grid and a vertical measurement taken from the ground level nearest the find or from a datum line based on a grid stake. All obviously recent materials from at or near the sur- face, unidentifiable fragments of unworked bone, stones, and small undecorated sherds, were discarded. In each case, however, a note was made of the quantity and level in which they appeared. After the sod was removed from an area to be excavated, the under- lying fill was skimmed with shovels so that architectural or other fea- tures might be seen as soon as they appeared. 'Trowels were used to scrape a clean floor at the end of each level and also to profile the trench and pit walls. In a few instances, where little or no cultural material was apparent, 0.5-foot increments were removed by spading. Except Feature 3, which was dug by natural levels, all excavation units were divided into 0.5-foot vertical intervals. Profiles in all excavation units were scraped clean and the stratigraphy noted. Earth pillars were left to support the grid stakes, and in Feature 4, the southeastern bastion, north-south and east-west balks remained after excavation. In most cases, the trenches and pits were excavated to a depth of 4.0 feet below surface, well into the gray silt or loess that underlies the site (see below). STRATIGRAPHY A uniform stratigraphic sequence was found throughout the site. Three distinctive units were present : Zone 1—surface to 0.4-0.8 foot is composed of dark brown, moist humus con- taining many small rootlets. Aboriginal material occurred in this de- posit, together with an abundance of modern refuse, Riy. Bas. Sur. r e .. - Riv. No 36] HICKEY SITE—CALDWELL, MADISON, GOLDEN 275 Zone 2—0.4-0.8 foot to 1.0-1.4 feet below surface is limited to a dark-brown humus horizon. It is lacking in some portions of the site so that locally, zone 1 may be in direct contact with zone 3. Artifacts were present in this layer together with small quantities of recent material. Zone 3—1.0-1.4 feet to limit of excavation is a dry, compact, silty deposit (loess), light gray in color. Aboriginal material was found in the upper part of this layer. As noted above, the majority of excavation units reached a depth of only 4.0 feet below surface; however, the central pit in Feature 8 was carried to 12.0 feet. Here there was a gradual transition (3.5-5.0 feet) from the matrix of zone 3 toa layer of silt and sand. The sandy texture increased from 5.0 to 6.6 feet. From 6.6 to 7.5 feet the deposit consisted of brown clay with areas of dark carbon staining. The pit terminated in gravel at 10.0 to 12.0 feet. In the section of the defensive ditch excavated adjacent to Feature 4, the southeast bastion, there was a black midden deposit slightly below the present ground surface. It undoubtedly accumulated as refuse thrown in the ditch (see p. 276). FEATURES The term “Feature” was applied to anything at the site which required description or discussion. After designating the site as Feature 1, and the defensive system as Feature 2, numbers were as- signed consecutively to cultural phenomena and to each excavation unit (i.e., trench or test pit) (see fig. 53). Actually, only a small proportion of the features represent the former; the majority desig- nate excavation units. In the course of the 1958 season, a number of the surface depressions were examined and a large portion of the site was investigated by means of test pits. In addition, one bastion was excavated, a section of the palisade wall was exposed, and two transverse trenches were cut across the defensive ditch. Test trenches——Four trenches were excavated; three of them (Fea- tures 8, 10, and 11) were intended to test well-marked depressions within the village enclosure. Feature 9 was excavated beyond the southern limits of the defensive ditch in order to test for cultural material outside of the fortified area. The test trenches were 5.0 feet wide and ranged from 35.0 to 50.0 feet long. Each was dug to a depth of 2.0 feet below surface, then a 5.0 by 5.0 foot section in the center of the trench was continued downward to ascertain if cultural material lay at a greater depth. In Features 8 and 9, the fill was entirely sterile. Metal, brick, crockery, and other objects of recent origin were present on the sur- 661-932-— 6420 276 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 face of Features 10 and 11, occurring to a depth of 0.5 foot in the former. The reexcavation of Feature 10 during the 1959 season pro- duced some evidence of a rectangular structure; this will be described in greater detail below. Test pits.—Since the initial trenches proved to be largely unproduc- tive, a number of smaller test pits were excavated in an attempt to locate more intensely occupied areas. Ficune 54.—Plan of bastion, Feature 4. BALK ty VG au ie a ; ee . Ly Pd [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 278 vauy aseunisia fr] sowon KS anid yava GRREEE] NOISMION! 1H9IN- \ "1 pue ¢ soinqeag ‘uvjd ‘sajyorg—*s¢ a¥undIy 37150Ud LS3M‘2 3uNLIVaS 31!S SY3H1ONS ASNOIH pW I16e B1308d 1S3M‘S 3YuNLV34 (11s) ‘e-3N0z [__] (SAWNH O10) ‘2 3NOZ [- (snwnk) | 3noz _ ZZ Riv. Bas. Sur. Pap. No. 36]. ICKEY SITE—CALDWELL, MADISON, GOLDEN 279 Five feet east of the stockade line and at a depth of 1 foot was a saucer-shaped area (1.3 feet in diameter, 0.3 foot thick) filled with soft white ash surrounded by reddish, burned earth. Sherds (Riggs Flared Rim in zone 1, plain body sherd in zone 2), stone artifacts, and fragments of animal bone were in association. HABITATION AREAS Evidence of domestic occupation or related phenomena was found in fiveareas. The following features are involved : Features 3 and 6 Feature 10 Features 25 and 14 Features 39 and 12 Features 40, 22, 23, and 41 Features 3 and 6—Feature 3 was a pronounced circular depression about 30 feet in diameter, situated in the south-central portion of the village enclosure. An area 30 by 35 feet was carefully skimmed to a depth of about 1.0 foot, following natural stratigraphic levels. Since little cultural evidence was appearing, work was concentrated in the southern one-half of the excavation, skimming the floor to 2.0 feet be- low surface. Two small tests were carried to a slightly greater depth. No structural features were noted, but the profile of Feature 6 (fig. 56, pl. 42, 6), a 5.0 by 5.0 foot pit excavated to a depth of 5.7 feet, suggests that a nebulous house floor may have been present. Some of the de- pressions originally labeled as rodent burrows are strikingly like postholes in profile. Again, the southeast profile of Feature 3 (fig. 56) indicates at least one, and perhaps two, cache pits deriving from the same level. The few artifacts from this feature were found above the 2.0 foot depth, mostly in the upper 0.6 foot of zone 3. Five rim sherds of the type Riggs Flared Rim and two of Anderson Low Rim were recovered from the 0.0 to 1.4 foot level. From 1.4 to 2.0 feet, the upper portion of zone 8, only Anderson Low Rim sherds were reported. This level is coincident with the putative house floor suggested in the previous paragraph. Other occupational debris included stones, unidentifiable bone fragments, and stone chipping detritus. Feature 10.—Feature 10 was a circular depression, approximately 50.0 feet in diameter, situated 340 feet south of the primary datum. An east-west trench (pl. 43, a), 45.0 feet long and 5.0 feet wide (be- tween N660-N665 and E940-E985) was excavated in 0.5-foot levels to a depth of 2.0 feet, following the surface contour. A central pit, 5.0 by 5.0 feet, was dug to a depth of 4.0 feet below surface. The stubs of five fenceposts and numerous items of recent manufac- ture were found in the 0.0- to 0.5-foot depth interval. These were in- [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 280 9 pur ¢ soinqeag ‘ueid ‘sajyorg—'g¢ 31150Yd LSVSHLNOS ‘¢ 3uNLVSS AVYD LHOIT Za EEE 1334 ¢ vauy a3zsunisia Ee4 (11S) ‘€ 3NOZ [__] (SAWNH O70) ‘2. 3N0Z FE22] (sawnk) ‘1 3Noz [7 51)430Yd LSVSHLYON ‘9 3uNLV34 VuS WNIGaN JLIS SYSHLOUG ASMOIH DW 16¢ AUN] Nvid ‘9 ONY € S3uNniv3ad Riv. Bas. Sur. ap. No. 36] HICKEY SITE—CALDWELL, MADISON, GOLDEN 281 cluded within the zone1 layer. Zone 2 began at about 0.6 foot and per- sisted toa depth of 1.8 feet. From 1.8 to 3.3 feet, in the central pit, was a brown fill, overlying the matrix of zone 8. The surface of the brown fill was broken by several small pits that resemble postholes in profile. In 1959, the central pit and a portion of the trench were reexcavated and, in addition, an irregular area 7.5 feet (east-west) by 18.0 feet (north-south) was opened at the western end of the feature (fig. 53). At a depth of 1.0 foot below surface, a distinct “wall line” was found. The fill was then skimmed to a “floor” at 1.5 feet below surface. Sev- eral posts were found in the process, but they could not be definitely related to either floor or wall line. The resulting pattern suggests a fragment of a rectangular structure with a rounded corner. The floor level at 1.5 feet is approximately the same depth as the putative posthole level observed in the central portion of the trench. Unfor- tunately, no additional structural details were observed, and artifacts were limited to several small, plain body sherds. Features 25 and 14.—Test pit 14 was situated in the north-central part of the village enclosure. When a patch of carbon staining was found at 0.5 foot below surface, the excavation was expanded to in- clude an area of approximately 13.0 by 15.0 feet and labeled as Feature 25 (pl. 48,6). Additional irregularly shaped, dark, mottled areas, to- gether with areas of fire-reddened earth, were found just beneath the sod line. One of these contained the remains of a charred log. Four large unworked stones were found in the center of the area, but they had no apparent arrangement. An unidentifiable bone fragment, a piece of glass, and some sherds were the only specimens recovered. All came from near the surface of zone 1. The only rim found falls into the Riggs Punctate category. No structural elements, such as post molds, were observed, and the area had obviously been disturbed by ranch activity, but Madison and Caldwell do not discount the possibility that the fill represents roof dirt, overlying a more deeply buried earth-lodge floor. Features 39 and 12.—Test pit 12 was intended to sample a shallow depression located near the east edge of the site. The initial test was enlarged to an area 27.0 by 30.0 feet (Feature 39) following the dis- covery of three post molds and fragments of animal bone. The entire excavation unit was stripped to an average depth of 0.8 foot (pl. 44,a). Twenty-one additional post molds and postholes were found, occurring from just below surface to a depth of 1.0 foot (fig. 56). They ranged in depth from 0.10 to 2.5 feet and in diameter from 0.138 to 0.72 foot. More than one-half contained wood or charcoal. In one instance a fragment of milled timber was found; several other posts may also be of a recent date. 282 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 TRENCH TRENCH ° ® BISON BONES—;? ane e BISON eee 8 uf « RECENT POST HOLE——9O SARTIFACTS > \ ARTIFACTS ——> O—RECENT POST HOLE CONCENTRATION OF SHERDS st AREA OF RED BURNED EARTH "5 CONCENTRATION OF CHARCOAL BS. ae, e \ Bison SCAPULA——C - @4 Z fi FIREPIT—+— | TRENCH \ LIMITS OF BURNED AREA—\ FEATURE 12 (TEST PIT) FEATURE 32 39LM4 TRENCH Ficure 57.—Feature 39, plan. A circular firepit filled with compact white ash was found slightly below the surface. Just to the east was an irregular area of reddish earth, mottled with carbon stains and charcoal fragments. The out- lines of charred maize cobs were also visible. Artifacts were found to the north and west of the firepit. Eight rim sherds of the type Riggs Punctate were present in the 0.0 to 1.0 foot interval (zone 2). In addition, a knife of Bijou Hills Quartzite, an end scraper, irregular flake scrapers, four shaft smoothers, a split metapodial flesher, a quill flattener, and a possible scapula hoe were found in this zone in association with glass and other recent objects. Despite the obvious disturbance of Feature 39, it is possible that the true floor level was not reached. The concentration of charcoal and burned areas suggests roof fall or above-floor fill. Similarly, the excavation may not have been enlarged enough to uncover the entire post pattern. Posthole Nos. 1, 2, 3, and 4, oriented about the firepit, suggest the four central roof supports of a circular earth lodge (fig. 57). On the other hand, post molds can be selected to Riv. Bas. Sur. Pap. No. 36] HICKEY SITE—CALDWELL, MADISON, GOLDEN 283 form the corner of a rectangular house. The case remains unproven, but considered with the associated rim sherds, the latter house type would appear to be the best possibility. Features 22, 23, 41, and 40.—Test pits 22 and 23 were excavated adjacent to the terrace edge. Since a number of artifacts were found, the two test pits were connected by a trench 95.0 feet long and 5.0 feet wide (Feature 40). Near the center of the trench, the strati- graphic sequence varied from the usual. Intervening between the uppermost humus (zone 1) and the old humus deposit (zone 2), was a brown deposit (0.7 to 0.8 foot thick) divisible into a darker upper half and a lighter basal portion. Zone 1 extended to depth 0.4 foot, zone 2 from 1.1 to 1.4 feet, and zone 3 below 1.4 feet. Two aboriginal pits were found within the limits of the trench. Near the center, originating at depth 0.4 foot (the surface of the brown matrix) and extending to 2.0 feet, was an excavation of in- determinate shape. It was filled with a sterile “whitish” fill, probably ash. About 3 feet to the north was a globular cache pit with a con- stricted neck. The pit apparently originated at or somewhat below the surface of zone 3. The mouth was approximately 1.6 feet in diameter, widening to 2.2 feet maximum width of body. The convex bottom extended to 2.2 feet below the upper margin of zone 3. The fill contained flecks of charcoal and ash, a few plain body sherds, and rodent bones. About 25 feet north of the cache pit was a burned area consisting of reddened earth and flecks of charcoal. At the south end of the trench, near Feature 23, another firepit (Feature 41) was discovered at a depth of 0.9 foot below surface (brown stratum). The fire basin was a shallow (0.5 foot) saucer-shaped depression 2.3 feet in diameter, containing a mass of white ash. A few sherds were present adjacent to the fire area. The Riggs Punctate examples with fingernail im- pressions were concentrated here. Three post molds were found close to the surface near the north end of the trench. Two contained wood fragments. Diameters ranged from 0.35 to 0.55 foot, depths from 1.0 to 2.0 feet. ‘The posts were spaced about 15.0 feet apart on a north-south alinement, suggest- ing that they are the remains of a recent fence line. With one exception, a plate chalcedony knife from zone 3, all arti- facts were restricted to zone 1 and the underlying brown stratum. Six rim sherds of the type Riggs Punctate and two of Riggs Flared Rim were excavated in association with leaf-shaped knives of Bijou Hills Quartzite, one plate chalcedony knife, a single end scraper, a single irregular spall with worked edges, and a large metapodial flesher. 284 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 ARTIFACTS The number of artifacts excavated from the Hickey Brothers site is inordinately small, totaling only 479 specimens, of which 485 are potsherds. Artifacts of bone, stone, and shell are largely undis- tinguished and are not of great help in suggesting areal or temporal relationships. ‘The most that may be said is that the bone and lithic sample is not out of character with the pottery. POTTERY A. RIM suERDS: The rim sherds fall within two well-recognized pottery series, Thomas Riggs (Hurt, 1953) and Anderson (Lehmer, 1954). Riggs Punctate: 15 sherds, portions of 5 vessels (pl. 46, a, d, 7) Temper: Moderately abundant sand and grit, probably crushed or de- composed granite, diameters to 2.0 mm. Texture: Medium, some specimens well smoothed. Color: Tan to dark gray (interior, exterior, and core). Hardness: 3.0 (Moh). Surface treatment: Plain. Decoration: Limited to fillet applied below lip. Motifs include deep hemi- spherical punctations (pl. 46, g), hemiconical notches (pl. 46, a), wide gouges, and oblique fingernail incisions (pl. 46, d). Rim profile: Straight to slightly flaring. Rim height: 35.0-45.4 mm. Rim thickness: 4.1-7.0 mm. Lip: Rounded to slightly beveled, thickened into exterior flange in a few cases. Thickness: 5.0-7.0 mm. Rim modification: None. Remarks: No additional indication of vessel form. The Hickey Brothers sample lacks the lip decoration reported for the collection excavated from the Thomas Riggs Site (Kleinsasser, in Hurt, 1953). The oblique fingernail incision does not seem to have been reported previously. Riggs Flared Rim: 9 sherds, portions of 4 vessels (pl. 46, B, c, e, h). Temper: Moderately abundant sand or decomposed granite; diameter of in- clusions is particularly large, ranging to 3.0 mm. Texture: Medium to fine, one specimen (39LM4-108) has a slight luster. Color: Tan to dark gray, core gray. Hardness: 3.5 (Moh). Surface treatment: Plain. Decoration: Restricted to lip. Motifs include deep punctation and notching (39LM4-108). Rim profile: Flared. Rim height: 50 mm. Rim thickness: 7.0-9.0 mm. Tip: Rounded and flattened, often distorted through application of decoration. Thickness: 6.0-9.0 mm. Rim modification: None. Remarks: No additional indication of vessel form, Riv. Bas. 3] | HICKEY SITE—CALDWELL, MADISON, GOLDEN 285 Anderson Low Rim: 7 sherds, portions of 2 vessels (pl. 46, f, i). Temper: Sand and grit, with inclusions to 3.0 mm. diameter. Texture: Medium. Color: Tan to gray, exterior, interior, and core. Hardness: 4.0 (Moh). Surface treatment: Smoothed; smoothing over cord roughening is apparent on two sherds. Cord roughening appears on lower rim and shoulder in one instance (89LM44/2). Decoration: Oblique punctates or incisions on lip bevel and lip-rim juncture (pl. 46, f), hachure on lip (pl. 46,7). Rim profile: Flared. Rim height : 19.0-30.0 mm. Rim thickness : 2.5-5.9 mm. Lip: Everted or beveled. Remarks: No rim modification, vessel form unknown. B. Bopy SHERDS: Plain: 276 sherds; temper, color and hardness are similar to the rims de- scribed. In addition, there is a group of 60 sherds with slightly burnished surfaces. Cord roughened: 36 sherds (pl. 46,7). Temper: Similar to described rims. Color: Tan to dark-gray exterior, interior, and core. Hardness: 3.5-4.0 (Moh). Simple stamped: 30 sherds (pl. 46, &), temper, color, and hardness are similar to the rims described. Brushed: 1 sherd. LITHIC MATERIALS A. PROJECTILE POINTS (4 specimens). 1. Small, triangular, straight edges, flat to slightly concave bases, with medium-to-wide side notches (3 specimens) (pl. 45, 6-d). Two addi- tional body fragments probably belong with this group. Length: 21.0-22.9 mm. Width: 4.6 mm. Thickness: 2.0-2.7 mm. 2. Small, triangular, straight edges and slightly concave base (39LM4-64) (pl. 45, a). Length: 18.0 mm. Width: 3.9 mm. Thickness: 4.0 mm. B. KNIvEs: 1. Large leaf-shaped blades (fragmentary) of Bijou Hills Quartzite, bi- facially flaked, retouched edges, cross section biconvex (5 specimens) (pl. 45, h,n). Length: —. Width: 26.0-58.0 mm. Thickness: 9.0-11.0 mm. Angular intersection: 30°-50°. 2. Fragments of plate chalcedony with retouched edge (4 specimens) (pl. 45, g, j, m). 391.M4-30 may also have been used as a drill or reamer. 286 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 8. Trapezoidal blade, bulb of percussion intact, minute retouches along one edge, longitudinal flake scars with hinge fractures (39LM4-58) (pl. 45, e). Length: 54.0 mm. Width: 29.0 mm. Thickness: 8.0mm. C. SCRAPERS: 1. Large end scrapers, thick, heavy flakes, retouched from one surface to form a thick body with a low keel (5 specimens) (pl. 45, 7). Length: 29.0-58.0 mm. Width: 25.0-42.0 mm. Thickness: 10.0-15.0 mm. 2. Irregular spalls with purposefully retouched or use-retouched edges (10 specimens) (pl. 45, k, f). D. MISCELLANEOUS STONE OBJECTS: 1. Biface chopper or large blade formed from a massive flake, convex tri- angular body, irregular convex bit (89LM4-22) (pl. 44, e). Length: 90. 0 mm. Width: 37.0 mm. Thickness: 23.0 mm. 2. Uniface chopper formed from a quartzite cobble spall, chopping edge formed by removing large flakes from cortical surface (39LM4-61). Length: 109.0 mm. Width: 76.0 mm. Thickness: 37.0 mm. 8. Polished celt (diorite), wide asymmetric bit tapering to a flattened poll, bit edge crushed (39LM4-53) (pl. 44, b). Length: 120.0 mm. Width: 47.0 mm. Thickness: 31.0 mm. 4, Ovoid hammerstone of granite, deeply pecked over most of surface (39LM4-182). Length: 72.0 mm. Diameter: 61.0 mm. 5. Arrowshaft smoothers of sandstone, elongate ovoid body, one fiat sur- face with deep longitudinal groove, body cross section irregularly convex (4 specimens, 2 fragmentary) (pl. 44, c,d). Length: 40.0-78.0 mm. Width: 30.0—-40.0 mm. Thickness: 18.0-26.0 mm. 6. Fragmentary pipe bowl, material unknown, subrectangular bowl, thick rectangular flange projecting around orifice (89LM441, 45) (pl. 45, J, 0). 7. Tubular concretion (?) (39LM4-98). III. BONE MATERIALS A. Bison metapodial flesher (fragmentary), cancellous tissue removed, proxi- mal end of bone forms butt, body beveled abruptly to form a thin, serrated bit (89LM4-168) (pl. 44, 9). Length: 180.0 mm. Width: 32.0 mm. Thickness: 13.9 mm. Riy. Bas. Sur. aoh‘a Pap. No. 36], HICKEY SITE—-CALDWELL, MADISON, GOLDEN 287 B. Bison metapodial flesher, split segment of shaft, distal portion beveled to a broad, thin bit (89LM4-190) (pl. 44, f). Length: 154.0 mm. Width: 30.0 mm. Thickness: 38.3 mm. C. Pottery tool or quill flattener (fragmentary), split section of bison rib with rounded end, cancellous tissue partially smoothed (39LM4-155, 156) (pl. 44, i). Length: 65.0 mm. Width: 14.0 mm. Thickness: 5.0 mm. D. Arrowshaft wrench, segment of bison rib perforated with gouged hole (39LM4-78) (pl. 44,7). KE. Proximal fragments of bison scapula; 39LM4-170 has the suprascapular spine removed (3 specimens) (pl. 44, h). CONCLUSIONS It is obvious that the occupation of the Hickey Brothers site was not intensive, yet despite the relative scarcity of artifacts and archi- tectural data, it is altogether probable that two components are pres- ent. Pottery forms the most distinctive unit of the artifact inventory. Of the total group of rim sherds, 72 percent (23) resemble the typical pottery from the Thomas Riggs site. The remaining nine rims are plainly Anderson Flared Rim Ware. The small quantity of cord- roughened body sherds (about 1 percent) is somewhat less than one might expect, yet it does provide a basis for the inference that a con- siderable proportion of the Anderson pottery had smoothed bodies. Pottery of the Riggs grouping occurs only in the upper portions of the cultural deposit, particularly in zone 1. Anderson Flared Rim ware is restricted to zone 3, particularly to the surface and uppermost part of that zone, and to the lower part of zone 2. The stratigraphic occurrence of nonceramic artifacts is preponderantly in the upper zones in association with the Riggs pottery. The assemblage contains nothing not reasonably “at home” in the Thomas Riggs context (see Hurt, 1953). Plate chalcedony knives, end scrapers of the type de- scribed above, and split metapodial fleshers apparently are not char- acteristic of earlier horizons. Recognizing the disabilities imposed by the small number of arti- facts, still it is not unreasonable to postulate the presence of two components, the early represented by the few Anderson sherds, the late by the Riggs-like occupation. While the architectural data is even less adequate, it is a fair inference that long-rectangular houses were present, too. The limited information from Feature 10 offers good support for such an assertion. Evidence from other features is in- conclusive, although one might contend that Feature 39 is a house 288 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 representative of the Thomas Riggs occupation. Stratigraphically, it falls within the Riggs pottery zone. Unfortunately, there is no direct data linking the fortification sys- tem with either occupation. A few cord-roughened body sherds, pres- ent in Feature 4, might suggest an Anderson tie, but the fortification itself is more characteristic of a later time, in fact it resembles the Huff site in south-central North Dakota, in most significant respects. The Hickey Brothers bastion is longer and narrower in proportion than the excavated example at the Huff site (Howard, 1959, personal communication) and the “tower room” or strongpoint features at the Huff site are lacking at Hickey Brothers. It is worth noting that the situation at the Hickey Brothers site suggests that the village defenses were completed prior to the large-scale building of houses. Tentative as the foregoing observations are, they cannot be con- strued to offer any explanation for the negligible occupation found at the Hickey Brothers site. The insignificant artifact return and the diffuse architectural remains are difficult to rationalize with the large- scale, well-engineered project represented by the village fortifications. A number of possible explanations come to mind, but there is no real support in fact for any of them. One of the more enticing suggestions is based in the possibility of a local ecological shift following the beginning of village construction. A season of severe drought, or even a large prairie fire, might have made the area temporarily uninhabitable, or might have destroyed usable building material. Today there is virtually no bottom land adjacent to the site. An ice jam or a shift in riverflow by some other agency could well have reduced the bottoms to such a degree that the tillable land available to the village could not support the expanding settlement. The explanation might even be more simple. A mere change in leadership, the heavy toll of an epidemic, or pressure from enemies might have made a shift of locale necessary before the village was really well established. A sudden attack could have ended the in- cipient village at a single blow. The concentration of charcoal in house areas, particularly Features 25 and 39, offers some support for this view. The sum of these speculations (and postulations) amounts to an enigma. The data will support no firm statements. In extirpation of the insecure position that we have been forced to assume, we do not feel too self-conscious in restating the useful cliche—more work needs to be done to solve the case of the unfinished village. Riv. Bas. Sur. Pap. No 36] HICKEY SITE—CALDWELL, MADISON, GOLDEN 289 LITERATURE CITED Coogan, ALAN H., and IRvING, WILLIAM N. 1959. Late Pleistocene and recent Missouri River terraces in the Big Bend Reservoir, South Dakota. Journ. Iowa Acad. Sci., vol. 66. Hurt, WESLEY R., JR. 1953. Report of the investigations of the Thomas Riggs site, 39HU1, Hughes County, South Dakota, 1952. South Dakota Archaeol. Comm., Archaeol. Studies, Cire. No. 5. LEHMER, DONALD J. 1954. Archeological investigations in the Oahe Dam area, South Dakota, 1950-51. Bur. Amer. Ethnol., Bull. 158, Riv. Bas. Surv. Pap. No. 7. mie ee ote do riansi:, } tar ante vii: m? in ; ‘The Hiskias } Watt af loner tried a ey jelodatl iio note eas ade ort st! euottmyliooiidh ia nis AM, gal .r102 wag ial, PAL Aud ,fondnt sammA ‘jiyd Lite 2% pact Ou Pier rate . Ete si ing $< ju’ ive vit Fis ait eee Cre 4 a» ibn 4 fa they a i Stig Ra bs CG ptitir Wy the la reredibe aia OF pee ‘ » « i ‘ bs 7 iy ‘ ne Ty tio § iv wary BP Oral tae “Sieg Le d ; ' F ( ‘ aT wry, <7 us . el ae ay. hae aU? ici aly rithy ann yea pis i) iy ‘4 iy hig 2? uth are * Vk ricant artifact mies it Sabena wetvis is lithoul’ io rations tive witty ii ty t ; A yi Pa 7 vA MAYS hee a + , f , : Be 7 bi cx AF 2!) ive Gehtivey iG, eat, (oe). 1S ee “tty i 7 - 7 fot ; j : ur tw? that ures @ th TEKS vie Mase Sa ; 116 | ean, | aged ie | ‘ ¥ fh NTE? OF VE id’? is sehr et A ; . ay ia aes wr rout +. tal gtk hae beg j . 5 | ’ 7 pi) | i \ aki stry ee Lie? FTA Ti h merry Ait (na i 4 . iT" ' L aaets > & ‘ ) ei “(s ‘ a ; (iit : A Oe i att ihe Vs tT OMe ' yan ‘20 TOR VE a we alae 7 } Fig; i : 2 5 ” fi sanig tv. A rN Li Ais mn Tar % ry ; ; ' POLO Cla aye 4 p =a 7 | . be — dae alt: i ha nae o ine Ab ss 7 rrIAA ie reat ; —_ * i : J Ae i « ‘ pai : ¥% ate > i Patil sy bear: yi a in ni 2 Wy 4 ,M ri 40 SET ae a . : ; : ‘ x F ‘ : ‘ mis Pw. 4 en L1 fie ag rare , BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 39 a, The Hickey Brothers site, looking across the northwestern bastion, Missouri River flood plain in middle distance; 39LM4-3. 6, Feature 4, the southeastern bastion, prior to excavation; 39]L.M4-16. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 40 a, Feature 4 cleared of sod and staked in 10-foot squares; 39LM4-19. », Surface contour of ditch in the vicinity of Feature 5, prior to excavation; 39LM4-22. EE etree en BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 41 a, The stockade line of Feature 4, excavation completed; 39LM4~+46. b, Feature 18, post- holes of the stockade along the inner edge of the defensive ditch, fire area about 5 feet to the east; 39L.M4-48. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 42 a, Profile of west wall, Feature 5; the dark fill outlines the aboriginal ditch excavation; 39LM4-30. , Profile of east wall, Feature 6, showing rodent disturbance and possible postholes; 39LM4-24, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 43 a, Feature 10, excavation completed, 1958; 39LM4-40. 5, Feature 25, a burned area with a concentration of sherds, possibly roof fill of a house; 39LM4~42. BULLETIN 189 PLATE 44 BUREAU OF AMERICAN ETHNOLOGY ARN ‘ d J a, Feature 39, a house of indeterminate form, postholes, firepit on pedestals; 39LM4-51. b, Celt; c, d, shaft smoothers; e, chopper; f, g, fleshers; h, scapula hoe; i, quill flattener; j, shaft wrench, a BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 45 ————___— 3 CM a-d, Projectile points; e, g, h, 7, m, n, knives; f, 7, k, scrapers; J, 0, pipe fragments. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 46 Ceramic remains from Hickey Brothers site. a, d, g, Riggs Punctate; b, c, e, h, Riggs Flared Rim; f, i, Anderson Low Rim; j, cord-roughened body sherd; k, simple-stamped body sherd. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 189 River Basin Surveys Papers, No. 37 The Good Soldier Site (39LM238), Big Bend Reservoir, Lyman County, South Dakota By ROBERT W. NEUMAN 291 661-932—64——21 Bic ie pes ae CERO. medtsat! ope: i 7 CONTENTS PAGE MHpEUCUCEION 2 Sees = te ee ee eee Bee eee ra EPR Dene ee I) 295 Acknowledements._._..........22e Meee? sal Sai i BOI OF 2 DORL OE ATE 295 ieenhion ane cescripwon- woes ee Pa OT MOL) So uiosgs ST ISt 296 PReavauionses = So. oe wee see ee UME DIAS EEE LD TOT) oh 296 Good Soldier’Component (39 M238)i 2 2Ye_ = Ve LL ee iol) apie 297 Peabures. 8) Se ee eee ee ee ME Bet SOE DSA are 297 Aytifaets 2. 2S SS eG as eae Oils HOT oem oede sroie hin yyat 299 GUSTY See Ae ee eee ee Se ee ee 299 LGina PIMeN be GS ae ees ome Lae ome) ee Sa eS ee eee 300 Grey (Cloud Horizontal=Inciseds 4.0) 2 8 pe ee ee 300 dang Crow Stren Rint. - 2° a he bee eee 301 Gadoibe Collared =7- 22 ose = er Pe ee Pos eee 301 Sasnley “Pool lmpressed 2.22 3. se ke ee 302 MiGeIIANeG Use = orcs Mee ea es Sep ta ae a ae 302 SEU E05) Ae yA ae nah i el ace ye Aah at he OR Yocom 303 ineajectila pomtes oe. eros AEA aes SURES Rey ee 2 303 Rng SCrapersee 5 24g) oho t A iant he cig Bk Reach anne apple, 303 LCi (S.A a OS MORE. -- ARE. RENE: Wy eves. mire 304 OPO) eS a he ai Biche a SY CO eae Le nl 304 Miscellaneous thipped'stone!..... 2-2. 22.5.5 2-222 2l ee 304 Ground istone artifacts 32222 ee Ts ee See 304 MEL SMM ODENEEsE Mt 2 ee oe Ue Le i ek See 304 HELGE 2 eet ne Sch Sm ENP Re ee or ee ap a Ne al Poe 305 Bone ariiiacts: <. 5 CE See Verte Pee ROP OE a 305 Good Soldier Component and related sites.____---------------------- 305 Badger Component, (G9 Mae8) so. 90 el ee 308 | SH SCG CE 5s SL i AP, a a Pe RN Ee PSE ED 9 308 1 ey cg lial, el a ot AM ai fad sR dia ly S eng. Wha es pe Oras ented 308 Group. Ays &.(erur he 59. AA er ed Derk PB SE as ee 308 Crolip ibe ee 621k br) aie tine tp aah cee pe Pepys t dete eee) per cere 310 Group Get. eet 3 ha ee aE Ge Ne Bie oN EER he 310 Si Cats SNe ael ON EF ha ROR el eee ren er Seen oes Set eee eee ee 311 eLOPEC ILE DOMMES 15s See ae Sey hte eg eee ee ee 311 EMGISCEA Perse se et re eee Nene wan ene Peete a. eee ae oe 312 GAT hi{ eo DRo Nee, See a oS) SPS ods BONS Phd eR eR Gem t a! | Bese 313 Workedifigkes t.13 0. AW tay! ob Monk Vee Eee eee ce 314 PTOOlematicall ObIGCh. ae ees oe Ske ee ee ee 314 Discussion of the Badger Component... 3224-23 oe eas 314 OVE S PITS aaah iy slap tind pathos albeit, ihe Neto = ale claps 315 Ye) 312) 116 Ub th EDL MMA AI A Sg i DE ly eS RN a UO 316 ieiteralGrewitedieest wet err feet Mes 27 ended te aS 2 eee 317 294 47. 48. 49. 50. 58. ILLUSTRATIONS ILLUSTRATIONS PLATES (All plates follow p. 318) a., View east-northeast of Good Soldier site and surrounding vicinity. Mis- souri River flows eastward in the upper left. eross-sectioning the long axis of the natural mounds. are adjacent to the large pit (Feature 4). Pottery specimens from the Good Soldier Component: a, b, Iona Indented ; c, d, Grey Cloud Horizontal-Incised; e, f, Talking Crow Straight Rim; g, Cadotte Collared ; h, Stanley Tool Impressed ; i-k, miscellaneous. Stone and bone specimens from the Good Soldier Component. Pottery and stone specimens from the Badger Component. Plan and cross section drawings of excavations at the Good Soldier TEXT FIGURE b, View north of trench Broken rocks PAGE THE GOOD SOLDIER SITE (39LM238), BIG BEND RESERVOIR, LYMAN COUNTY, SOUTH DAKOTA By Rosert W. NreumMAn INTRODUCTION In July of 1958 a field party of the Missouri Basin Project, Smith- sonian Institution, spent 14 days conducting archeological excavations at the Good Soldier site (39L.M238), a prehistoric Indian camp site in the Big Bend Reservoir area, South Dakota. The site was first recorded and tested in 1956 by a survey team of the Missouri Basin Project under the direction of Harold A. Huscher (Huscher and McNutt, 1958). The 1958 investigations were supervised by the writer; James J. Stanek acted as field assistant.1 Excavations at this prehistoric site were made possible through Federal funds provided for the Inter-Agency Archeological Salvage Program of the Smith- sonian Institution, the National Park Service and cooperating Fed- eral, State, and local institutions. ACKNOWLEDGMENTS A number of people from the Missouri Basin Project assisted in the compilation of this report, and to the following I am sincerely grate- ful: Warren W. Caldwell, G. Hubert Smith, and Robert L. Stephen- son, who read the original manuscript and offered helpful suggestions; Evelyn B. Stewart, who proofread and handled the illustrations; Ione Wilson, who did the final typing; Jerry Livingston, who did the draft- ing; and Wayne Nelson, who did the photographic work for the plates. Special thanks are due Hobart Eagle, Superintendent of Main- tenance, Bureau of Indian Affairs, at the Crow Creek Indian Reserva- tion. It was through his cooperation that I was loaned a boat, motor, and other water equipment for our daily crossings of the Missouri River from Fort Thompson to the Good Soldier site. To the crew of the 1958 field party I am also deeply indebted, because no task or suggestion was beyond their endurance. 1 Members of the 1958 field party were as follows: Lee Azure, Joseph Benthall, Eugene Brother-of-All, Adolph Burns Day, Michael Forth, Donald Howe, Milo Kearney, Frederick Middle Tent, Horace Slow, and Junior Yellow Back. 295 296 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 LOCATION AND DESCRIPTION The Good Soldier site (391.M238) lies on a low, flat terrace (Mis- souri River Terrace 1, Coogan, 1960) in a narrow valley formed by Good Soldier Creek (formerly Badger Creek) just south and west of the confluence of this tributary and the Missouri River in Lyman County, central South Dakota. The valley is bounded on the east and west by steep, thoroughly dissected breaks that rise abruptly to heights of 30 to 90 feet above the valley floor. The occupational debris at the site was concentrated on a low hummock located about 400 feet south of the right bank of the Missouri River and 150 feet west of the left bank of Good Soldier Creek (the approximate intersection of latitude 44°02’30’’ N. and longitude 99°27’ W.) at an elevation of 1,372 feet m.s.l. (Missouri River Map (1: 24,000) Omaha District, Corps of Engineers, Sheet No. 52, 1949). The hummock measured 130 feet north-south and 65 feet east-west and had a maximum apical height of 5 feet (pl. 47, a). EXCAVATIONS During the 1956 reconnaissance a test pit was dug at the southern end of the rise to a depth of 1.0 foot below the surface. Artifacts recovered from this test include pottery sherds from a single vessel and several hematite-coated bison rib fragments. The restorable portion of the vessel was reconstructed in the laboratory and is iden- tified as of the type Iona Indented (Smith and Grange, 1958, pp. 98-100). At the time of the 1958 investigations a grid of 5-foot squares oriented with the cardinal directions was superimposed over the nat- ural rise and the peripheral area. Each square was given a letter and number designation noting its location in the grid. The squares were scraped horizontally with shovels, and artifacts were placed in bags labeled according to their vertical and horizontal provenience within the grid. All excavations, except eight widely scattered 5- by 5-foot test. pits, were conducted within the northwest quadrant of the grid. Depth measurements were made from the top of the present soil sur- face downward. Excavations were initiated by digging a trench 5 feet wide and 130 feet long across the north-south axis of the mound (pl. 47,5). As work progressed it became apparent that the mound was a natural structure of stream-deposited gravel capped with five distinct soil strata. The stratigraphic sequence in the trench profiles read as follows from top to bottom: Stratum 1, the top soil zone, consisted of a layer of undis- turbed sod and humus about 0.3 foot thick. Only rarely were artifacts present in this zone. The underlying soil zone, stratum 2, was com- Riv. Bas. Sur. = Pap. No. 371 GOOD SOLDIER SITE—-NEUMAN 297 posed of a brownish silt. This deposit ranged from 0.3 to 0.5 foot in thickness and became increasingly thicker along the east and west slopes of the mound. Stratum 2 contained the artifactual material belonging to the latest component at the site. The third deposit, stra- tum 38, was a zone of yellowish, loess-like material generally 0.3 foot thick. This zone was sterile of cultural material except for the in- trusion of postholes from stratum 2. ‘The next soil layer, stratum 4, was composed of a fine, grayish silt, ranging from 1.2 to 1.4 feet in thickness. Stratum 4 contained the archeological specimens belong- ing to the earliest component found at the site. The deepest soil de- posit was stratum 5; it was composed of a thin, yellowish layer iden- tical in appearance to stratum 3. This deepest deposit was about 0.2 foot thick and was present only in the central area of the mound (fig. 58). Soon after beginning excavations at the Good Soldier site it became evident that two distinct archeological deposits were involved. The deposits were separated by the yellow band of silt designated as stra- tum 3. A closer examination, in the laboratory, of the artifact pro- veniences and other archeological data left no doubt that at least two human occupations were represented. In this report the top, and last, occupation is called the Good Soldier Component; the deepest oc- cupation is referred to as the Badger Component. GOOD SOLDIER COMPONENT (89LM238) All of the archeological material belonging to this component was recovered from the two top soil zones, stratum 1 and stratum 2. The most characteristic trait is sherds from globular vessels that have straight to flaring rims and are simple-stamped on their exterior surface. Another trait, somewhat less diagnostic, was the occurrence of numerous posthole remains; 29 post molds were randomly dis- tributed over the central portion of the mound. Of the 29, 13 con- tained only soft, dark earthfill. Bison bone wedges or supports were found in 14 of the postholes; 2 of these contained small rock frag- ments, 2 others contained portions of the original posts, and in an- other a stone projectile point was found. Two postholes contained the remains of unwedged wooden posts. The diameters of the postholes ranged from 0.3 foot to 1.1 feet, the majority being between 0.4 and 1.1 feet. The depths of the postholes ranged from 0.2 to 1.9 feet. Whether or not these postholes are the remains of an earth lodge or lodges cannot be determined from the available data. FEATURES In the field certain peculiarities, such as soil discolorations that marked the locations of pits or concentrations of archeological re- [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 298 ‘OUIS JOIPJOG Pood ay} Iv sUOTIwAOXA JO SZUIMIP UOI}As ssolD pue UL[G—'gso AUN WNLVYLS ASVG GNNOW "VBAvY9o 4331 St OL s te} NOILVAVOXA JO NOILO3SS-SSOYD ANY Id (8EZN16E) 3LIS YAIGIOS GOOD VAAVYES JO SLNNOWY TIVWS ONINIVLNOD 1711S MO11344 VVINSLVW LNSNOdWOD Y3MO7 ‘L1IS A3YD 3NI4 AMYALS "LTS MO113A 3NI4 WINBLVN LNINOdWOD YAddN ‘GOS AG GaddOL 11IS NMOBS AYVG S3univ3s J Q10W1S0d 03903M ANOS GIOWLSOd G3903M 4D0N7ANOB Oo 1SOd N3000M 03903M ANOE A QOOM ONINIVLNOD S1IOHLSOd + GIOWLSOd & Riv. Bas. Sur. Pap. No. 37]. GOOD SOLDIER SITE—NEUMAN 299 mains, were designated as features. Each feature was numbered, carefully described on special “Feature Forms,” and photographed in the order of its discovery regardless of the component with which it was associated. Therefore, the feature numbers assigned to the various components are not in consecutive order. Features 2, 7, 8, and 9 were associated with the Good Soldier Component and are described in detail below. Feature 2.—One posthole and three articulated bison vertebrae. The bone was 2.0 feet north of the posthole. After being cored, the posthole was found to contain a portion of coyote skull. No artifacts were located in direct association with these remains. The top of this feature was in square N140W50 at 0.6 foot below the surface. Feature 7—A concentration of pottery, a fragment of worked bison scapula and several catfish bones. The pottery fragments, 3 rims and 14 body sherds, all belong to a single vessel. This section of the vessel was re- stored in the laboratory and is identified as the type Iona Indented (Smith and Grange, 1958). Feature 7 was located in square N140W35 at 0.8 foot below the surface. Feature 8—Consists of five articulated and two separate bison vertebrae. The remains were located in situ at square N140W30 between 0.3 and 0.5 foot below the surface. Feature 9—A line of stones 5 feet in length, two postholes, and a _ basin- shaped pit. The stone line, oriented almost exactly north-south, con- sisted of 17 hand-sized cobblestones. The two postholes were located symmetrically on opposite sides and about 1 foot from the stones. A line joining the centers of the postholes would be perpendicular to the line of stones and would intersect it at a point 2.0 feet from the south end. The easterly posthole had bison bone wedges while the western one contained the remains of a cedar post. The basin-shaped pit was located less than 1 foot east of the eastern posthole. Fill from the pit was composed of hematite-stained soil and a small amount of burned earth. The pit measured 2.8 feet north-south, 2.0 feet east-west, and had a maximum depth of 0.5 foot. This feature was located in square N145W45 at 1.5 feet below the surface. ARTIFACTS POTTERY The ceramic collection associated with the Good Soldier Component at 39LM238 includes the following pottery types: Iona Indented, Grey Cloud Horizontal-Incised, Talking Crow Straight Rim, Cadotte Col- lared, Stanley Tool Impressed, and a few miscellaneous rims. A total of 241 body sherds was recovered; 83 of these were simple stamped and 158 were plain surfaced. In addition, there were 51 decorated sherds, almost all of which are from the shoulder area of vessels; 39 of these sherds bear parallel incised lines, 2 have parallel trailed lines, 8 are incised with opposed diagonals, 1 has trailed, op- posed diagonals, and 1 bears an incised herringbone motif. These decorative designs are common to all of the above-named pottery types, 661-932-6422 300 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 and I have not attempted to correlate any of the decorated sherds with particular rim-sherd types. IONA INDENTED This pottery has been described in detail and assigned as a compo- nent type of Iona Ware by Smith and Grange in the Spain site (389LM301) report (Smith and Grange, 1958). There are no significant differences between the Iona Indented pottery from the Spain and Good Soldier sites. SAMPLE: Two partially restored vessels (pl. 48, a and b) and four separate rim sherds. METHOD OF MANUFACTURE: Probably built up from lumped clay by the paddle and anvil technique. There is no evidence of coiling. PASTE: Temper: Moderate amounts of tiny grit particles composed of quartz, mica, and feldspar. Texture: Medium coarse to smooth. Color: Buff, grading into gray. Many areas on the interior and exterior surfaces are blackened from use. ForM : Lip: Thickened, T-shaped or resembling an inverted |. Flattened or rounded on top. Lip thickness ranges from9mm.tollmm. At the Spain site most lips were from 9 mm. to 16 mm. thick. Rim: Slightly flaring to straight. Heights range from 36 mm. to 46 mm. Most rims at the Spain site were from 26 mm. to 47 mm. in height. Shoulder: Quite rounded. The rims extend out from the vertical at an angle of about 25 degrees. The angle between the exteriors of the rim and the shoulder ranges from 85 degrees to 110 degrees on the two measur- able specimens. Base: Presumably rounded. Appendages: None recovered at the Good Soldier site; however, four strap handles and one lug are reported for the Spain site. SURFACE FINISH: Interior: Smoothed, otherwise unmodified. Haterior: Simple-stamped, large areas have been smoothed to the extent of almost obliterating the simple-stamped impressions. Decoration: Limited to lip and shoulder areas. Two rims have a con- tinuous series of opposed diagonals on the lip. Four other rims are decorated with a series of diagonal lines interrupted at equal intervals by single, opposing lines. ‘The shoulder of one vessel bears incised opposed diagonal lines. GREY CLOUD HORIZONTAL-INCISED Pottery of this type has been described in detail and assigned the above name by Smith and Grange (1958, pp. 102-103) in the Spain site report. Only three Grey Cloud Horizontal-Incised rims were re- covered at the Good Soldier site. The first rim is grit tempered, buff shading into gray and black on the exterior and light buff on the interior surface; the coreis gray. The rim is slightly flaring and has a flat lip. Decoration consists of a se- Riv. Bas. Sur. Pap. No. 37] GOOD SOLDIER SITE—NEUMAN 301 ries of diagonal punctates on the lip and four horizontally trailed lines on the rim exterior (pl. 48, d). ‘The rim measures 27 mm. in height and the lip is 6 mm. thick. This specimen most closely resembles the type Grey Cloud Horizontal-Incised, Variety A. The second rim is incomplete, it resembles the first rim in all respects except that it is decorated on the lip with a series of tool identations. I have also classified this specimen as Grey Cloud Horizontal-In- cised, Variety A (Smith and Grange, 1958, p. 103). The third rim is grit tempered, buff colored on the exterior and gray on the interior surface; the core is also gray. The rim is slightly flaring and the lip is somewhat rounded and beveled toward the ex- terior. Decoration consists of four horizontally trailed lines on the rim exterior and a series of vertically oriented tool indentations on the lip interior (pl. 48, ¢). This sherd falls within the Grey Cloud Hori- zontal-Incised, Variety C category as described by Smith and Grange. This type is also represented by a small number of sherds from the Talking Crow site (89BF8). TALKING CROW STRAIGHT RIM Talking Crow Straight Rim is a component type of Talking Crow ware (Smith, 1951, pp. 36-37; Smith and Grange, 1958, pp. 101-102). Charles H. McNutt, in a subsequent classification of certain Plains pottery (McNutt, 1959), prefers to assign similar sherds to the Russell Plain Rim type, a component of Russell ware. I have grouped five of the rim sherds from the Good Soldier Component into the Talking Crow Straight Rim category (pl. 48, e, f). The rims are grit tempered. The exterior surfaces are buff, grading into gray, and one sherd is black and orange. Interior surfaces are buff to gray on four of the specimens and black on the fifth. Cores are generally lighter than the darkest side-surfaces. In form the rims are straight and the lips are flat to slightly rounded. Decoration is confined to the lips and consists of a series of diagonal tool indenta- tions on three of the specimens and punctations on the remaining two. Three of the rims are complete enough to measure; they range from 31 mm. to 41 mm. in height. Lips range from 5 mm. to 7 mm. in thickness. CADOTTE COLLARED Pottery of this type has been described in a manuscript prepared by Carlyle Smith. The ceramic classification is based upon specimens excavated at the Two Teeth site (89BF204). Smith (personal com- munication) also includes four collared rims, listed under the Miscel- laneous category in the Spain site report, as Cadotte Collared (Smith and Grange, 1958, p. 104, pl. 31, 6). Only one rim sherd from the Good Soldier Component can be classified within this grouping (pl. 48,9). 302 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The rim is grit tempered; gray shading into black on the exterior and buff shading into black on the interior surface. Only a small portion of the shoulder area remains; however, the collared area and the lip are complete. The collar is straight in form and measures 39mm. in height. The lip is slightly rounded and beveled toward the exterior; it is 6 mm. thick. Decoration consists of a series of verti- cally oriented tool indentations on the lip interior and four rows of punctations on the exterior of the collared area. The only difference between this sherd and Cadotte Collared, Variety D, is that the type description of Variety D makes no mention of lip decoration. STANLEY TOOL IMPRESSED This pottery type was established by Donald J. Lehmer (1954, p. 45) as a component type of Stanley Braced Rim ware. Only one rim fragment from the Good Soldier Component is classified under this category (pl. 48, i). The rim is grit tempered, buff to gray on the exterior and gray on the interior surface. The rim is slightly flaring and has a braced fillet extending down onto the rim exterior. The lip is flattened. Decoration is limited to the braced area and the lip and consists of a series of opposed diagonal impressions. MISCELLANEOUS Included in this category are three rim sherds, one appendage, and one incised body sherd. The first specimen is a grit-tempered rim sherd; it is gray on the exterior and interior surfaces. The rim is straight and the lip is rounded. Decoration consists of two diagonal tool impressions on the lip exterior and three horizontally incised lines on the rim exterior (pl. 48, 2). The second specimen is also a grit-tempered rim sherd. It is buff to gray on the exterior and interior surfaces. The rim is flaring and has a rounded lip. Decoration is limited to the rim interior at the lip in the form of vertical tool indentations. Incised lines, which undoubtedly decorated the shoulder area, extend up to the lower extremity of the rim (pl. 48, 7). The specimen measures 18 mm. in rim height and the lip is 5 mm. thick. This rim resembles pottery classified by W. Raymond Wood as Evans Indented, a component type of Evans ware reported from late prehistoric sites in Holt and Knox Counties, northeast Nebraska (Wood, MS.) The third rim is grit tempered, buff on the exterior and gray on the interior surface. The rim is straight in form and the lip is rounded. An appendage has been broken from the upper portion of the rim. The remaining portion of the projecting appendage is round and sug- gestive of a loop handle or lug. There are two parallel incised lines Riy. Bas. Sur. ee BNO STI GOOD SOLDIER SITE—NEUMAN 303 extending longitudinally on the lip. To one side of the appendage, on the rim exterior, there are two horizontally trailed lines (pl. 48, %). The fourth specimen appears to be a loop handle. It is gray and lacks decoration. The appendage measures 15 mm. in height and is 6 mm. thick. The fifth specimen is a body sherd with an incised motif on the exterior surface. The sherd is grit tempered, buff on the exterior and black on the interior, and measures 6 mm. in thickness. It is quite jagged around the edges and unsymmetrical. Portions of the design have been broken off; however, the remaining portion consists of a circle and two diagonal lines that intersect each other in the approxi- mate center of the circle. STONE PROJECTILE POINTS (5 specimens) Triangular, straight base (one specimen) .—It is a planoconvex in cross sec- tion and displays jagged percussion flake scars on both faces and along all side edges. The point is made from a reddish-brown quartz. It meas- ures 30 mm. long, 22 mm. wide, 7 mm. thick and weighs 4.8 gm. (pl. 49, a). Triangular, concave base (one specimen).—The point displays fine, bifacial pressure flaking. It is made from a brown jasper and measures 20 mm. long, 11 mm. wide, 3 mm. thick and weighs 0.8 gm. (pl. 49, b). Triangular, basal notched.—This point lacks a portion of the tip. There is some doubt as to whether or not the very shallow basal notch was in- tentionally made. The specimen is made from a yellowish-brown quartz- ite. It measures 34 mm. long (estimated), 20 mm. wide, and 6 mm. thick (pl. 49, c). Triangular straight base, side notched (two specimens).—The largest point is made from a gray chert. It measures 19 mm. long, 14 mm. wide, 3 mm. thick, and weighs 0.8 gm. (pl. 49, d). The smaller point is made from a whitish chalcedony. It measures 14 mm. long, 12 mm. wide, 3 mm. thick, and weighs 0.5 gm. (pl 49, e). END SCRAPERS (2 specimens) The first specimen is triangular in outline and planoconvex in cross section. The “keel” has been removed from the convex face, and secondary chipping on that face is limited to the working end and one side edge. A single percussion flake scar extends down the total length of the long axis on the convex face. The scraper is made of Knife River flint and measures 40 mm. long, 28 mm. wide, and 8 mm. thick (pl. 49, #). The second specimen is rectangular in outline and concavoconvex in cross section. The convex face has been retouched along the working end and the two side edges. The planer surface is flat near the working end, but it has a pronounced concavity at the 304 BUREAU OF AMERICAN ETHNOLOGY {Bull. 189 opposite end. The specimen is made of a clear, banded quartz. It measures 27 mm. long, 21 mm. wide, and 5 mm. thick (pl. 49, g). EKNIVES (2 specimens) Specimens in this category are bifacially flaked along one or more edges. Each piece is but a fragment of the original tool. The first fragment is rectangular in outline and has two flat faces. Chipping is limited to one end and one side edge. The chipping along the side edge has resulted in a steep bevel from one face. This speci- men is made from plate chalcedony; it measures 49 mm. long, 30 mm. wide, and 7 mm. thick. The second specimen appears to be the main portion of an elliptical knife. One end and a side have been chipped from only one face. The other side edge has been bifacially chipped, but it has a definite bevel from one face. The knife is made of Bijou Hills quartzite; it meas- ures 63 mm. long, 43 mm. wide, and 8 mm. thick. DRILL (1 specimen) This piece appears to be the butt end of an expanded base drill. The sides and the base are quite straight. Both faces have been flaked along the edges, and in cross section the drill resembles a flattened oval. The specimen is made from tan quartz; it measures 32 mm. long (esti- mated), 12 mm. wide, and 5 mm. thick (pl. 49, 2) MISCELLANEOUS CHIPPED STONE Four specimens, generally ovoid in outline, bear a small amount of retouching along one or more edges. These may have been modi- fied intentionally for immediate use and then discarded. Two of the flakes are made of quartz, one of quartzite, and the other of gypsum. They range from 21 mm. to 41 mm. long, 19 mm. to 27 mm. wide, and 6 mm. to 15 mm. thick. GROUND STONE ARTIFACTS SHAFT SMOOTHERS (1 specimen) This piece consists of a fragment of scoria. It is slightly triangular in outline and planoconvex in cross section. There is a shallow U- shaped groove extending longitudinally down the entire length of the planer surface. Toward one end, the groove becomes quite shallow and narrow. Riv. Bas. Sur. Soe a te ee GOOD SOLDIER SITE—NEUMAN 305 ABRADERS (3 specimens) The first specimen somewhat resembles a discoidal. One of the faces has been smoothed from use, the opposite face is rough and un- altered. It is made from a piece of scoria and measures 26 mm. thick and 56 mm. in diameter. The second specimen is shaped like a parallelogram. The two faces are flat and have rounded side edges. The sides vary from wedge shaped to rounded to flat. The specimen is made of gypsum. The top and bottom measure 50 mm. and 48 mm. long respectively ; it is 29 mm. in width and 15 mm. thick. The third specimen is a rectangular piece of sandstone that ap- pears to have been artificially shaped. Both faces and one long side are quite smooth. One face has a shallow, ovoid concavity and the side surface of one end bears pecking scars. The tool measures 98 mm. long, 57 mm. wide, and 27 mm. thick. BONE ARTIFACTS (4 specimens) Scapula fragments (three specimens).—Each is a fragment of bison scapu- lae. The first fragment is from a right scapula which has had the spine removed. The proximal and distal portions are lacking, as is the posterior border. The remaining distal portion of the anterior border has been smoothed from wear. The specimen measures 231 mm. in length. The second specimen is also a fragment of a right scapula. It is a rec- tangular piece lacking the proximal and distal portions and the posterior border. The proximal end of the fragment, as well as the lateral sur- face and the anterior border, has been smoothed from wear. This speci- men is 174 mm. long (pl. 49,7). The third fragment is a thin, elongated piece of scapula that converges to a rounded point at one end; the opposite end is fractured. Both side edges along the pointed half of the bone are sharp and smooth from wear. Each face of the bone is marked with numerous longitudinal scratches, most of which are concentrated toward the pointed end. The specimen is 182 mm. long (pl. 49, 7). Cut bone (one specimen).—A portion of the proximal end of a bison or elk right metacarpal. The bone has been cut diagonally down from the articular facets to a point along the vascular groove. No other area on the bone appears to be artificially altered. The specimen is 39 mm. long. GOOD SOLDIER COMPONENT AND RELATED SITES In the following discussion I have attempted to describe the Good Soldier Component more fully by relating its remains to similar material from other sites. This method, I hope, will afford the reader 306 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 a meaningful understanding of the people represented by these arche- ological data. The excavated remains that have been assigned to the Good Soldier Component at site 39L.M238 were recovered from the two top strata of anatural mound. Features, of which there were four, consisted of concentrations of worked and/or unworked mammal bone, stone arti- facts, and pottery fragments. One rather interesting feature was a line of cobblestones associated with postholes and a firepit. The inventory of pottery, stone, and bone was indeed meager; of these specimens, only pottery and specifically vessel rims, are diagnostic enough to show relationships between the Good Soldier Component and certain other sites. The ceramic collection from the Good Soldier Component includes vessel rims assigned to five pottery types. Each of these types is characterized by globular vessels having constricted necks. The rims are straight, slightly flaring or collared. The trait of simple- stamping the exterior surface of the vessels is common to each of the types. This trait is generally associated with the “farming-hunting villages” of the late prehistoric and historic peoples who lived on the northern and central Great Plains. The earliest date on a simple- stamped pottery occupation in these regions is from the Thomas Riggs site (39HU1), a rectagular house village in Hughes County, S. Dak. A charcoal specimen from this site provided a carbon-14 date of A.D. 1228+-200 years (Missouri Basin Chronology Program, Statement No. 2, Missouri Basin Project, June 15, 1959, Lincoln). A site bearing close artifactual relationships to the Good Soldier Component is the Spain site (391LM301). This small, compact village is located about 1 mile upstream from the mouth of Bull Creek in Lyman County, S. Dak. The creek flows eastward into the Missouri River about 33 miles downstream from the Good Soldier site. Excavations at the Spain site were conducted in 1953 by a field party of the University of Kansas under the direction of Carlyle S. Smith. The principal occupation at the Spain site has been designated Com- ponent A and is characterized by a small, prehistoric village situated in the tree-sheltered bottoms of a minor stream course (Smith and Grange, 1958). The remains of two and probably four earth lodges were located on low hummocks or rises. The ecological situation closely duplicates what was found at the Good Soldier site, where the occupational remains were concentrated on a low, natural rise in a narrow stream valley. The one completely excavated house at Spain had a basin-shaped floor 29 feet in diameter and an extended entryway 7 feet long and 4 feet wide. The entryway faced southeastward toward Bull Creek. Features within the house included a central hearth, a secondary Riv. Bas. Sur. Pap. No. 371 GOOD SOLDIER SITE—NEUMAN 307 firepit, and two small pocket caches. A bell-shaped cache pit and a midden deposit were located along the outer edge of the house. Abundant remains of charred beams indicated that the structure had been destroyed by fire. TABLE 1.—Oollection of identifiable rim sherds from Spain site (Component A) Pottery type Total Percent fons Modan fede is. lS seek ial a ye tae el Eo 328 26. 5 lone .Horzonunalincised <2 205522 26S ee 311 25. 2 Jona Myagangl-Meised_o. 2b ie es kL 82 6. 6 IigeayS-Rim, Vertepy Au... F 2. - Se iT (Mi 6. 2 Tonacsoiwnn VY arleLy is see ee ee 50 4.1 kona Shine Watiety Ol ioe) Pe ee eet 14 Theat Tons (Shun ay anieny (Ds x.) 4- Soe es hers 8 ay? Palkine Grow oiraipbt Rim... oe oll 25.2 Grey Cloud: Horizontal-Incised _i-_-_.-_-=-_2-___. 54 4.4 There are also 320 rims, representing 20.5 percent of the total collec- tion, that are listed under the heading of Miscellaneous. As mentioned previously, four of these were later classified as Cadotte Collared. It is apparent that the ceramic inventories and their propor- tional representation, at the Spain site (Component A) and the Good Soldier Component at 39L.M238, are very similar, indicating a close archeological relationship. The presence of one Stanley Tool Im- pressed rim at Good Soldier and the absence of such a type at Spain may be reason to suspect that the former site was occupied ata slightly later date. Component A at the Spain site has been assigned to the Shannon Focus of the Chouteau Aspect; a postulated date for the component lies between A.D. 1550 and 1650. (Smith and Grange, 1958). Another component of the Shannon Focus is the Two Teeth site; excavated in 1955 by a crew of the University of Kansas under the supervision of Carlyle S. Smith. This village is located across the Missouri River and about 4 miles upstream from the Good Soldier site. Excavations here were conducted into the remains of two circular earth lodges and several midden deposits. Smith (1959) states that the same pottery types found at Spain plus “Cadotte Collared, a small amount of Stanley Braced Rim Ware, and one stray Campbell Creek Cord Marked sherd” were present at Two Teeth. It would seem that the pottery inventories from Two Teeth and Good Soldier are almost identical and that each of the components may be contemporaneous occupations. On this basis the Good Soldier Com- ponent fits into the Shannon Focus of the Chouteau Aspect and dates no later than A.D. 1600. 308 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 BADGER COMPONENT (89LM238) FEATURES On the basis of their location within stratum 4, Features 1, 3, 4, 5, and 6 were assigned to the Badger Component of the Good Soldier site. The features and their associations are described below. | Feature 1—A small pile of 31 unworked stone chips, some of which fit together. The concentration, 0.3 foot by 0.4 foot, measured 0.3 foot from top to bottom. The chips range from 38 mm. to 53 mm. long, 26 mm. to 35 mm. wide, and 8 mm. to 14 mm. thick. The cache was located at a depth of 1.2 feet in square N125W50. Feature 3—A small, basin-shaped firepit containing charred and uncharred bone fragments and a small quantity of charcoal. The pit was 2.6 feet in diameter and had a maximum depth of 0.4 foot. The top of the pit was defined at 2.2 feet below the present ground surface in square N1380W50. Feature 4—A large, basin-shaped firepit containing charred and uncharred bone fragments, charcoal, and 225 fire-cracked, handsized, granitic rocks. Most of the rocks were in the bottom two-thirds of the pit. The feature measured 3.3 feet by 4.6 feet and had a maximum depth of 14 feet. A definite outline of the pit was apparent at 2.0 feet below the surface; however, indications of the pit were noticeable at a depth of 1.5 feet in square N125W50. A charcoal specimen from Feature 4 was submitted to the University of Michigan carbon—14 laboratory; it was cataloged M-1090a and dated 419 + 150 years B.C. Feature 5—Two shallow, basin-shaped firepits. The fill in the northern- most pit was stained red with hematite. It measured 1.9 feet in diameter and had a maximum depth of 0.19 foot. The second pit, located about 0.3 foot south of the first, measured 0.9 foot in diameter and 0.4 foot in depth. Fill from this pit contained charcoal, flint chips, and fragments of bone. The pits became apparent at 1.2 feet below the present ground surface in square N155W50. Feature 6.—A shallow, basin-shaped firepit containing burned earth, small flecks of hematite, and fragments of bone, stone, and shell. The pit, 1.38 feet in diameter and 0.4 foot in maximum depth, was located at 1.6 feet below the surface in square N180W50. POTTERY The pottery collections from stratum 4 at the Good Soldier site consist of sherds, most of which have been cord-paddled on their exterior surfaces. At least two vessel shapes are represented, one being globular with a constricted neck and a straight, everted rim. The other is an almost shoulderless vessel with a conoidal base. The shapes and decorations on the rims are such that they may be sorted into three groups. GROUP A This category consists of two undecorated rims, apparently from the same vessel (pl. 50, and c). They measure 24 mm. in height and are 7 mm. thick, becoming slightly thinner in the area immediately Riy. Bas. Sur. Pap. No. 87]. GOOD SOLDIER SITE—NEUMAN 309 below the lip. The flattened lip slopes slighty downward toward the vessel exterior and is 8 mm. wide. The interior suface is smooth, almost polished in appearance, and dark gray to black. The exterior, tan to dark brown, has faint, vertical, cord-paddled impressions ex- tending a short distance up the rim. Both specimens are abundantly tempered with fine to medium-sized grit particles. The largest rim was located at a depth of 1.6 feet in square N145W15. ‘The other rim came from a depth of 2.1 feet in square N145W20. These fit the descriptions of the plain Great Oasis pottery reported by Lloyd A. Wilford (1945 pp. 35-86; 1955 p. 188). Elden Johnson, of the Uni- versity of Minnesota, examined the sherds and stated “The flat-lipped rims... are identical to Wilford’s Great Oasis undecorated rims” (Personal communication March 24, 1961). Other Great Oasis sherds have been collected from at least four sites in South Dakota; the Oldham site (839CH7) in Charles Mix County (Cooper, 1955, p. 60), and the Gavins Point site (89 YK203) in Yankton County (Hall, 1961). In November 1960, this writer was allowed to examine the private collection of Mr. Francis Deuder of Ree Heights, Hand County. The collection included a number of decorated Great Oasis rims from the Ree Heights Buffalo Kill site (89HD3). Finally, sherds bearing some resemblance to Great Oasis pottery are men- tioned in the Swanson report (Hurt, 1951, p. 88). The comparison is made with the type Chamberlain Incised Triangle; Hurt notes that “. . . none of the Chamberlain Incised Triangle found at the Swanson site has a cord-marked body like Great Oasis pottery.” This should make little difference in the comparison, since a fair percentage of Great Oasis wares do not have cord-paddled exterior surfaces (Wil- ford, 1945, p. 36). It is worth mentioning here that a wooden post, excavated from a rectangular house at the Swanson site was analyzed at the University of Michigan carbon-14 laboratory; it was cataloged M-839 and dated A.D. 858-250 years. In a 1949 publication Wilford refers to a vessel and some sherds “. . which are clearly of the Great Oasis type” and speculates that they may represent prehistoric Omaha pottery because they are from the Ryan site (25DK2) which is thought to be a burial place for the Omaha “Large Village.” The vessel and its resemblance to Great Oasis is mentioned again in Wilford’s 1955 report. John L. Champe, of the Laboratory of Anthropology, University of Nebraska, permitted an examination of this vessel and another from the same excavation. The specimens excavated in 1939 are from a natural mound designated 25DK2A near the town of Homer in Dakota County, Nebr. Stanley Bartos, Jr., supervised the excavations and his notes, on file at the University of Nebraska, state that the mound measured 125 feet north-south, 20 to 50 feet east-west, and had a height of 2.5 feet (also see Champe, 1946, pp. 117-118). It contained prehistoric secondary 310 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 burials, and at a later date primary interments ascribed to the Omaha, were intruded into it. GROUP B These are five undecorated rim sherds, four of which fit together. All five apparently belong to a single vessel (pl. 50, d). The rim section is about 9 mm. thick and has a rounded lip. The interior surface is black and smooth. The exterior, ranging from tan to dark gray, is covered with large (3 per centimeter), parallel, cord- paddled impressions extending diagonally downward from the lip. The sherds are tempered with minute particles of calcite and occasional large and small grit inclusions. They were recovered from the 2.0-2.5 foot level in square N135W40. The specimens bear a general re- semblance to several varieties of Plains Woodland vessels that are almost shoulderless and have conoidal bases. Calcite tempering is a characteristic of Harlan Cord-Roughened, the diagnostic pottery of the Keith Focus (Kivett, 1953). Two sites within this focus have been dated by radiocarbon methods. The Woodruff Ossuary in north- central Kansas (Kivett, 1953) was dated at A.D. 6114240 years (Wedel and Kivett, 1956), and site 25FT18 in south-central Nebraska was dated at A.D. 828200 years (Crane and Griffin, 1960, p. 40). The Valley Cord-Roughened is characteristic of the Valley Focus (Hill and Kivett, 1941), the type site being 25VY1 in central Nebraska (Kivett, 1949). A dendrochronological estimate of A.D. 1000 to 1150 was assigned to a Valley Focus occupation at Ash Hollow Cave in southwest Nebraska (Champe, 1946, p. 86). The trait that allies the Group B specimens from the Good Soldier site to Valley Cord- Roughened is the diagonal orientation of the cord impressions on the vessel exterior. GROUP C Only one large rim and shoulder section of a vessel is represented here (pl. 50, a). The rim is vertical with a very faint outward bulge between the neck and the lip. A very weak shoulder is evident below the neck. The lip is flat and slants downward toward the vessel exterior. Rim height, measured between the neck and lip top, is 34 mm. and the thickness is 6 mm. The specimen is abundantly tem- pered with grit particles ranging up to 2.0 mm. in size. The interior surface is a brownish-orange and smooth. The exterior is the same color, but it is covered with extremely fine, parallel, cord-paddled impressions that extend vertically down from the lip. The rim is decorated with parallel, notched-stick indentations that extend from the lip top diagonally downward onto the rim exterior. The indenta- tions are about 13 mm. in length and V-shaped in cross section. The notches run perpendicular to, and average about six per indentation. The rim was recovered at a depth of 2.0 feet in square N145W50. Al- Riv. Bas. Sur. Pap. No. 37] GOOD SOLDIER SITE—NEUMAN 311 though this rim does not resemble any of the reported pottery types on the northern and central Plains, I was able to find two similar specimens that belong to a single vessel. They are from site 39BR11, located on the opposite side of the Missouri River and about 21 miles downstream from the Good Soldier site. Field parties of the Smith- sonian Institution located and made tests at the site in 1947, 1953, and 1954. Paul L. Cooper’s notes (on file at the Lincoln office of the Smithsonian Institution) show that an occupation area at 39BR11 was overlain by two distinct strata. The uppermost was the plow zone or present surface, under which was a sterile zone of fine, yellowish silt. Beneath the silt was a “dark brown” layer of soil containing small, basin-shaped firepits and artifacts. One of the rims (cat. No. 39BR11-18) was recovered from this stratum at “Profile 6... 2.6’ to 3.3’” below the surface; the other rim (cat. No. 89BR11-87) came from dirt that had sloped down onto the riverbank. There are also certain artifactual similarities between the Badger Component and the Clear Lake site, which is located along the Illi- nois River in Tazewell and Mason Counties, Ill. (Fowler, 1952). I believe the rim from Group C (Badger Component) resembles those from the conoidal vessels of Weaver ware, the predominant pottery type at Clear Lake. The projectile points from Group 2 (Badger Component) also bear a resemblance to those of Group A that are illustrated on page 155 of the Clear Lake Report. Melvin L. Fowler kindly examined specimens from the Badger Component and his statements, regarding the Group C rim and the Group 2 points, are as follows: “. . . I would not compare it closely with Weaver. The general resemblances are there in the rather tightly twisted parallel cords and general vessel shape. The paste is different and not as com- pact as Weaver. . . . In general, if such a sherd were found in IIli- nois, it would be classified as Late Woodland on a general Weaver Horizon. .. .” (Fowler, personal communication February 2, 1960). “The projectile points are not Weaver but are similar to those asso- ciated with the so-called Maples Mills and Dillenger cultures in TIli- nois....” Weaver ware is generally assigned to the Woodland period in the Illinois Valley and the ware has been dated by carbon-14 from the Irving Village at A.D. 770250 years, and from the Ruther- ford Mound at A.D. 425-£200 years (Griffin, 1958, pp. 12-18 and 15). STONE PROJECTILE POINTS (6 specimens) Group 1. Triangular, straight base (two specimens).—Each has slightly convex sides and shows haphazard bifacial flake scars. The ore BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 largest specimen lacks the tip and is made from a grayish-brown quartzite (pl. 50, 7). It measures 20 mm. long (estimated), 18 mm. wide, and 5mm. thick. The other point (pl. 50, e), made from a erys- tal clear quartz, is 15 mm. long, 17 mm. wide, 4 mm. thick and weighs 1.8 gm. Both points were recovered from the 1.0-1.5 foot level in square N145W45. Group 2. Triangular, corner-notched (four specimens).—None are complete, but all show straight to slightly convex sides. Evenly placed, bifacial flake scars extend outward at right angles from the longitudinal axis to the side edges. One point, broken almost its en- tire length down the long axis, has a slightly convex base (pl. 50, ¢). It is made from a brown chalcedony, but one face is completely coated with a white patination. The specimen measures 38 mm. long, 27 mm. wide (estimated), and 5 mm. thick; it was located at a depth of 1.3 feet in square N125W70. The next piece also has a convex base, but lacks the tip portion (pl. 50, 2). It was made from a very fine- grained, light-brown quartzite and measures 32 mm. long (estimated), 24 mm. wide and 4 mm. thick. The specimen was recovered 2.2 feet below the surface in square N175W50; it is similar to a point (Cata- logue No. 43) found at site 39BR11, Brule County, S. Dak. The third specimen, made of basalt, lacks the base, and the tangs were acci- dentally broken off in the laboratory (pl. 50,2). It was located at a depth of 2.3 feet in square N190W40. The fourth piece, made from a fine-grained, olive-drab quartzite, consists only of a point section (pl. 50,7). It was recovered from a depth of 1.8 feet in square N150W35. END SCRAPERS (10 specimens) Group 1 (four specimens).—AlIl are planoconvex in cross section, ovoid in outline, show little or no secondary flaking on the convex surface, and have been retouched at one end and both side edges. The first specimen (pl. 50, &), made from a gray chalcedony, is 40 mm. long, 28 mm. wide and 7 mm. thick. It was located at a depth of 2.3 feet in square N155W15. The second scraper (pl. 50, 7), made from a brown chalcedony, is patinated on the convex surface. It measures 35 mm. long, 29 mm. wide, 10 mm. thick and was recovered from the 1.5-2.0 foot level in square N155W25. The third piece (pl. 50, m), also made from a brown chalcedony, is patinated on the base and along the side edges. It measures 26 mm. long, 12 mm. wide, 7 mm. thick and was found in the 1.0-1.5 foot level at square N185W50. The fourth specimen has been fractured opposite the working end; it was made from a brown chalcedony and is heavily patinated on the convex surface (pl. 50, 7). The fragment measures 22 mm. wide, 9 mm. thick and was located in the 1.0-1.5 foot level at square N145W40. Riv. Bas. Sur. Pap. No. 37], GOOD SOLDIER sITE—NEUMAN 313 Group 2 (three specimens).—All are planoconvex in cross section, rectangular in outline and have a keel toward one side of the convex surface extending down the long axis for almost the entire length of the scraper. The first specimen, made from a brown chalcedony and patinated on one side, has a small semicircular notch chipped into one side edge (pl. 50, 0). The piece measures 18 mm. long, 18 mm. wide, and 7 mm. thick. It was recovered from the 2.0-2.5 foot level at square N150W40. The next specimen, made from a mottled gray flint, is patinated along portions of one end and a side edge and has secondary chipping along the opposite side edge (pl. 50, p). It is 20 mm. long, 18 mm. wide, 5 mm. thick and was located in the 2.0-2.5 foot level at square N145W35. The third specimen is incomplete; it is made from a brown chalcedony and shows retouching only at the working end (pl. 50, g). The fragment was recovered from the 1.5- 2.0 foot level in square N160W50. Group 3 (three specimens) —These pieces are planoconvex in cross section, triangular in outline and have been retouched over the entire convex surface. The largest scraper, made from a brown chalcedony (pl. 50, 7), is 30 mm. long, 23 mm. wide, and 6 mm. thick. It was lo- cated in the 1.0-1.5 foot level at square N135W35. The second scraper, also a brown chalcedony (pl. 50, s), is 19 mm. long, 21 mm. wide, and 6 mm. thick. It was located in the 1.0-1.5 foot level at square N135W35. The last specimen, made from a yellowish-brown jasper (pl. 50, ¢) is 20 mm, long, 19 mm. wide, 6 mm. thick and was recovered from the 1.5-2.0 foot level at square N140W50. KNIVES (7 specimens) Group 1 (three specimens).—Ovoid in outline, each of these pieces displays large, bifacial, percussion flake scars. The largest knife has been finely retouched on one face almost entirely around the outer edge. It is made from a brown chalcedony and has a thin, whitish patination on one face (pl. 50, w). The knife is 87 mm. long, 65 mm. wide, 13 mm. thick and was recovered from 1.8 feet below the surface at square N150W30. The second specimen is incomplete, but the recovered portion lacks any secondary flaking (pl. 50, v). It is made from a fine-grained, mottled brown and tan quartzite and measures 46 mm. long (estimated), 35 mm. wide, and 10 mm. thick. The piece was located in the 1.0-1.5 foot level at square N155W55. The third specimen, also incomplete, is made from a fine-grained, yellowish quartzite (pl. 50, w). It measures 47 mm. wide, 9 mm. thick and was recovered from the 1.5-2.0 foot level at square N140W50. Group 2 (two specimens).—Triangular in outline, each shows hap- hazard, bifacial chipping along the two side edges. The first piece, made from a fine-grained, purple quartzite, measures 82 mm, long, 314 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 28 mm. wide, and 12 mm. thick (pl. 50, 2). It was found in the 1.0- 1.5 foot level at square N145W45. The second specimen is a knife fragment. It consists of a fine grained, grayish quartzite that is patinated along portions of one face and a side edge (pl. 50, y). The piece is 20 mm. long, 34 mm. wide, and 6 mm. thick. It was located at a depth of 1.3 feet in square N150W50. Group 3 (two specimens).—Both pieces are steeply beveled along the side edges, rectangular in outline, and are made of Bijou Hills quartzite. The larger specimen is incomplete and appears to be the upper portion of a knife, or possibly a drill with opposite faces alter- nately beveled. It was located in the 1.0-1.5 foot level at square N145W30. The smaller piece is the midsection of a bifacially flaked blade that has been steeply beveled along one side edge from both faces. It was located in the 1.0-1.5 foot level at square N145W45. WORKED FLAKES (14 specimens) These pieces consist of asymmetrical flakes showing unifacial re- touch along one or more side edges. A few are no doubt fragments broken from side and/or end scrapers. One specimen has a semi- circular notch chipped into each of its two side edges. It has been suggested that this type of tool may have been used for cutting arrow shafts (Cosner, 1956). All of these flakes were recovered from within 1.0 to 2.5 feet depths. PROBLEMATICAL OBJECT (1 specimen) This specimen is tubular and broken at one end. It may be a section of a bead, or simply a limestone concretion (pl. 50, 2). The piece measures 22 mm. long, has an outside diameter of 14 mm. and an inside diameter of 6 mm. It was located in the 1.5-2.0 foot level at square N190W50. DISCUSSION OF THE BADGER COMPONENT In describing the pottery and certain projectile point specimens from this component I have referred to other sites from which similar artifacts were collected. It is evident that the remains of at least three types of pottery vessels are represented in the relatively thin soil zone of stratum 4; evidence as to whether or not the makers of these vessels occupied the site contemporaneously or at different times is inconclusive. A comparable archeological situation involv- ing the kinds of specimens assigned to the Badger Component has not been reported. Nevertheless, in light of the carbon-14 dates related to pottery specimens similar to those in Groups A, B, and C, Riv. Bas. Sur. Pap. No. 37] GOOD SOLDIER SITE—NEUMAN 315 I do not think that the possibility of a simultaneous occupation for the component should be ruled out. Two discrepancies remain to be discussed. The first concerns the question of whether or not Great Oasis pottery is prehistoric Omaha in origin. Statements alluding to the possibility of this origin have appeared in print at least twice (Champe, 1946, p. 48; Wilford, 1949, p- 86), and each reference concerns the vessel or vessels recovered from the excavations at 25DK2A mentioned earlier in this report. I have examined the field notes and sketches from the site and am of the opinion that the vessels, both of which are good Great Oasis speci- mens, belong to the prehistoric burial remains rather than to the intrusive Omaha interments that were accompanied by European trade goods. I am not aware of any Great Oasis ceramics found in undisturbed, direct association with White trade material. The trait of cord-paddling that appears on the exterior surface of some Great Oasis vessels is not a late prehistoric technique in the central Plains or Middle Missouri region, nor is this pottery type very similar to others that are diagnostic of the protohistoric or historic periods in those areas. Great Oasis sherds have been more aptly compared to pottery from the Middle Ceramic Period (Champe, 1946), and trac- ing this pottery type up through time to Omaha occupations has yet to be validly demonstrated. The second problem is in connection with the carbon-14 date ob- tained from charcoal at Feature 4. This date, 419 B.C.+150 years, is certainly not compatible with the artifactual material of the compo- nent; however, it may be a perfectly acceptable date for the basin- shaped pit in Feature 4. As other sites are dug, in locales where stratigraphy is not so subject to stream erosion and deposition, arti- facts may be found that show significant relationships to some of those in the Badger Component, thus validating an earlier occupation not recognized now. In conclusion, the Badger Component represents an early occupa- tion site, the artifacts of which are most closely comparable to those from other sites assigned to the Plains Woodland Period, and a date no later than A.D. 1200 is suggested for the occupation of this com- ponent. CONCLUSIONS The Good Soldier site (839LL.M238), located along the Missouri River in central South Dakota, consists of two stratigraphically separated occupations. The latest has been named the Good Soldier Component. It comprises remains closely resembling those from two other sites in the vicinity and has been assigned, together with those sites, to the Shannon Focus of the Chouteau Aspect. The early occupation at 661-932-6428 316 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 39L.M238 is designated the Badger Component. Its artifactual re- semblances to certain other sites in Minnesota, Nebraska, and South Dakota indicate that it falls within the Plains Woodland Phase. The possibility of a still earlier occupation at the Good Soldier site is suggested. APPENDIX BONE AND VEGETAL IDENTIFICATION FROM THE GOOD SOLDIER SITE? BONE The small quantity of bone from the Good Soldier site was more or less evenly distributed throughout the arbitrary half-foot levels of the excavations. Buffalo (Bison bison) remains were by far the most common, while the whitetailed deer (Odocoileus speleus), coyote (Canis latrans), kit fox (Vulpes velox), prairie dog (Cynomys ludo- vicianus), and the cottontail rabbit (Syvilagus floridanus) are also present. Included in the inventory are a few fish bones; however, there was no evidence of fowl] in the excavations. I have assigned all of the osseous remains recovered from the sur- face, down to 1 foot in depth to the Good Soldier Component and all below that depth to the Badger Component. The results are shown in table 2. TABLE 2.—Osseous remains from Good Soldier and Badger Components GOOD SOLDIER COMPONENT Bison Num-}| White-tailed |Num- Coyote Num. Kit fox |Num- Catfish — ber deer ber ber er er Lower jaw -- 1 || Upper tooth. 3 || Muzzle... 1 || Humerus.- 1 || Dorsal Occiput=-==" 2 || Lower jaw-.-- 1 || Lower jaw--- 1 vertebra - - 1 Scapula_-__.__ 1 || Meta- Atlas os = 1 Pectoral umerus..--- 1 carpal____- 11) | i. 9 Ce 1 girdle_--__- 1 Radius=..=--— 3 || Phalanges_-_ 1 || Humerus_.-_ 1 Fin parts_.._ 4 Pibiges <<< 3 3 || Cuboid----- 1 || Caleaneum_- 1 Metacarpal - - 2 Metatarsal___ 1 Phalanges-_-__ ll BADGER COMPONENT Bison Num-|| Prairie dog |Num-|} Cottontail |Num- ber ber ber Lower jaw--- ‘Reeth=- 335232 Lower jaw..- 1 ||} Humerus-.-- 1 Astragalus___ Cuboid_-_____ 2 4 1 1 Metacarpal _ - 3 F 1 3 1 Phalanges.._- 4 _ 2 The fish bone was identified by David H. Dunkle, United States National Museum. All other bone was analyzed by Theodore E. White, Dinosaur National Monument. Norton H. Nickerson, Washington University, identified the vegetal specimen. Riv. Bas. Sur. Pap. No. 37] GOOD SOLDIER SITE—NEUMAN 317 The lists show buffalo, deer, coyote, kit fox, and fish associated with the Good Soldier Component, and buffalo, prairie dog, and rabbit with the Badger Component. Interestingly, almost all of the buffalo re- mains are from the limbs of the animals; previous writers have sug- gested that this phenomena indicates that the kills and the butchering took place some distance from the main camp or village (Lehmer, 1952; White, 1952 a, 1952 b, 1954). VEGETAL Only one vegetal specimen, a pit of the Prunus americana, was re- covered. It came from square N140W40 in the 1.0-1.5 foot level. LITERATURE CITED Barros, STANLEY, JR. Field notes. Manuscript on file at the Laboratory of Anthropology Univ. Nebraska, 1939, Lincoln. CHAMPE, JOHN L. 1946. Ash Hollow Cave. Univ. Nebraska Stud., n.s. No.1. Lincoln. Coogan, ALAN H. 1960. Geological age of Soldier Creek, Buffalo County, South Dakota. Proc. Iowa Acad. Sci., vol. 67, pp. 314-325. Cooper, Pau L. 1955. The archeological and paleontological salvage program in the Mis- souri Basin, 1950-1951. Smithsonian Misc. Coll., vol. 126, No. 2. 99 pp. Cosner, AARON J. 1956. The “Stone Scraper” and Arrow “Wrench.” Amer. Antiq., vol. 21, No. 3, pp.. 300-301. CRANE, H. R., and GRIFFIN, JAMES B. 1960. University of Michigan radiocarbon dates V. Amer. Journ. Sci., Radio- carbon Suppl., vol. 2, p.40. New Haven. Fow ter, MEtyin L. 1952. The Clear Lake site: Hopewellian occupation. Hopewellian Com- munities in Illinois, edited by Thorne Deuel. Illinois State Mus., Sci. Pap., vol. 5, No.4. Springfield. GRIFFIN, JAMES B. 1958. The chronological position of the Hopewell Culture in the Eastern United States. Univ. Michigan, Anthrop. Pap., No. 12. Hatt, Rosert L. 1961. An archaeological investigation in the Gavin’s Point area, Yankton County, South Dakota. Museum News, W. H. Over Museum, vol. 22, No. 7%. Hit, A. T., and K1vett, MARvIN F. 1941. Woodland-like manifestations in Nebraska. Nebraska Hist., vol. 21, No. 3. Hort, WESLEY R., JR. 1951. Report of the investigation of the Swanson site 39BR16, Brule County, South Dakota, 1950. South Dakota Archaeol. Comm., Archaeol. Stud., Circ. No. 3. 318 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 HuscHeEr, HArotp A., and McNutt, CHARLES H. 1958. Appraisal of the archeological resources of the Big Bend Reservoir, South Dakota. Mimeographed report prepared by the Missouri Basin Project, Smithsonian Institution. Lincoln. KIvetT, MARVIN F.. 1949. A Woodland pottery type from Nebraska. Proc. Fifth Plains Conf. Archeol. Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 67-69. 1953. The Woodruff Ossuary, a prehistoric burial site in Phillips County, Kans. Bur. Amer. Ethnol., Bul 154, Riv. Bas. Surv. Pap., No. 3, pp. 103-142. LEHMER, DONALD J. 1952. Animal bone and Plains archeology. Plains Archeol. Conf. News Let- ter, vol. 4, No. 4. 1954. Archeological investigations in the Oahe Dam area, South Dakota, 1950-51. Bur. Amer. Ethnol., Bull. 158, Riv. Bas. Surv. Pap., No. 7. McNutt, CHARLES H. 1959. Comments on two Northern Plains pottery wares. Abst., Proc. Sixty- Ninth Ann. Meet. Nebraska Acad. Sci. SMITH, CARLYLE §, 1951. Pottery types from the Talking Crow site, Fort Randall Reservoir, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 3, pp. 32-41. 1959. The temporal relationships of coalescent village sites in Fort Randall Reservoir, South Dakota. Actas del xxx11I. Congreso Internacional de Americanistas, tomo II, pp. 11-123. San Jose, Costa Rica. SMITH, CARLYLE S., and GRANGE, ROGER T., JR. 1958. The Spain site (89LM301), a winter village in Fort Randall Reser- voir, South Dakota. Bur. Amer. Ethnol., Bull. 169, Riv. Bas. Surv. Pap., No. 11. WEDEL, WALDO R., and KIvETT, MARVIN F. 1956. Additional data on the Woodruff Ossuary, Kansas. Amer. Antiq., vol. 21, No. 4, pp. 414-416. WHITE, THEODORE FE. 1952 a. Observations on the butchering technique of some aboriginal peoples: I. Amer. Antiq., vol. 17, No. 4, pp. 337-338. 1952 b. Suggestions on the butchering technique of the inhabitants at the Dodd and Phillips Ranch sites in the Oahe Reservoir Area. Plains Archeol. Conf. News Letter, vol. 5, No. 2. 1954. Butchering techniques at the Dodd and Phillips Ranch sites. Bur. Amer. Ethnol., Bull. 158, Riv. Bas. Surv. Pap. No. 7, Appen. 4, pp. 165-172. WILFORD, LLoyp A. 1945. Three villages of the Mississippi pattern in Minnesota. Amer. Antiq., vol. 11, No. 1, pp. 32-40. 1949. Archeological field work in Minnesota. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 34-36. 1955. A revised classification of the prehistoric cultures of Minnesota. Amer. Antigq., vol. 21, No. 2, pp. 130-142. Woop, W. RAYMOND. The Redbird Focus. MS., MA thesis, on file at the Lab. Anthrop., Univ. Nebraska, 1956. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 47 a, View east-northeast of Good Soldier site and surrounding vicinity. Missouri River flows eastward (upper left). b, View north of trench cross-sectioning the long axis of the natural mound. Broken rocks are adjacent to the large pit (Feature 4). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 48 Pottery specimens from the Good Soldier Component: a, b, Iona Indented; c, d, Grey Cloud Horizontal-Incised; ¢, f, Talking Crow Straight Rim; g, Cadotte Collared; h, Stanley Tool Impressed; 7-&£, miscellaneous. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 49 Stone and bone specimens from the Good Soldier Component. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 50 me EE Ee u SCM Pottery and stone specimens from the Badger Component. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 189 River Basin Surveys Papers, No. 38 Archeological Investigations in the Toronto Reservoir Area, Kansas By JAMES H. HOWARD 319 i Tal pie a) oe fi an ee ralgi a via PEN aa ae 1 / . Nr } ef! nh 1 ‘ ' ao fii Pact ij ft hay + ele alien nivale bss beams oe ‘x u : “> wise vs if ce aa a : oa | , a i ie ae ‘ unt y ; : re ‘ +) i) } @ : y + | i ‘ i i : ; ' Sy) on) ij 7 ¥ ° : it é é Py Os aaa ; & : ; ¥ ! § ite } — ab Slee paageich oraes er \ y : eS * n2 i * nv a y * ‘ tes 4 hb ad a i Facey se 7se | wa ia oy A f= (4 testis = = ' CONTENTS PAGE Introduction]. .- 2. 22 Seen A ee eee Soe re AOS) Bes ee 323 ANEYSILCS Seen ee Oe eS TLE A eT Oe Re Ae) PEL SO UI Oe 327 Dito WOU aco. es es SO AS SORA AOE RINT 8 2 328 PIZCAVAIONSes ) a2 see Ok ean Baek NOT ae ees 328 Specimcm aescripuionssos / 2 2.) Sa TOU eaoblL Wea Bhs ee 329 Culturaliatuliation = =~ = 28 2 Se eee ee A 332 SipealaWwOlOg ms soe" aU ee) Me PE 2. Se Se Gy ce 332 HIXCaVaAtlOnS= 2242-8 0s oe a I Se ae See 2 332 Specimen deseriptiongs 2! S781 one 2 oh ap ek ST Se 334 Culturallatnliation st 2222s bee Oa ee. ee Eee OP 337 LUCE AW o Te se ans te i ee eo ne a eas 2 337 IBIS CAVAUION SE) eee ee Sh Le tps Oe te ete, Se ee Oe erage 337 Bpceimien GescriptiOns as 6-45. .ke ea Lo oe ee eee 338 Culturaliafiliations’ == 45—es-.. seen oa ie ae rs. eee 341 The waleve Hackshelter (14WO222)_..) 22.202 2. ye see 341 ET RCS Wav ULONIS psn ees ar ee a eh er ee nee Ee Se yen ee 342 1 ESF) A 1 eas i Nae a ape a ee ee ee et 342 of BOPP TVEIDIS | ECTey gs 100 1 aa SA a ln lea Oe pel SR IG 344 Culluraltattiiationss = see eee Ne tpt A ee a a ee ee 346 DryiCreek Rockshbelter (1l4WO224): 2 ie oe ae Sh ae 347 IER O AVAL LOTS ea et eee ce ery eae ne a 348 E207 oc PN a ga NN ieee La He pean 2 et 348 SPCOnMEn GESCrIPMONSs ©. 2 he een sen sae See 350 Calturallafiliation ses) ses 520 ee ee oe eee aoe ore ee 353 The: Outlaw Rockshelter(14WO225)i 2s. 5.2 eee eee Se ce 353 Little-Water-Man petroglyph (14W0226)__-._____-__------------- 353 he: Possum Point site: (14WO22S)< <2. 2. = ot ee ee 354 IS CAV SUIONS! ee Me eee ee oe see ae ne ee eet 354 Bpecinten Ceseripuons. oo 22a sue ee eek ee ee 354 Cirlt urs alah Orig ak eee AER Se edt eee es Pete ote ee 357 Sitesl4GR2 Qe 5 Oil wee eS ye hy meh eal pe, a) Rs eee ee Saf TUR CAV ATION SSeS a ee a ee aes ae re 357 Specimen, GESCEIDtIOUS a= == ey ook Go ee 358 Cultumalta tila bionic eee ne ee a eee ee ee ree 360 Sted 4G RI21Gsie sss eee aes a ee ee ee 360 TBI RCA VATIONS: ees ete NECN Oe CC ee ee a eee eee ed ae 361 | OEE 5502 21= aR ee es be ae Sela ye Ns ON AR Oy Sn 361 Specimen Gescriptions. 22 see es es eee eee 361 Culturallathitations se eae ee eee le 366 SINHIMmAlyeanercOneClisiOng.. 82 a fo een nee en eae 366 Appendix. Tabulation of Toronto Reservoir sites._____..------------- 368 Piierapure Gi paem seen eae Ste 0s nO a LO ee a ee Be ce 369 322 51. 52. ILLUSTRATIONS PLATES (All plates follow page 370) a-l, Artifacts from site 14W0203. m-k’, Artifacts from site 14W0O209. a-d’, Artifacts from site 14WO215. e’-p’, Artifacts from Walleye Rockshelter (14W0222). 53. Petroglyphs from Walleye Rockshelter (14W0O222). 54. a, Excavation of Dry Creek Rockshelter (14W0O224). 56-t, Artifacts from Dry Creek Rockshelter (14W0224). 55. a, b, Petroglyphs from Dry Creek Rockshelter (14WO224). c, Petroglyph at the Outlaw Rockshelter (14W0O225). d, e, Petroglyph from site 14W0226. 56, a, Excavation at Possum Point site (14W0228). b-k, Artifacts from Possum Point site (14W0O228). 57. a, View of site 15GR210. 6-m, Artifacts from site 14GR210. 58. a, Portion of excavated area in site 14GR216. b-w, Artifacts from site 14GR- 216. TEXT FIGURES PAGE 59. Site map of the Toronto Reservoir area__.....---.----------- (facing) 327 60. Plan of site 14WO2038, showing excavations_______----------------- 329 61. Plan of site 14WO209, showing excavations_____-__---------------- 333 62. Plan of site 14W0O222, showing excavations.___.__.._.-------------- 343 63. Plan of site 14W0O224, showing excavations____-------------------- 349 64. Plan of site 14W0O228, showing excavations_____-_----------------- 355 hb. Plan of site 14G R210; showime excavations 22" 5222S eee 359 66. Plan of site 14GR216, showing excavations_-______----------------- 362 ARCHEOLOGICAL INVESTIGATIONS IN THE TORONTO RESERVOIR AREA, KANSAS* By James H. Howarp INTRODUCTION The Toronto Dam and Reservoir, a flood-control and conservation project of the U.S. Army, Corps of Engineers, Tulsa District, is located on the Verdigris River in Greenwood and Woodson Counties, southeastern Kansas (see fig. 59). The dam isa rolled, earthfill struc- ture, 4,712 feet in length and 90 feet in height, above the river channel. It is 314 miles south of the town of Toronto in Woodson County, Kans., 55 miles north of the Oklahoma border, and 75 miles west of the Missouri border. It forms a reservoir some 1114 miles long, with a maximum width of approximately 114 miles at full-pool level of 931 feet (m.s.l.). This reservoir extends up the Verdigris River and its tributaries in the eastern portion of Woodson County and the western portion of Greenwood County to cover a total area of approximately 10,000 acres. Dam construction was begun in November 1954 and was essentially completed in December 1959. Flooding of the reservoir began in the spring of 1959 with maximum pool level to be reached intermittently after completion. In this area the Verdigris River flows in a generally southeasterly direction, forming a wide flood plain between bluffs of Pennsylvanian sandstone and limestone. Numerous bends and meanders indicate that the drainage system is old. A heavy growth of trees, principally scrub oak, elm, cottonwood, and hackberry, clothes the valley and the sides of the bluffs. Above the valley proper, the bluffs slope gradually upward to rolling, grass-covered plains. This upland country is used mainly for grazing, since numerous sandstone and limestone outcrops hinder the use of the plow for breaking the land. In early historic times, two Thegiha-speaking Siouan tribes, the Osage and the Kansa, are known to have lived in this area along the Verdigris River (McDermott, 1940). However, no sites that could definitely be attributed to either of these tribes were located during the archeological investigation. 1 Submitted September 1959. 661-932—64——_24 323 324 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Archeological sites here are characteristically located on slight rises close to a river or stream. Numerous sandstone fragments lie on or near the surface of the sites, and other than these stones, all surface archeological features have been obliterated by intensive cultivation. The principal stone used in the manufacture of chipped-stone arti- facts is a chert from the Flint Hills region, some 50 miles to the northwest, in Marion, Chase, and Morris Counties, Kans. A popular source of material for ground-stone tools is a coquina-like Permian limestone of the Cottonwood formation, which outcrops in Greenwood County. The Inter-Agency Archeological Salvage Program has conducted three seasons of archeological investigations within the area of the Toronto Dam and Reservoir (see report of the Committee for the Recovery of Archaeological Remains, 1958). The first season of in- vestigations consisted of a brief, initial survey of a portion of the area in May 1953 by Edward H. Moorman, working out of the Region 3 Office of the National Park Service. The short time available for this survey limited the results to the locating, visiting, and recording of but three archeological sites in Greenwood County (14GR1-3) and two sites in Woodson County (14WO1-2). No testing or excava- tion was possible, but the survey was sufficient to indicate that con- siderable archeological potential existed in several sections of the flood area. Additional survey and investigation was recommended within the framework of the Salvage Program. The second season of salvage in the Toronto Reservoir was under- taken in September and October 1956, by Alfred E. Johnson and two assistants, working out of the Lincoln, Nebr., office of the River Basin Surveys, Smithsonian Institution. Johnson, then a student at the University of Kansas and a temporary staff member of the River Basin Surveys, and his party worked intermittently for 6 weeks in the area, revisited the sites located by Moorman, made additional surface collections from them, and excavated three 5-foot test squares in one of the most promising of them (14WO1). In addition, this party located, visited, and recorded 18 more sites in Greenwood County (14GR201-218) and 13 more sites in Woodson County (14 W0201-213), to bring the total of sites in the Toronto Reservoir area to 36. Johnson also dug three 5-foot test squares in site 14W0O203 and excavated three hearths in site 14GR209 that had been partially exposed by heavy equipment in road-construction activities. Surface collections of artifacts were made from nearly all sites located, petro- glyphs were recorded in one site, and data regarding terrain, geology, and general archeological considerations were recorded. Johnson (1957) prepared a report of these sites, analyzing the materials re- covered, recommending certain sites for more intensive investigation, Riv. Bas. Sur. ; ies Day. No 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 325 and suggesting that some portions of the reservoir area still might provide additional sites. On the basis of surface collections and the three brief, subsurface tests, Johnson was able to identify tentatively the cultural assem- blages of nine of the sites as follows: TAG ee Upper Republican IAGIR2O2 oe oo eee ee Kansas City Hopewell and Keith Focus IAGR20 Ges Les Ae Upper Republican TAGR2U2 420 peeves) the 4 Kansas City Hopewell TAGR216 9 ne eee Upper Republican, Keith Focus, and Archaic IGRI see ee Keith Focus TZ ALY (07 oe ee ees Woodland TAWIOZ0G2 25 Sees Kansas City Hopewell 14W0209____________- Upper Republican The remaining 27 sites did not provide sufficient material upon which a cultural identification could be based, though many of them ap- peared to be simple, nonceramic (Archaic?) camps. The nature of most of these sites was such that no significant results would appear likely from further investigations, and none were recommended. However, there were seven sites (14GR2, 14GR202, 14GR210, 14GR 212, 14GR216, 14WO0203, and 14W0O209) where further work ap- peared likely to produce clarifying information. These sites were, therefore, recommended for additional investigation. Johnson then drew the following conclusions regarding the cultural manifestations in the area: Preceramiec remains are rare. Only one component of a three-component site (14GR216) can definitely be assigned to an Archaic complex. Some of the sites presented under the heading ‘““Nonceramice Sites,’? may also be associated with an Archaic complex, but the remains are so scanty that no positive state- ment can be made in that respect. Woodland remains are fairly common. There are four sites or components of sites affiliated with this pattern. Of these, three have artifacts which allow them to be placed within a subdivision of the pattern, the Keith Focus. The remaining site contains traits which do not permit its placement in one of the previously defined foci, but this may be because of the insufficient quantity of material collected. The fact that Keith Focus remains are present in the reservoir extends the distribution of this complex further to the east and south than previously realized. The Hopewell complex has a known distribution through central Kansas as far as Ellsworth. The presence of three sites in the Toronto Reservoir, having an affiliation with this complex, extends its distribution farther to the south and indicates that the complex may have a much wider distribution throughout the eastern section of the state than was previously realized. The presence of sites of an Upper Republican affiliation, situated within the boundaries of the Toronto Reservoir, extends the limits of this cultural complex considerably farther to the south. The nearest site with an Upper Republican affiliation which could be found in a perusal of the literature is located in the Kanapolis Reservoir, on the Smoky Hill River to the north and west (Smith, 1949, p. 295). 326 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Using a recently published date as a starting point, we find that the area to be inundated by the waters of the Toronto Reservoir was occupied, at least, from A.D. 611 (plus or minus 240 years) on. This date is from the Woodruff Ossuary, situated just south of the Nebraska line in Phillips County, Kansas, which has been assigned to the Keith Focus of the Woodland Pattern (Wedel and Kivett, 1956, p. 414). This beginning date would probably be extended considerably . into the past if the material from the Archaic component at 14GR216 were sufficient to allow its identification with other sites from that complex. [Johnson, 1957, pp. 58-59.) The third and final season of archeological salvage in the Toronto Reservoir area was accomplished during a continuous 6-week period from May 15 to June 30, 1957, by Dr. James H. Howard and four assistants, working out of the Lincoln, Nebr., office of the River Basin Surveys, Smithsonian Institution. Howard, then a temporary staff member of the River Basin Surveys, and his party revisited many of the sites located by Moorman in 1953, and by Johnson in 1956, made additional surface collections from them, and conducted excavations in four of the sites that had been recommended by Johnson for further work. All four were open, occupation areas. One of these was site 14G-R210, where four 5-foot test squares were excavated, and to which Johnson’s data did not permit assignment of a cultural affiliation. A second was site 14GR216, where a trench 65 feet long and 5 feet wide was excavated and later widened by excavation of three 5-foot squares on each side of the base trench. Johnson had assigned the three com- ponents of this site to the Upper Republican Aspect, the Keith Focus, and the Archaic, respectively. A third was site 14WO203, one of those briefly tested by Johnson with three 5-foot test squares. The Howard party dug three additional 5-foot test squares. Johnson’s identifica- tion of this site was Kansas City Hopewell. The fourth was site 14W0209, to which Johnson had assigned a cultural designation of Upper Republican, and within which Howard excavated two 5-foot squares. The other three sites recommended by Johnson for further work (14GR2, 14GR202, and 14GR212) were revisited in 1957, and additional surface materials were collected, but the sites were not considered of sufficient potential, at that time, to warrant excavation. In addition to revisiting these previously recorded sites, the 1957 party located, visited, and recorded 3 more sites in Greenwood County (14GR219-221) and 18 more sites in Woodson County (14W0O214-231) to bring the final total of recorded sites in the Toronto Reservoir area to 57. Only four of these new sites were of sufficient archeological significance to warrant excavation or assignment of cultural affiliation, though petroglyphs were recorded in two others (14W0O225 and 14W0226). The four significant sites included 14W0O215, in which Howard excavated two 5-foot test squares; the Walleye Rockshelter (14W0224), in which nearly the entire surface area (eleven 5-foot 0 ed 81 SOWB I= Se: ear TSSOWws! beth 1 SHOR LOW IA’ TOSOWPR IA. au SSO, SOWA g SAAB” a4 EGG Or ossowsl\\ ’ WR issowe| \\\ \ 7) ‘ : ‘OTMOR ves ei wt \ A vhs ee Dae. Ne i ts , J BISOWPI OT CISOWR] § pale :, posows | —A— =e Aw ESSOWs | Bsa sya SSSOWs| rd’ Bccpwn-—— Mates - dwhe cnr, r : AN e;SOwai— 7 ae a ae ¢ r o. Sy < Ns ee ., iain a ah Sead TaSSows | ‘ Se OlSsOwal ; . Miss aa a. vanaly | . yimb ‘i ,, a) * | OSOWF | 7 21KA 2 ay en ; «Webb, Jeot a al 34 _¥TVUOD WO2CoOW _ YTMUOD YO2 IW t BheViAN WABVIA 21 AOlGAAaV AZAA ALOWARS3A OTMOAOT , 9 GAM 2@A33V 013 AO 25R00 hae), MOR G3TFAGA . Pinctet 50.—Site mop of the Toronte Reser ar en eT 661-932 O - 64 (Face p. 273) sh KANSAS | TORONTO RESERVOIR ie. Riv. Bas. Sur. ey Pep. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 327 squares) was excavated; and the Possum Point site (14W0O228), in which he excavated six 5-foot squares. The present report will detail the work done by Howard’s party in the eight excavated sites and the two petroglyph sites mentioned above. The work of Edward H. Moorman in the 1953 season was done under the supervision of Dr. Erik K. Reed and Charlie R. Steen of the Region 38 office of the National Park Service, Santa Fe, N. Mex., under whose jurisdiction the southern Kansas area was ad- ministered at that time. The work of Alfred E. Johnson and his party in 1956, and that of Dr. James H. Howard and his party in 1957, was done under the administrative supervision of Dr. Frank H. H. Roberts, Jr., Director of the River Basin Surveys, and Dr. Robert L. Stephen- son, Chief of the Missouri Basin Project. The Toronto Reservoir is situated outside the Missouri Drainage Basin, and funds for the work were transferred by the National Park Service from the appro- priations for work outside the Basin. For purposes of convenience and economy, the work was administered through the Missouri Basin Project office of the River Basin Surveys, Smithsonian Institution, in Lincoln, Nebr. Assisting Johnson in the field were Wayne O. Wallace and Gay- lord S. Tefft. Tefft and Richard Fischer assisted Johnson in his laboratory analyses, and Sidney Anderson identified the animal bones. All five were students at the University of Kansas, where Dr. Carlyle S. Smith kindly made laboratory facilities available to Johnson. Assisting Dr. Howard in the field were Edward A. Danaczko, D. Wil- liam Chatfield, August Love, and Joseph Marshno, employed by the Smithsonian Institution for the project. The regular staff of the Missouri Basin Project in Lincoln assisted Howard in his laboratory analyses and preparation of this report. Others who assisted in a number of ways in the archeological salvage in the Toronto Reservoir area include George Fritz, Mark Sample, George Phillips, Albert Webb, George Webb, Frank J. Adenauer, J. E. Sower, H. W. Pashe, and Fred Jamison, all of Toronto, Kans.; Richard Phillips and Lester Harding of Yates Center, Kans., and Mr. and Mrs. Kenneth Landes and Mrs. James H. Howard of Kansas City, Mo. Mr. Lloyd Tanner of the University of Nebraska State Museum in Lincoln identified the bone materials from the 1957 excavations. Sincere thanks are ex- pressed to all of these people for their generous assistance, without which the work would have been much less pleasant and rewarding. THE SITES The 57 archeological sites recorded in the Toronto Reservoir are located on the reservoir map (fig. 59) and tabulated in the Appendix. Johnson (1957) has described and identified some of these sites, and WOODSON COUNTY GREENWOOD COUNTY f 14GR209 | 14GR208 14GR207 fe 14GR206 i 14wO206 |4GR216¢ ) .. 0 1 4GR22 YY 5, 14GR202 14GR20 1 ie. |4GR203 a = 4W0217 i aa 16 wy, 14GR204 LF |\4wo227 14GR21| 14wo23! 14wWO207 |4Wo| 22 14GR215 14WO2 || oz |4wo220 0 gon? = 14wo22! pes i a ( eee o —__—_4GR220 *——_|4GR214 *—14G6R213 14GR212 _To 14wozia, @ |4woz219 14wo204 14wO0223 —|4w0222 14w0228 14wo202 14w02!5 14GRi/ 14wo213- if ayes SS . 6 Stes |4W0225 ~14wW02 10 14w0209 14wO0208 \ | | Be Pa J va ee LEN 3. —- VERDIGRIS RIVER, KANSAS TORONTO RESERVOIR AREA ADAPTED FROM, 1954, CORPS OF ENGINEERS MAP Figure 59.—Site map of the Toronto Reservoir area. ——— — le To 328 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 others are so insignificant as to warrant but little comment. These latter produced so few artifacts and so little archeological data that no more may be said than that they are probably places where aborig- inal peoples of unknown affiliation have camped at some unknown time. A third group of these sites did, though, provide sufficient information, when excavated, to support cultural identification and thus warrant — rather full descriptions here. The following analyses of sites include only those in the third group, the sites that were partially or fully excavated in the 1957 season. As 6 of these 10 sites have more than 1 component, they will be described in the order in which they were investigated, rather than in the order of cultural sequence. The Woodson County sites are described first, followed by the Greenwood County sites. Five culture complexes have been identified in the Toronto Reser- voir area. These are: Archaic (5000 to 200 B.C.) ; Kansas City Hope- well (200 B.C. to A.D. 500); Plains Woodland (A.D. 500 to 900) ; Aksarben (A.D. 1100 to 1500) ; and Great Bend (A.D. 1500 to 1700). The dates given in parentheses are provisional and subject to any re- visions which new information may suggest. Srre 14W0203 This site is located on a slight rise on the northwest side of the Verdigris River (fig. 59). It is bounded on the northeast by an abandoned railroad grade, on the southeast by the Verdigris River, and on the southwest and northwest by an oxbow lake. An area of refuse concentration approximately 150 feet in length by 140 feet in width lay near the center of the site. The entire site has an over- all length of 900 feet. Long utilization of the site for agricultural purposes has obliterated any traces of surface features, if they were ever present (fig. 60). EXCAVATIONS Johnson and his party tested this site by means of three 5-foot- square test pits. Two of these were carried to a depth of 1.5 feet, and a third to a depth of 2.0 feet. Arbitrary levels of 0.5 foot were maintained during the excavation. As agricultural operations had obliterated all traces of these test pits by the time the site was mapped in 1957, the locations of these pits do not appear on the site lan. , In the summer of 1957, Howard and his party further tested the site by means of three more 5-foot-square test pits, placed in areas of greatest surface concentration and artifacts. Two of these were carried to a depth of 4.0 feet, and one to a depth of 2.0 feet. In the first of these, no artifacts were recovered below the surface until a Riv. Bas. Sur. Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 329 TEST PIT no.2 0 TEST PIT D NO. 1 14W0 203 TORONTO RESERVOIR FEET Ficure 60,—Plan of site 14WO203, showing excavations. depth of 1.0 foot was reached. Here the broken end of a drill or projectile point and a fragmentary scraper were recovered. Nothing was found below this depth. In the second pit, the only artifact, a large body sherd, was re- covered at a depth of 0.5 foot. The third pit also yielded but a single artifact. This was a large, expanding-stem point, which ap- peared at a depth of 1.5 feet. SPECIMEN DESCRIPTIONS (pl. 51, a-t) Rim sherds.—Two rim sherds were recovered, one from the sur- face of the site and one from a test pit dug by Johnson’s party. The specimen from the surface (pl. 51, a) is slightly flaring and has a 330 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 rounded lip. Decoration consists of a series of vertical, incised lines, running downward from the lip for a distance of 22 mm. and, slight- ly below the lower limit of these lines, a single row of dentate stamp impressions. The sherd is tempered with grit, and its color ranges from tan through gray to black. The lip thickness is 9 mm. and the lower rim thickness is the same. ) The second specimen (pl. 51, 6) shows considerable weathering and lacks the lip. Its decoration consists of square dentate stamping, arranged in parallel lines. Below this is a horizontally incised line, and below this, and here and there superimposed upon it, is rocker- stamping. Temper, color, and thickness are the same as in the speci- men described above. Body sherds—Of the 51 body sherds obtained, 32 were from the surface of the site, 18 were recovered by Johnson’s party in their test pits, and 1, mentioned above, was recovered in Howard’s test pit 2. Forty-six are plain, and five are cord marked. Superimposed on the surface of the plain sherds are examples of rocker-stamping, dentate-stamping, and parallel, incised lines. Temper and color are the same as in the rim sherds described above. Thickness ranges from 6 to 12 mm. Projectile points—Two complete projectile points and the bases of three others were recovered. These can be divided into two categories on the basis of size. Four of the points are in the “large point” tradi- tion and probably served as atlatl dart points (Fenenga, 1953). They all have expanding stems with straight or slightly convex bases. The only complete point of this style (pl. 51, /) was recovered at a depth of 1.5 feet in Howard’s test pit 3. It measures 84 mm. in length, is 27 mm. wide at the widest point, which is just above the notches, and has a maximum thickness of 9mm. Its base is slightly convex. The point is made of a grayish-white chert. Another point of this style (pl. 51, e), consisting of only the basal portion, is made of a gray chert with white inclusions. It has a maximum width of 40 mm. and a maximum thickness of 7 mm. Its base is straight. The third point of this style, also a fragment, has a slightly convex base (pl 51,d). It has a maxi- mum width of 40 mm. and a maximum thickness of 8mm. It is made of gray Flint Hills chert. The fourth large point, very fragmentary, is of dark gray chert. Its maximum thickness is 5 mm. The remaining point is in the “small point” tradition (pl. 51, ¢) and probably served as an arrowhead (Fenenga, 1953). It is triangular in outline and unnotched, made of gray Flint Hills chert. Its length is 11 mm., its maximum width 9 mm., and its maximum thickness 2 mm. Points of this type are generally considered to be “late horizon” markers, i.e., characteristic of late prehistoric and historic cultures. Drill——One possible drill point, fashioned from gray chert, was Riv. Bas. Sur. ite Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD ool obtained. This specimen was found in Howard’s test pit 1 at a depth of 1 foot. The point of the drill is 2 mm. wide. It has a maximum thickness of 6 mm., measured at the point where this portion was broken off, and a maximum width of 15 mm., measured at the same place. Graver—One graver, suitable for incising wood or bone, was obtained by Johnson in his testing. It has been fashioned from an ir- regularly shaped flake of gray Flint Hills chert. The only modifica- tion of the flake is in the point itself, which is extremely narrow. The width of the point is 2 mm.; the width of the base 20 mm.; and the length of the tool 83 mm. Its maximum thickness is 3 mm. Blades—The bases of three blades were recovered, two from the surface of the site and one from one of Johnson’s test pits. The largest blade fragment is of gray Flint Hills chert (pl. 51,¢). It is rounded, indicating that the complete blade may have been pyriform in shape. It is 47 mm. in width and has a maximum thickness of 11 mm. A second rounded base, also of gray Flint Hills chert, measures 28 mm. in width and has a maximum thickness of 6 mm. The third basal fragment has a contracting stem with a rounded base. This piece is of pink chert. Its dimensions are: maximum width, 29 mm., maximum width of stem, 22 mm., maximum thickness, 5 mm. End scrapers—F ive objects of this sort were collected. Three are of gray Flint Hills chert, two are of pink chert. In all examples the working end is thicker and has been sharpened by the removal of flakes almost at right angles to the plane of the artifact. Two of the scrapers are pyriform in outline, one is trianguloid, and the remaining two, fragments, appear to have been rectanguloid when complete. Lengths of the pyriform specimens are 51 and 72 mm., maximum widths are 30 and 42 mm., and maximum thicknesses are 10 and 14 mm., respectively (pl. 51, 2,7). The trianguloid specimen, which appears to be the re- worked tip of a projectile point, measures 24 mm. in length by 17 mm. in width, with a maximum thickness of 4 mm. (pl. 51,7). The maxi- mum widths of the rectanguloid specimens are 40 and 20 mm., their maximum thicknesses 7 and 6 mm. (pl. 51, %, 7). Flake scraper.—One fragmentary artifact bearing traces of second- ary chipping along two edges was recovered. It is made of gray Flint Hills chert, is quite irregular in form, and has a planoconvex cross sec- tion. It is 43 mm. long, 33 mm. wide, and has a maximum thickness of 14mm. Chopper.—One fragmentary chopper, made from a fine-grained green quartzite, was found on the surface of the site. The flaking is entirely of the percussion type, and is restricted to the edges of the specimen. The edges show evidence of repeated pounding, indicating a possible secondary use as.a hammerstone. Dimensions: width at point of breakage, 48 mm., maximum thickness, 18 mm. 332 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Miscellaneous chipped stone fragments——In addition to the above specimens, which can be placed in artifact categories, 31 specimens were collected that are too fragmentary to be classified. These con- sist of 6 projectile-point tips, 5 projectile-point or blade fragments from the area between the base and the point, 2 probable end-scraper fragments, and 18 fragments which show secondary chipping along © only one edge. The predominant material is gray Flint Hills chert. One unfinished projectile point was also recovered. Unidentified artifact—One piece of highly siliceous, worked hema- tite with a hardness of 6.5 (vesuvianite in the modified Mohs hardness scale) was recovered. It shows definite signs of rubbing and scratch- ing on two surfaces. Were it not for its hardness, one might suspect that it had served as a source of pigment. The piece measures 54 mm. in length, 34 mm. in width, and averages 7 mm. in thickness. Paint.—Three limonite fragments and four burned limestone frag- ments were collected. All are probably of local origin, and probably served as sources of yellow and white paint. Unworked stone—There were 327 irregularly shaped flakes of chert, lacking any traces of secondary chipping, recovered at the site. Gray Flint Hills chert is the predominant material, although tan and white cherts are also represented. CULTURAL AFFILIATION Potsherds with dentate- and rocker-stamping and large, expanding- stem projectile points indicate an identification with Kansas City Hopewell as defined by Wedel (1943). Since no cultural stratification could be discovered and the artifacts appear to be culturally homogene- ous, it is assumed that the site has but one component. Therefore, specimens from the surface were grouped with those from the exca- vations for descriptive purposes. Srre 14W0209 This site is located on a slight rise on the northeast side of the Verdigris River (fig. 59). It covers an area approximately 1,100 feet in length (NW-SE) by 300 feet in width (NE-SW). A slightly heavier concentration of material is found at the southeast end of the site in an area 200 feet by 180 feet. Any surface features that may once have been present have been destroyed by cultivation (fig. 61). EXCAVATIONS Two 5-foot-square test pits were laid out in the area of greatest surface concentration. The first of these was excavated to a depth of 1.0 foot, the second to a depth of 0.7 foot, at which point work was TORONTO RESERVOIR AREA, KANS.—HOWARD ga0 Riy. Bas. Sur. Pap. No. 38] “SUOTJEAVIXO sutmoys “60ZOMPI ays JO uUe[G—']9 TUNIS T 1334 '— WANJZLNI YNOLNOD YIOAYSSSY OLNOYOL 6020M 1 SASAYNS NISVE YSAIY NOILNLILSNI = NVINOSHLIAS @ anys? NOLWYINSONOS Jo vauy = @ did 1s310 \ \ 1 t 4 J did 16319 < “ 2 - ~ - t ayvis® "Rouv7 Jo 3svea lv 334 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 suspended at the insistence of the owner of the site, who feared that his machinery might become mired in the pits during farming opera- tions. It appeared, however, that at this depth the occupation layer had already been passed, as no artifacts had been recoved below a depth of 0.5 foot in either pit. Potsherds and flint chips were recovered in both pits to a depth of 0.5 foot from the surface. SPECIMEN DESCRIPTIONS Lim sherds—Three rim sherds were recovered, all apparently from vessels of a cord-marked Aksarben type. The first bears vertical, cord-marked lines on the outer surface and has a rounded lip. The width of the lower rim is much greater than that at the lip, indicating that the fragment was once part of a collared rim (pl. 51, m). The second rim is also cord marked on the outer surface, but the cord marks have been wiped away for a distance of 10 mm. below the top of the lip. The lip is rounded and the rim is straight or very slightly in- sloping (pl. 51, 0). The third rim sherd is similar to the second in most respects, but seems to have come from a much smaller vessel (pl. 51,”). In color the sherds grade from buff to gray. The tempering is coarse grit. Lip thickness on all three sherds is 5 mm.; the lower rim thickness is 9 mm. on the first rim and 5 mm. on the other two. Body sherds.—All of the 114 body sherds recovered represent pot- tery of the Aksarben Complex. Eighty-four of these show cord mark- ing on the outer surface, twenty-three are smooth, and seven are in- determinate (i.e, split or badly weathered sherds). In color they range from buff through gray to black. Tempering is coarse grit. Thickness ranges from 4 to 12 mm. Loop handle.—A single pottery loop handle was recovered by John- son’s party from just north of the main concentration area of the site (pl. 51, p). The handle had been fastened to the vessel by drilling holes in the vessel wall, inserting tabs of clay from the handle, and welding the tabs to the inside of the vessel. The smoothed exterior bears a bright red-orange slip, the interior is gray. Coarse grit is the tempering material. The diameter of the handle is 14 mm. This fragment has been assigned to the Geneseo Red Filmed type (Wedel, 1949, p. 89). In Rice County, in central Kansas, Wedel found sherds of this type associated with Rio Grande Glaze Paint sherds (ca. A.D. 1525 to 1650), and small amounts of European and Southwestern trade goods. Pottery of the Geneseo Red Filmed type is one of the diagnostic traits of the Great Bend Aspect or Paint Creek Culture, which Wedel has tentatively identified with the late prehis- toric and early historic Wichita (Wedel, 1942, p. 10). Projectile points—Twenty projectile points that were sufficiently complete for classification were recovered. These have been grouped Riv. Bas. Sur. roe Dap. Ne 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 335 into six different styles. The first, represented by one incomplete specimen, is large and thick, and has in-sloping shoulders and a slightly expanding stem with a straight base. It is made of red chert and measures 45 mm. in length by 21 mm. maximum width by 10 mm. maximum thickness. Typologically, it would seem to belong to the Archaic Culture (pl. 51, ¢). The second style, also Archaic typologically, is represented by two complete points. Both are large and thick, with in-sloping shoulders and contracting stems. The first measures 49 mm. in length, 21 mm. in maximum width, and 11 mm. in maximum thickness. It is made of tan chert (pl. 51,7). The second is 56 mm. in length, 22 mm. in maxi- mum width, and has a maximum thickness of 11 mm. It is made of gray Flint Hills chert (pl. 51,s). The third style is represented by five incomplete specimens. These points are large, with expanding stems and straight or slightly con- vex bases. The first measures 25 mm. in width and has a maximum thickness of 8 mm. (pl. 51, ¢).. The second is 27 mm. wide and has a maximum thickness of 11 mm. (pl. 51, wv). Both are of gray Flint Hills chert. The third specimen, of gray Flint Hills chert, has a max- imum width of 28 mm. and a maximum thickness of 7 mm. The fourth is 21 mm. wide, with a maximum thickness of 5 mm. It is of gray Flint Hills chert that has white banding. The fifth is 27 mm. in maximum width and 5 mm. in maximum thickness, made of gray Flint Hills chert. The fourth style is represented by one complete point, one reworked point, and three fragmentary specimens. Points of this type are dis- tinguished by their large size and their contracting stems. The com- plete specimen measures 70 mm. in length, is 23 mm. wide, and has a maximum thickness of 10 mm. (pl. 51, w). Other points of this type are wider, the widest being 33 mm. in maximum width. Three are of tan chert, the other two of gray Flint Hills material (pl. 51, v,2,Y). All of the above are in the “large point” tradition. The two remain- ing styles are “small point” forms. The fifth style is made up of small, triangular, “late horizon” points. Two specimens are unnotched. They measure 24 and 20 mm. in length, 13 and 14 mm. in width at their bases, and 8 and 5 mm. in maximum thickness, respectively. Materials are white and gray Flint Hills cherts (pl. 51, 2, a’). The remaining points are notched. Two of the points have double side notches, one has single side notches, and one (perhaps accidentally) has one side notch and a basal notch. Two are of gray Flint Hills chert, the others are of a pinkish chert. They average 20 mm. in length by 15 mm. maximum width, with a thickness of 3mm. (pl. 51, 0’, c’, d’). A sixth style, possibly a drill rather than a projectile point, is repre- sented by a single broken specimen. It is slender, with shallow side 336 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 notches, and is slightly curved when viewed from the side. It was probably about 33 mm. in length when complete, has a maximum width of 10 mm., and a maximum thickness of 3mm. It is made of gray Flint Hills chert (pl. 51, e’). Blades——¥our pyriform blades, two nearly complete and two frag- © mentary, were recovered. All are of gray Flint Hills chert. They vary in width from 22 to 40 mm., and in thickness from 9 to 13 mm. The two nearly complete specimens measure 53 and 59 mm. in length (pl. 51, f’). End scrapers—Two complete planoconvex end scrapers were re- covered. Both are manufactured of gray Flint Hills chert. The first measures 35 by 30 mm. and has a maximum thickness of 7 mm. (pl. 51, h’). The second measures 45 by 30 mm. and has a maximum thickness of 10mm. (pl. 51,9’). Flake scrapers—Two artifacts of this description were recovered. The first is made of pink chert and measures 26 x 20 mm. with a maxi- mum thickness of 4mm. The second, of tan chert, is 72 mm. long, 27 mm. in width, and 11 mm. thick (pl. 51, 7’). Both are ovate in form, flat on one side, worked on the other. Chopper.—The basal portion of a heavy stone chopper was recov- ered at the site. It is 70 mm. in width at the point of breakage and has a maximum thickness of 25 mm. Celts —Three small celts, similar to each other in shape, but of var- ied materials, were found at the site. The first and smallest has been chipped from a fragment of dark gray micaceous schist. Its cutting edge is polished from use. It measures 80 mm. in length, 43 mm. in width, and has a maximum thickness of 13 mm. (pl. 51,7’). The sec- ond is made of the Cottonwood formation limestone so common in artifacts from this area. It has apparently been cut into shape by an implement of some harder stone. This celt is 100 mm. long, has a maximum width of 46 mm., and a maximum thickness of 27 mm. It seems rather odd to find a celt made of this soft, rather porous stone (hardness 3.5, celestite), yet collections from this area contain celts, boatstones, and even grooved axes fashioned of this material. The third celt has been chipped from a light tan chert. It is 123 mm. long, has a maximum width of 60 mm., and a maximum thickness of 27 mm. (pl. 51, %’). Manos.—Kight manos, and mano fragments, were recovered. Seven of these are of sandstone, and the smallest specimen is made of mica- ceous schist. All are subrectangular, with smooth grinding surfaces on the top and bottom and partially smoothed edges and corners. The largest specimen is 153 mm. long, 96 mm. wide, and 40 mm. thick. Six of the others are but slightly smaller than this. The eighth specimen is but 55 mm. long, 50 mm. wide, and 23 mm. thick, and appears to have Riv. Bas. Sur. = SITE Pap. No. 38}, TORONTO RESERVOIR AREA, KANS.—HOWARD 337 been used as a grinding tool for small materials such as medicines or perfumes. Three of the larger manos appear to have been used also as cup- stones or anvils, as they have paired depressions on the two opposite, flat surfaces. The only complete specimen of this sort measures 117 mm. in length, 55 mm. in width, and is 56 mm. thick. The pecked de- pressions are 23 mm. in diameter and 4 mm. thick. Hammerstones.—Three hammerstones were recovered. All are un- worked river pebbles of a size to fit conveniently in the hand of the user, and all show signs of battering on one or more surfaces. Core.—A pyriform core of gray Flint Hills chert was found at the site and probably represents a piece of material roughly shaped at the quarry for transportation to the site, where it could be made into points, scrapers, etc. It is 102 mm. long, 72 mm. wide, and has a maxi- mum thickness of 35 mm. Miscellaneous stone artifacts——In addition to the artifacts de- scribed above, there were 28 chipped-stone specimens too fragmentary to classify. One of these seems to represent the end of a small end scraper. Three seem to be portions of flake scrapers, and the remain- der may be portions of projectile points or blades. The predominant materials are gray Flint Hills chert, tan chert, and pink chert. CULTURAL AFFILIATIONS Although no cultural stratification could be discovered, the arti- facts from this site indicate that at least two components are present: Archaic (as defined by Spaulding, 1955, pp. 15-19) and Aksarben.? A single loop-handle fragment may represent the third and most re- cent component at the site, Great Bend. Srre 14W0O215 This is a small site, limited to a low knoll and an adjacent level area near the Verdigris River (fig. 59). The area has been under cultiva- tion for some time and has also been subjected to continual flooding during periods of high water. Nevertheless, it is still quite rich in surface materials. EXCAVATIONS The site was tested by means of two 5-foot-square test pits, exca- vated in the area of greatest surface concentration. The first pit was excavated to a depth of 2.5 feet, the second to a depth of 2.0 feet. The fill was clay loam at the surface, grading into hard clay at a depth of 1.0 foot. Numerous sandstone blocks were encountered at the top 2'The term ‘‘Aksarben” is used for the archeological remains which make up the Upper Republican and Nebraska cultures. 338 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 of this hard clay. Plow disturbance was evident to a depth of 0.5 foot. The only artifact recovered in the first pit was a cupstone, which appeared at a depth of 1.1 feet in the northwest corner of the square. In the second pit an occupation level was reached at a depth of 1.0 foot and continued to a depth of 1.5 feet. It contained a flake scraper and numerous small chips. SPECIMEN DESCRIPTIONS Rim sherd.—Only one rim sherd was recovered from this site (pl. 52, 6’). It is, however, a rather unusual specimen, since it seems to have come from a bowl-shaped vessel. The sherd is cord marked on the outer surface, smooth on the inner. It is grit tempered, with a light-gray interior and buff exterior. It is 7 mm. in thickness. The upper surface of the lip and the outer rim, for a distance of 12 mm. below the lip, are decorated with diagonal cord marking. Below this, the cord marking is vertical. At a distance of 18 mm. below the lip edge, there is a very pronounced shoulder. Although rather atypi- cal, this sherd seems to be identifiable as of Aksarben origin. Body sherds—Seven cord-marked body sherds and five smoothed body sherds were recovered. ‘The cord-marked body sherds are all of Aksarben types. They range in color from dark gray to orange- buff, with gray interiors. Thickness ranges from 6 to 8 mm. (pl. 52,0’). The smoothed sherds are also grit tempered, but the particles of grit are much larger than in the cord-marked sherds. They vary in surface color from reddish-gray to buff, and interiors are uni- formly gray. Thicknesses vary from 5to9 mm. One of the sherds bears a decoration of two fine, incised, parallel lines on its outer surface (pl. 52,d’). The smooth, undecorated sherds have been tentatively assigned to the Geneseo Plain type (Wedel, 1949, p. 88). The incised sherd fits no described type known to the writer, although, like the Geneseo sherds, it probably represents a late time period. The incised lines on this sherd are reminiscent of those on sherds of the Oneota Aspect (ca. A.D. 1600-1800). Projectile points.—A total of 34 projectile points, sufficiently com- plete to be classified, were recovered. ‘These have been grouped into six categories for descriptive purposes. The first style is represented by two incomplete specimens. These are rather large, thick, leaf-shaped points, with only a slight indica- tion of a shoulder. The first is made of a brown chert with white inclusions. It measures 22 mm. in width, has a maximum thickness of 11 mm., and when complete was probably 64 mm. long (pl. 52, 0). The other point of this type, made of gray Flint Hills chert, has a Riv. Bas. Sur. asx Pap. No 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 339 maximum width of 20 mm., is 8 mm. thick, and when complete meas- ured approximately 62 mm. in length (pl. 52, a). The second style, represented by two complete points and four frag- ments, is of the contracting-stem type generally identified with the Archaic or Early Woodland cultures. The longer of the two speci- mens is made of tan chert. Its dimensions are: length, 62 mm.; maxi- mum width, 23 mm.; and maximum thickness, 8 mm. (pl. 52,¢). The other, apparently a reworked point, is 53 mm. long, with a maximum width of 30 mm. and a maximum thickness of 8 mm. (pl. 52, d). Maximum widths of the other four specimens are 87, 38, 25, and 24 mm. The third and most numerous style is large, with barbed shoulders and an expanding stem, either straight or convex at the base. Four nearly complete specimens range in length from 56 to 42 mm., while a clearly reworked point is only 80 mm. long. Widths vary between 56 and 22 mm., and maximum thicknesses between 6 and 10 mm. Seven incomplete specimens of this type were also recovered. Ma- terials are tan and gray Flint Hills cherts (pl. 52, e-<). A fourth style is represented by four nearly complete specimens and one fragment. Here the stems are straight, and the barbed shoulders characteristic of the previous style are absent. Lengths of the four nearly complete specimens are 59, 51,48, and 38mm. Widths are 24, 22,24,and24mm. The average maximum thickness is 8 mm. Materials are gray Flint Hills chert, tan chert, and pink chert (pl. 52, k,l). The fifth style is represented by one complete projectile point and three fragments. These are small, expanding-stem points with barbed shoulders, and resemble those recovered in Plains Woodland sites in Nebraska and Kansas (Kivett, 1952, pl. IX A, Nos. 6, 9, and 10; pl. XVI A, No. 4; 1953, pl. 23 a, bottom row, 3d, 4th, and 5th specimens). Materials are gray and pink cherts. The only complete specimen is 37mm. long. Widths of the points are 17, 16, 16,and 12mm. Maxi- mum thicknesses are 4, 4, 4, and 3 mm. (pl. 52, 7, m-o). The sixth style is represented by five complete specimens. This is the familiar “late horizon” triangular style. All lack side notches, but one has a basal notch. Lengths of the specimens vary from 30 to 19 mm.; widths, measured at the base, between 15 and 11mm. The points average 3.5 mm. in maximum thickness. Materials are gray and cream-colored cherts (pl. 52, p—t). In addition to these specimens, 30 fragments of projectile points, all in the “large point” tradition, were recovered. Materials em- ployed are gray Flint Hills chert, tan chert, and cream-colored chert. Enives and blades—Twenty-seven fragments of knives or blades were recovered at the site. Twenty-three of these were apparently pyriform in shape when complete. The other four were undoubtedly 661-932—64——_25 340 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 of the style known as the “Harahey knife.” Typically, these Harahey knives are lozenge shaped, with alternate beveling on the upper right and lower left edges, as one views the object from either face (pl. 52, 2). Such artifacts are frequently found at Aksarben sites, but are by no means limited to those sites. One of the pyriform blade fragments is of white chert, five are of tan chert, two are of pink chert, two are of gray-green chert, and eight are of gray Flint Hills chert (pl. 52 x, y). Four fragments, appar- ently representing the bases of blades, are of a reddish-gray jasper, while another is of banded tan chert. The four Harahey knife frag- ments are of tan chert. Drill.—One slender, chipped piece apparently represents a section of a drill stem. It is lozenge shaped in cross section, with a width of 10mm. and a maximum thickness of 7 mm. Shaft scrapers.—F ive artifacts, each distinguished by a semilunar notch on one edge, have been referred to as shaft scrapers. Three of these have small, deep notches, and would have been ideal for shaping arrowshafts with a diameter of 7 or 8 mm. (pl. 52, a’). Two others, with shallower notches, may have served as bowshaft scrapers. Both deep- and shallow-notched specimens are small, ranging in length from 25 to 40 mm. Three are made of light-gray chert, one is of tan chert, and one is of gray Flint Hills chert that has white inclusions. End scrapers——A. total of 45 end scrapers were recovered, 27 of which are complete and 18 of which are recognizable fragments. The complete specimens range in length from 30 to 72 mm., in maxi- mum width from 20 to 36 mm., and in maximum thickness from 7 to 13 mm. Light gray, tan, and gray Flint Hills cherts are the usual materials (pl. 52, ww). Flake scrapers.—Ninety-one flakes with one or more edges bearing secondary chipping were recovered. These vary greatly in size and shape. The materials are tan, white, and gray Flint Hills cherts. Choppers.—Two choppers, both of gray Flint Hills chert, were recovered. They are ovoid in outline, and show percussion flaking on both faces. They measure 92 and 90 mm. in length, 69 and 68 mm. in maximum width, and 17 and 22 mm. in maximum thickness, re- spectively. These tools are so alike in material and workmanship that it seems very likely that they were made by the same person. Miscellaneous worked fragments.——Five small fragments which cannot be assigned to any particular category were recovered. These may be portions of projectile points, knives, blades, or scrapers. All show some evidence of secondary chipping. Materials are tan, brown, white, and gray Flint Hills cherts. Cores——Four cores were recovered. These are the irregularly shaped pieces remaining after numerous flakes have been removed Riy. Bas. Sur. Aree Bap. No 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 341 from a large quarry blank. Three are of tan chert and one is of gray Flint Hills material. Manos——Four complete manos and one fragmentary specimen were recovered. They are subrectangular, with the upper and lower sur- faces worn smooth. Three of the complete specimens are of sandstone, as isthe fragment, while the remaining artifact is of limestone. The three sandstone manos have the following dimensions: lengths, 140, 125, and 72 mm.; widths, 94, 90, and 60 mm.; thicknesses, 55, 41, and 30 mm., respectively. The limestone specimen is 117 mm. long, 94mm. wide, and 55 mm. thick. All of the specimens show plow scars. Hoe.—A single hoe or hoe-like implement was found. It is made of a quartzite river pebble, roughly triangular in outline, which has been sharpened at the broad end by percussion flaking. It is 108 mm. in length, has a maximum width of 66 mm., measured at the broad end, and a maximum thickness of 32 mm., measured at the narrow end. Celt.—A single celt was recovered. It is made of Cottonwood lime- stone, a rather poor material for a cutting tool, one would think, be- cause of its softness. The celt is typical in shape of those pecked from harder material. It is 83 mm. long, has a maximum width of 51 mm., measured at the blade end, and a maximum thickness of 23 mm., measured at the poll. Unidentified stone objects —Two pieces of worked siliceous hema- tite were recovered. Both appear to be too hard to have served as sources of paint, although it is difficult to conceive any other use for them. Perhaps they were rubbed on some abrasive surface to secure the desired pigment. One measures 27 mm. in length, 23 mm. in width, and 9 mm. in thickness. The other is 22 mm. long, 10 mm. wide, and 4 mm. thick. Unworked stone-—A small piece of micaceous schist found at the site was probably brought there by an individual who was attracted by its sparkle and texture. It is 43 mm. long, 31 mm. wide, and 8 mm. thick. CULTURAL AFFILIATIONS No cultural stratification could be distinguished at the site, since all of the diagnostic artifacts were surface finds. On the basis of typology, three components may be distinguished: Archaic, Aksarben, and Great Bend. Tue Watteyve RocksHenter (14WO222) This site is a small rockshelter overlooking Finger Creek, which enters the Verdigris from the north (fig. 59). Several rather striking petroglyphs are present on a fallen slab in front of the shelter (pl. 53). A narrow occupation area about 10 feet in width and 40 feet in length 342 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 is located immediately below the bluff overhang. The fill of the occu- pation area is a rich humus mixed with fallen sandstone fragments from the bluff. The area has been disturbed by rodent activity, picnickers, and the activities of local enthusiasts who believe that the petroglyphs are a treasure map. For this reason, no cultural stratifi-. cation could be distinguished. EXCAVATIONS The entire area beneath the bluff overhang was laid out in 5-foot squares (fig. 62). Each of these squares was carried down to bed- rock, maintaining arbitrary 0.5-foot levels during the excavation. A total of 13 squares was excavated. The 0-foot to 0.5-foot level yielded the following artifacts: Two small triangular points; a large expanding-stem point; the tip of another point or blade; a slender drill; portions of three end scrapers; two flake scrapers; three cord-marked, grit-tempered potsherds of Aksarben affiliation; one thick, plain, grit-tempered sherd; one sand- stone shaft smoother; one cupstone; one stone hoe; one hammerstone; one core; and one plum seed. The 0.5-foot to 1.0-foot level yielded the following: One large ex- panding-stem point; two blade or point fragments; one Aksarben sherd; one shell-tempered sherd; and one fragmentary limestone celt or hoe. The 1.0-foot to 1.5-foot level yielded the base of a large, expanding- stem point; the tip of a large blade; a large, smooth, shell-tempered sherd; and a piece of charred vegetal material. The 1.5-foot to 2.0-foot level yielded the tip of a large point or blade and a complete point of the small, triangular, “late horizon” type. The 2.0-foot to 2.5-foot level, present in only three of the squares, yielded only one artifact, a modified flake. Charred wood also ap- peared in this level. Unworked flakes, animal bone, and other refuse were apparent at all levels. PETROGLYPHS Several petroglyphs, both incised and pecked, were found on a large sandstone slab which once formed a part of the overhang of the shelter. From the orientation of the petroglyphs it seems evident that they were made at a time before this slab had broken loose from the bluff face. If the slab were restored to its original position, the majority of the figures would be upright and on the outer face of the overhang, while one of the figures (pl. 53, a, 6), a warrior with a bow and arrow, would be on the undersurface. 343 TORONTO RESERVOIR AREA, KANS.—HOWARD Riv. Bas. Sur. Pap. No. 38] *suOIqeAvoXa ZuLMOYS ‘77ZOMPFI AIS Jo ue[G—'79 ANNI A334 5—WAYFINI YNOLNOD REEEP] 3NI1 43m18 BIOAMASIY OLNOMOL : YANTSHSOOY SASTIVM % Z2zzOMvl Mae SA3AUNS NISVa U3AtY ake NOJLALILSNI NVINOSHLIWS Bee ae 3 fax QYuvH ee Foe ao asve Zs 3NI0 33m18 344. BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Although the figures are highly conventionalized, the work is skill- fully executed, and seems to conform to a definite aboriginal art style. One of the figures (pl. 53, c, extreme left, and f) has been painted with a weather-resistant red paint, still visible after many years of exposure to the elements. Three of the incised drawings are . anthropomorphic, including the painted figure and the archer men- tioned above. A fourth represents a man mounted on a horse (pl. 53, ¢). In addition, there are pecked representations of cervid (deer and elk) hoofprints (pl. 53, c-e). Details of dress and adornment do not permit definite identifica- tion of the human figures as to tribal group. The roached hairdress was favored by many tribes which frequented the area, including the Kansa, Osage, and Wichita. The faces of two of the figures, the “painted man” and the “archer,” have lines about the eyes and on the chin, and triangular figures on the breast that might represent tattoo marks (pl. 58, a-c,e). This suggests that the Wichita are depicted, since members of this tribe were famous for their facial and body tattooing (Mooney, 1912, p. 947). Moreover, Bienville mentions a tattoo design identical with that of the “painted man” which was used by the Nakasa (a Caddoan group) : “all the savages here have a circle tattooed around the eyes and on the nose and three lines on the chin.” (MS. quoted by Swanton, 1942, p. 143.) The bow and arrow carried by the “archer” (pl. 53, a, 6) indicates that this weapon had not yet been supplanted by the rifle at the time the petroglyphs were made, while the horse indicates a date sometime after about 1750 for the work. Numerous grooves, apparently made by sharpening wooden or bone tools on the sandstone, appear at the south end of the slab. Two iso- lated grooves of this sort are also present between two of the anthro- pomorphic petroglyphs, the “painted man” and his neighbor (pl. 53, ¢). SPECIMEN DESCRIPTIONS Body sherds.—Seven body sherds were recovered from this site. Four are cord marked, with grit tempering; two are smooth, with shell tempering; and one is smooth with grit tempering. The four cord- marked sherds are typical of Aksarben sites (pl. 52, e’, 9’). They have buff exteriors and dark gray interiors. In thickness they range from 5to7mm. Three of these were found in the 0-foot to 0.5-foot level, and one in the 0.5-foot to 1.0-foot level. The two shell-tempered sherds (pl. 52, f’) vary in thickness from 5 to 8 mm. They are buff on the exterior surfaces, with interiors of dark gray to buff, flecked with white from the tempering material. In some places the tempering has leached away, leaving a pitted sur- Riv. Bas. Sur. Bab. No $8] TORONTO RESERVOIR AREA, KANS.—HOWARD 345 face. One of these sherds was found in the 0.5-foot to 1.0-foot level, the other in the 1.0-foot to 1.5-foot level. These sherds have been tentatively identified with the Cowley Plain type (Wedel, 1949, p. 87). The single smooth, grit-tempered sherd is light buff on its exterior surface, gray on the interior. Its thickness is 8.5 mm. The outer surface of this sherd seems to have been wiped with a bunch of grass when the vessel was ina “leather” state. It was recovered in the 0-foot to 0.5-foot level of the site. It has been tentatively identified with the Geneseo Plain type (Wedel, 1949, p. 88). Projectile points—Five complete projectile points and three frag- ments were recovered at the shelter. These are readily divisible into two groups on the basis of size. Two of the complete points and the three fragments are “large point” styles, while the remaining speci- mens are “small point” forms. All are manufactured of gray Flint Hills chert. The largest of the projectile points (pl. 52, m’) has an expanding stem with a convex base. It is 50 mm. in length, has a maximum width of 36 mm., and a maximum thickness of 9mm. This projectile point was found in the 0-foot to 0.5-foot level. The second large projectile point (pl. 52, 2’) also has an expanding stem, but the base isconcave. The shoulders, in the form of barbs, are carried back from the notches to the level of the base of the stem. This point style seems to be quite rare in this section of the country, judging from local col- lections. It is 39 mm. in length, has a maximum width of 30 mm., and a maximum thickness of 7 mm. It was recovered in the 0.5-foot to 1.0-foot level. One of the small points is of the plain, triangular, unnotched variety. It is 20 mm. in length, has a maximum width of 18 mm., and a maxi- mum thickness of 8 mm. (pl. 52,7’). Another is shouldered and has a straight stem with a convex base. It is 17 mm. in length, has a maximum width of 12 mm., and a maximum thickness of 3 mm. (pl. 52, k’). Both of these were recovered in the 0-foot to 0.5-foot level. A third small projectile point, recovered in the 1.5-foot to 2.0-foot level, has side notches and a basal notch. It is 20 mm. long and has a maxi- mum width of 12 mm., with a maximum thickness of 3.5 mm. (pl. 52, 7’). Drill—A beautifully chipped drill of pink and tan chert was recov- ered in the 0-foot to 0.5-foot level. It is 61 mm. in length, has a maximum width of 13 mm., and a maximum thickness of 8 mm. (pl. 52, 7’). End scrapers—Fragments of three end scrapers were recovered, all in the 0-foot to 0.5-foot level. An almost complete specimen meas- ures 40 mm. in length, has a maximum width of 20 mm., and a maxi- 346 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 mum thickness of 6 mm. This specimen is made of pink chert, the other two are of light-gray chert (pl. 52, ’). Flake scrapers——Two artifacts of this type were recovered, both in the uppermost level. The first has three edges that show secondary chipping. It measures 60 mm. by 30 mm., with a maximum thickness of 8mm. The second also has three working edges. It measures 57 mm. in length by 35 mm. maximum width, with a maximum thickness of 7mm. Both aremade of gray Flint Hills chert. Sharpening tool——A. sandstone sharpening tool, probably employed to sharpen and smooth wooden, bone, and antler implements, was found in the uppermost level of the site. It has smoothed faces on opposite sides, and a narrow groove toward one end. It is 74 mm. long, has a maximum width of 40 mm., and a maximum thickness of 11 mm. Cupstone.—A single cupstone, made of reddish sandstone, was found in the 0-foot to 0.5-foot level. It is irregular in shape, measuring 85 mm. by 85 mm., with a maximum thickness of 30mm. On one face there is a circular, pecked depression, 26 mm. in diameter and 5 mm. in maximum depth (pl. 52, 0’). Hoes.—One complete hoe and a fragment that apparently represents the blade of another were recovered. ‘The complete hoe is made of a grayish-green indurated sandstone. The broad edge of the tool has been sharpened by the removal of several large percussion flakes. The pointed end could have been hafted to a wooden handle. This object measures 160 mm. in length, 63 mm. in width at the widest part, and is 19 mm. thick. It was recovered in the 0-foot to 0.5-foot level (pl. 52, p’). The fragment is made of Cottonwood limestone. It is 56 mm. wide at the point of breakage and has a maximum thickness of 15 mm. It was recovered in the 0.5-foot to 1.0-foot level. Hammerstone.—A single hammerstone, consisting of a river pebble with signs of battering on several faces, was recovered in the 0-foot to 0.5-foot level. It is 50 mm. in length and 35 mm. in diameter at the center. Core—A gray chert core was recovered in the 0-foot to 0.5-foot level. It is 40 mm. in length, 32 mm. in width, and has a maximum thickness of 37 mm. CULTURAL AFFILIATIONS At least three, and perhaps four, components are present at this site. The petroglyphs represent the latest, and can be attributed to some historic tribe, since one of them depicts a man mounted on a horse. Smooth shell and grit-tempered sherds identify a component of the Great Bend Aspect. These sherds may well be associated with the Riv. Bas. Sur. eine Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 347 petroglyphs, as Great Bend materials often occur with European trade goods. Thin, cord-marked pottery identifies the next-oldest compo- nent, Aksarben. Large, expanding-stem projectile points may indi- cate a still earlier Woodland or Archaic occupation. Dry Creek RocxsHetter (14WO0224) This is a fairly large rockshelter in the valley of Dry Creek (fig. 59). This site, as well as site 14W 0226, is outside the reservoir area proper, but due to potential destruction by picnickers, was considered as a part of the salvage project at the time of the 1957 excavations in the Toronto Reservoir area. Low hills with grass cover and some scrub pine and blackjack oak surround it. The area of occupation consists of the entire floor of the shelter. Several petroglyphs are to be seen on a slab of rock toward the front of the shelter (pl. 55,a). The site is quite disturbed, as it has been a favorite picnic spot for local people since at least 1880. Apparently, many petroglyphs once present in the shelter have been completely obliterated by initials, dates, and other personal memorials left by visitors. Andreas’ “History of the State of Kansas,” published in 1883, men- tions the shelter and some of the petroglyphs visible at that time. Under the heading “A Prehistoric Cave,” appears the following description : This cave is situated about twelve miles north of Toronto, on Section 13, Town- ship 24, Range 14. Its mouth is about fifty feet wide and ten feet high, and the cave extends back about twenty feet. In the mouth of the cave lies a rock about nine feet long by six feet wide, the surface of which is nearly horizontal, the rock having evidently fallen from the roof of the cavern. On the surface of this rock are cut numerous figures of various sizes and shapes, some of which are in- describable. No system of regularity was observed by the inscribers, but the different figures and groups of incisions are scattered promiscuously, often overlapping and interlacing each other, as if done more for pastime than for the purpose of leaving any record of events then occurring, to be read by future generations. Some of the figures represent the human body, others parts of the body, as the head, with a small hat on, and marks down the chin, which may have been meant to represent the beard. One may have been designed to repre- sent a little idol, another a bird’s foot, another looks like a capital A, ete. Great interest is manifested in them by the people of Woodson County which is doubtless altogether owing to the fact of their mysteriousness. The same interest will probably always attach to them. There is but little reason to hope that they will be so deciphered as to throw any light on the history of the past. These tracings, or figures, or hieroglyphics, as some call them, were discovered about May 15, 1858, by Esquire Robert Daly, while out on a private hunting ex- pedition. At the time of discovery, they were covered over with dirt and debris, and partially overgrown with moss. Mr. Daly, who was one of the first settlers in this part of the county, has resided in the vicinity ever since, and now lives about one and a half miles south of this prehistoric cave. (Andreas, 1883, p. 1190.) 348 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 EXCAVATIONS Eleven 5-foot-square test pits were excavated in the floor of the cave (fig. 63). Each of these was carried down to bedrock. Arbitrary 0.5- foot levels were maintained throughout the work (pl. 54, a). It was found that the floor fill had been badly disturbed to a depth of from 0.7 - foot to1.1 feet. Since no diagnostic artifacts were recovered below this level, no cultural stratification could be obtained at the site. The fill is not more than 4.5 feet at the maximum. It consists of organic material combined with roof fall. A section from the deepest part of the fill, which is toward the front of the shelter and in the center, revealed five layers. The top stratum, which had an average thickness of 0.3 foot, consisted of a dark clay-humus. Underlying this was a layer of light-yellow sandy soil with an average thickness of 0.6 foot. This layer was underlain by a very thin stratum of light-brown sandy soil, 0.2 foot thick. The fourth layer consisted of orange-red sand with thin, angular fragments of sandstone. It was 0.6 foot thick. The fifth layer was gray clay, containing large fragments of slightly decomposed sandstone and small lenses of organic humus with a clay admixture. Ata depth of 1.3 feet the color of the clay changed to a deep gray-blue shade. PETROGLYPHS Of the several petroglyphs mentioned in the Andreas account, only two major examples remain intact. The first of these occurs on the upper surface of a large slab, undoubtedly the same one described in the Andreas account. It is incised in the rock and is apparently intended to represent a horse, since the tail is quite long (pl. 55, a). Below the head of the creature, which is square rather than elongated, there is a line through the neck. A bilobed arrow is incised above the horse’s back. The treatment of this horse reminds one of the representations of the mythical Underwater Panther in Plains and Woodland In- dian song pictographs. Here such a line through the neck indicates that the creature is emerging from beneath the earth (James, 1956, pp. 845-346). The second petroglyph occurs on the side of a large boulder a few feet south of that bearing the horse design. It seems to be a repre- sentation of some growing plant, such as corn, or grass (pl. 55, 6). Because this petroglyph is so different in style from the others and from those at other sites in the area, it may well be spurious. On the other hand, it is scarcely the sort of thing a White picnicker might dash off on a Sunday afternoon. No tracing was made of this petro- glyph because of the roughness of the stone on which it is inscribed. Of those petroglyphs described in the Andreas account that have since been destroyed, one seems to merit special mention. This is the TORONTO RESERVOIR AREA, KANS.—HOWARD 349 Riv. Bas. Sur. Pap. No. 38] *SUOI]VAVIXO SULMOYS “PZ7ZOMPT OS JO ue[G—'f9 AUNT] 1334S5—vASSLNI YNOLNOD 1333 YIOANSSSY OLNOYOL SNVHY3SA0 ani 39037 : YSLTSHS MOOU Wasayo AYd PeZOMUl SASAUNS NISVE YSAIY NOILALILSNI NVINOSHLIWS NYA HBV SITLL, HdATOOML Sd g2zOMel 350 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 representation of the human head “with a small hat on, and marks down the chin.” It sounds very much like a description a Victorian scholar might have made of a pictograph similar to those at the Wall- eye Rockshelter. The “small hat” may very well have been the rep- resentation of the roach hairdress and the marks on the chin, of tattooing. If this were so, this would tend to connect the two shelters and also to correlate the petroglyphs with the Great Bend sherds that are present in both. SPECIMEN DESCRIPTIONS Rim sherds.—Five rim sherds were recovered at the shelter. Three of these are of Aksarben affiliation, one is apparently of Great Bend affiliation, and one is of an indeterminate type. One of the Aksarben rims is quite large (82 by 66 mm.) and repre- sents the rim, neck, and a considerable portion of the body of the vessel. The rim is slightly flaring and undecorated except for the cord marking, and the lip is rounded. The exterior is buff in color; the interior is dark gray. Tempering consists of fine grit (pl. 54, ¢). The other two Aksarben sherds are similar, except that they seem to be straighter than the large rim and are gray on their exterior surfaces, rather than buff. The thickness of all three sherds is 5 mm. The Great Bend rim sherd is straight, and tapers from a thickness of 7 mm. at the lip to 10 mm. at a point 25 mm. below the lip. Al- though no tempering material is visible, numerous small holes indicate that shell temper was employed. The sherd is brown on its exterior surface and dark gray on the interior (pl. 54,5). It has been assigned to the Cowley Plain type (Wedel, 1949, p. 87). The remaining rim sherd is straight, with a rounded lip. The sherd has been polished on the outer surface. It is a light-brown color throughout. No tempering material can be discerned. The sherd is 4mm. thick. It does not seem to belong to any ceramic type known in the Central Plains, and may be from a trade vessel originating in the Caddoan area to the southeast. Body sherds.—Sixty-six Aksarben sherds, one Great Bend sherd, and one polished brown sherd of indeterminate affiliation were recov- ered. The Aksarben sherds are grit tempered, cord marked on the outer surface, smoothed on the inner surface, and vary in thickness from 6 to 9mm. In color they vary from buff to dark gray, with occasional firing clouds. On some sherds the cord marking has been almost completely smoothed away. The Great Bend sherd is shell tempered and smoothed, but not polished, on both the inner and outer surfaces. It is brownish gray in color and is 6 mm. thick. It has been assigned to the Cowley Plain type (Wedel, 1949, p. 87). Riv. Bas. Sur. — Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 351 The polished sherd is brown in color throughout. The nature of the tempering material, if any, cannot be ascertained. It is polished on the outer surface and smooth, but unpolished, on the inner surface. Its thickness is 5 mm. Projectile points ——The 13 projectile points from the shelter fall into three distinct groups. The first, represented by two fragments, is in the “large point” tradition (pl. 54, g, 2). These points have expanding stems. One has a convex base; the base of the other is straight. Maximum widths of the two points are 31 and 30 mm.; maximum thicknesses are 9 and 7 mm., respectively. The second point shows evidence of attempted rechipping before abandonment. Both points are made of gray Flint Hills chert that has white inclusions. The second style is represented by two specimens. These points are small, and resemble those recovered in Plains Woodland sites in Nebraska and Kansas. Points of this style were also recovered at sites 14WO215, 14GR210, and 14GR216. The first, nearly complete, has an expanding stem with a concave base. It measures 30 mm. in length by 12 mm. in width, and is 3 mm. thick (pl. 54, ¢). The other has an expanding stem with a straight base. Its edges are serrated. Jt is 138 mm. in width and 3 mm. thick (pl. 54, 7). The first is made of gray Flint Hills chert, the second of tan chert. The remaining group is made up of nine points, two of which are complete. They are in the small, triangular, “late horizon” style (pl. 54, c,d). All are unnotched. The two complete specimens meas- ure 19 and 24 mm. in length by 11 and 15 mm. in maximum width, with maximum thicknesses of 2 and 3 mm., respectively. Four of the points are of gray Flint Hills chert, three of cream-colored chert, one of pink chert, and one of brown jasper. In addition to the above specimens, eight additional projectile point fragments, not complete enough for meaningful description, were recovered. Four of these seem to have been in the “large point” tradi- tion, the remainder of the “small point” type. The materials are tan and gray Flint Hills cherts. Blades.—One complete and one fragmentary blade were recovered. The complete blade is small, triangular in shape, and only 45 mm. long. It hasa maximum width of 23 mm. and a maximum thickness of 7 mm. (pl. 54,7). The large blade appears to have been slightly wider in the midsection than at the base. The maximum width of the fragment, measured at the point of breakage, is 47 mm. Width at the base is 34 mm., and the maximum thickness is 5 mm. (pl. 54, 0). Knives.—Fragments of two lozenge-shaped, alternately beveled, Harahey knives were recovered at the shelter. The more complete specimen is made of light-gray chert, the other from a dark reddish- gray chert (pl. 54,5). 352 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Flake knives—Two of these artifacts, which seem to represent a rather rare type in this area, were found. Both have been made from long lamellar flakes, rectanguloid in outline, and with a prominent dorsal ridge (pl. 54,4). One has been retouched with secondary chip- ping along one cutting edge, the other has been sharpened along both sides. The first specimen, of gray chert, measures 60 mm. in length by 17 mm. maximum width. It has a maximum thickness of 4mm. The other is made of tan chert. It is 61 mm. long, has a maximum width of 17 mm. and a maximum thickness of 8 mm. DrillA single fragmentary drill was found at the shelter. It is of the expanding-base type, and made of gray Flint Hills chert. The base is 20 mm. wide and the blade is 6 mm. wide at the point of break- age. Maximum thickness of the base is 3.5 mm. and maximum thick- ness of the blade is 2.5 mm. (pl. 54, m). End scrapers.—Fifteen artifacts of this type were found, repre- senting ten complete scrapers and five fragments. "Workmanship varies from fine to exceedingly crude. All of the scrapers are sub- triangular in shape. The complete specimens range in length from 46 mm. to 25 mm., and in maximum width from 27 to 15 mm. The maximum thickness of the largest is 14 mm., that of the smallest is 5 mm. Materials employed are tan jasper, a translucent brown chalcedony, and white, pink, and gray Flint Hills cherts (pl. 54, ty J, 1). Flake scrapers—Thirty-eight flake scrapers were recovered. They exhibit no uniformity in size or shape and are distinguished merely by one or more edges that show secondary chipping. Materials are cream-colored, tan, pink, and light and dark gray cherts. Choppers—F¥our choppers were recovered at the site. These are large, heavy pieces of gray Flint Hills chert worked to an edge by per- cussion flaking on both faces. Though quite irregular, they tend to be ovoid in shape. The largest specimen is 92 mm. long and has a maxi- mum thickness of 22 mm. The smallest is 49 mm. long and has a maximum thickness of 14 mm. Shaft smoother—tThe only shaft smoother recovered is made of sandstone (pl. 54,7). It bears a single straight groove along one face that was very likely employed to smooth arrowshafts. It measures 77 mm. in length by 38 mm. in width and is 27 mm. thick. The groove would accommodate a shaft 7 mm. in diameter. Sharpening stone.—This specimen is similar to that just described, but has semiconical grooves on three faces that seem to have been used to sharpen wooden, bone, or antler tools. It is of sandstone and meas- ures 57 mm. in length by 20 mm. in width and has a maximum thick- ness of 18 mm. Bone needles—Two bone needles were recovered in the shelter. The first represents the tip and a portion of the shank. It has been Riv. Bas. Sur. Riv. Bas. «= TORONTO RESERVOIR AREA, KANS.—HOWARD 353 worked from a section of mammal long bone. It is 43 mm. in length and 3.5 mm. in diameter (pl. 54, p). The second fragment represents a part of the shank of a much larger object, with a diameter of 7 mm. It has also been worked from a section of mammal, probably deer, leg bone. Bone awls.—Three bone awls were recovered. All are made from mammal long bones. These objects differ from the needles just de- scribed in that their shanks are much wider than their points. Lengths of the three objects are 55 mm., 52 mm., and 33 mm. Maximum widths are 14 mm., 13 mm., and 10 mm. Bone bead.—A tubular, bone bead is made from a section of the long bone of some small bird or mammal. It is 41 mm. in length and 8 mm. in diameter. Miscellaneous worked bone——Three pieces of bone, bearing recog- nizable tool marks, were recovered. They have no recognizable form and are probably scraps left after the manufacture of other objects. Metal objects—Two metal objects were recovered in the shelter fill. One is a fragment of a small iron kettle (pl. 54,7). It has a rather elaborate design on its outer surface and on the inner surface a raised numeral “2”. The second piece seems to be part of the handle of a brass spoon. Both objects may be assigned to either the historic Indian occupation of the site or to subsequent use of the shelter by White picnickers. CULTURAL AFFILIATIONS At least two, and perhaps three, components are represented at the site: Aksarben, Great Bend, and historic Indian. The first two are identified by ceramic materials. The last, which may prove to be identical with Great Bend, is identified by the petroglyph of a horse. Tuer Outtaw RocksHELTeR (14WO225) This site is a very small rockshelter with a petroglyph on the rear wall. The name derives from a local tradition that this shelter was used by one of the Dalton gang or Jesse James as a hideout. The shelter is very shallow, low-ceilinged, and damp. The petroglyph is possibly a representation of an Indian dwelling covered with grass thatching, pieces of bark, or hides. According to local tradition, the initials “N.C.” were added by a man named Nor- man Curtis, about 30 years ago, “to fool people” (pl. 55, c). Lrrrin-Water-Man Prrroeiypn (14W0226) This site consists of a single anthropomorphic petroglyph deeply incised in the rocky outcrop along Dry Creek about 300 feet upstream from the Dry Creek Rockshelter (14W0O224). 354 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 The name of the petroglyph was given by Mr. D. Wm. Chatfield, a member of the 1957 Smithsonian party, who discovered and reported it. According to Chatfield, who is an Ojibwa Indian of Cass Lake, Minn., the pictograph closely resembles Ojibwa drawings of a super- natural being called the “Little-water-man.” The figure holds an object, perhaps a banner, in his left hand (pl. 55, d, e). Tue Possum Point Sire (14W0228) This is a small occupation site which was apparently used for only a short time. It is located on a series of small, low knolls, bounded on the northeast and west by Finger Creek and by scrub growth and sandstone outcrops on the south (fig. 59). The area of occupation is approximately 1 acre in extent, and there seems to be a slightly greater concentration of material at the west end of the site (fig. 64). The site area has apparently been under cultivation for some time, and no surface features were visible. The topsoil layer, a dark, sandy loam, is from 0.8 to 0.4 foot in depth (pl. 56, a). Beneath this is a lighter-colored soil, perhaps an old erosion surface, of light-brown to yellow soil, probably derived from sandstone, which is from 1.5 to 2.0 feet in depth. A hard clay is found below this, with sandstone blocks at the top. There is evidence of plow disturbance to a depth of 0.8 foot. EXCAVATIONS A test trench was laid out in the area of greatest surface concentra- tion. This trench was 20 feet long and 5 feet wide, divided into four 5-foot squares. Later, following out the area of surface concentration, two additional squares were excavated to the east of this trench and adjoining it. Heavy orange-buff potsherds, projectile points, and chips were re- covered from the surface to a depth of 0.5 foot, where they thinned out and disappeared. No cultural stratification was found at the site. SPECIMEN DESCRIPTIONS Rim sherds—Seven rim sherds were recovered at the site. They represent at least four different vessels. All have been assigned to the Geneseo Plain type, although they differ slightly in color and seem to be, on the average, slightly thicker than those described by Wedel. Tempering material consists of large sand particles, ranging from 1 to 3 mm. in diameter. Hardness varies between 3.5 (celestite) and 4 (fluorite). The core color is buff to slate-gray, the exteriors orange buff to light gray or brown. There are occasional firing clouds. Sur- TORONTO RESERVOIR AREA, KANS.—HOWARD 5 H5y5) Riv. Bas. Sur. Pap. No. 38] *sUOI]BALIXS SUIMOYS “BZ7ZQO MPI 2S JO Ue[G—'F9 TUNITY Lie 4430709 ONY QHuO3 1334 S—1VAY3LNI HNOLNOD 4334 [a cs ee | 06 [“-) YIOAYSSSY OLNOYOL SLIS LNIOd WNSSOd 8dcOMMI SASAUNS NISVE YSAIY NOILALILSNI NYVINOSHLIWS —— OLNOHOL OL 661-932—64——26 306 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 faces are generally smooth. A few sherds show evidence of having been wiped with grass before firing. Rims are straight or slightly flaring. The lip is usually rounded, but in one example is flattened. Shallow, vertical indentations are present on the outer lip of five of the sherds (pl. 56, g, h, 7, &); the other two are undecorated (p. 56,2). One of the plain sherds tapers from a thickness of only 4 mm. at the lip to a thickness of 10 mm. at a point 50 mm. below the lip. The others have a more-or-less uniform thickness of from 7 to 9 mm. Body sherds.—A total of 55 body sherds was recovered, all either on the surface or in the 0-foot to 0.5-foot level. In thickness they range from 6 to 15 mm., averaging about 9 mm. It is likely that the thicker sherds are from the bottoms of the vessels. All but one of these sherds are undecorated. The decorated sherd displays two horizontal bands, 25 mm. apart, on its outer surface. The decoration consists of a shallow incised, herringbone pattern. An identical decoration is found on a sherd of the Neosho Focus of Oklahoma figured by Bell and Baerreis (1951, pl. 18,1). Projectile points —Four projectile points were found at the site. Three of these are in the “large point” tradition, while the fourth isa small “late horizon” form. The only complete large point is a rather blunt specimen (pl. 56, d). The stem is of the expanding type and has a convex base. It is 49 mm. long, with a maximum width of 34 mm. and a maximum thick- ness of 8 mm. It is made of cream-colored chert. Another of the large points (pl. 56, e), complete except for the tip, is in the same style, though not as blunt. It has a maximum width of 33 mm. and a maximum thickness of 7mm. When complete, it was probably about 60mm.long. This point is made of gray Flint Hillschert. The third large point lacks most of the stem (pl. 56, ¢), but enough remains to indicate that it was of the same type as the preceding ones. It has a maximum width of 24 mm., a maximum thickness of 7 mm., and was probably 35 mm. long when complete. It is made of light-gray chert. The small point (pl. 56, 6) is triangular, unnotched, and slightly concave at the base. It is 29 mm. long, 13 mm. wide at the base, and has a maximum thickness of 8 mm. Its material is the same cream-colored chert as the first large point described above. Blade——The base of a long, chipped blade was recovered. It is made of a pinkish-gray chert. Its base is rounded, and the piece is worked on both sides. It has a maximum width of 41 mm., measured at the point of breakage, and a maximum thickness of 10 mm. Abrader.—A. fragment of a sandstone abrader was found on the surface of the site. Both ends are missing. It is rectangular in cross section and measures 35 mm. by 30 mm. Riv. Bas. Sur. r ply, Bas. =~ TORONTO RESERVOIR AREA, KANS.—HOWARD 357 Cupstone.—A. single cupstone, with pits on opposite faces, was found at the site. The material is a reddish sandstone. It is 104 mm. in length, with a maximum width of 63 mm. and an average thickness of 833 mm. The two pits measured 31 and 20 mm. in diam- eter and are pecked to depths of 8 and 4 mm., respectively (pl. 56, f). Hoe—A rather crude hoe, made of Cottonwood limestone, was recovered. It is flat on one surface, rounded on the other, and has been sharpened at one end. It is 154 mm. in length, 70 mm. in width, and has a maximum thickness of 24 mm. Flake.—A very large gray chert flake was found on the surface of the site. One edge shows some secondary chipping, and perhaps the piece served as a scraper. It is 94 mm. long, has a maximum width of 77 mm., and a maximum thickness of 23 mm. CULTURAL AFFILIATIONS Heavy orange-buff potsherds and small triangular projectile points identify this site with the Great Bend Aspect, previously designated the Paint Creek Culture, and hitherto known only from central Kansas (Wedel, 1940, pp. 332-334; 1942, p. 10; 1949, pp. 86-90; Smith, 1949, pp. 292-800). As noted earlier in this paper, this may represent the late prehistoric and early historic Wichita tribe. Apparently very closely related is the Neosho Focus of northeastern Oklahoma (Bell and Baerreis, 1951, pp. 71-75). Possibly a second and much earlier Woodland or Archaic Component is represented by the large expand- ing-stem projectile points found at the site. Strmr 14GR210 This site is located on a slight rise on the south side of Walnut Creek, one of the main tributaries of the Verdigris in this area (fig. 59). The area of occupation is quite extensive, measuring approximately 1,000 feet in length (NW-SE) by 300 feet in width (NE-SW). The site was under cultivation in 1957 and has been for a number of years (pl. 57, a). EXCAVATIONS The site was tested by means of four 5-foot-square test pits, which were placed in areas of greatest surface concentration (fig. 65). The first two of these were excavated to a depth of 3.0 feet, the other two to a depth of 2.5 feet. Results were disappointing. The only artifact recovered from any of these pits was a fragmentary projectile point in the 0-foot to 0.5-foot level of test pit 2. The fill was a thin layer of clay-loam topsoil which graded into a heavy clay at depths varying from 0.3 to 0.7 foot. 358 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 SPECIMEN DESCRIPTIONS Projectile points.—A. total of 13 projectile points sufficiently com- plete to permit classification and fragments of 18 others were re- covered. The more complete specimens fall into three styles, two of the “large point” tradition and one of the “small point” tradition. The first style is represented by two points, one of which is com- plete (pl. 57, c, 4). These points have contracting stems and are quite thick. The complete specimen is 61 mm. in length. Maximum widths of the points are 25 and 27 mm., and their maximum thicknesses 12 and9mm. Both are made of gray Flint Hills chert. The second style is represented by nine fragmentary specimens (pl. 57, d,g). These points have expanding stems and pronounced, some- times barbed, shoulders. Three of the points have convex bases, two have straight bases, and two have concave bases. The bases are missing from the remaining examples. Lengths of the two nearly complete specimens are 40 and 30 mm., widths 28 and 24 mm., and maximum thicknesses 7 and 4mm., respectively. One of the fragments represents a point much larger than the rest, perhaps a knife or spear- point rather than a projectile point (pl. 57,2). It is 43 mm. in width and probably measured at least 95 mm. in length when complete. Materials are white, tan, pink, and gray cherts. The large point is of gray Flint Hills chert. The third point style is represented by only two specimens (pl. 57, é, f). Both have expanding stems with serrated edges. The complete specimen measures 23 mm. in length, and is 12 mm. wide at the base, with a maximum thickness of 4 mm. It is made of tan chert. The second specimen, a fragment, is 2mm. thick. It is of gray Flint Hills chert. As noted elsewhere in this paper, points of this style have been recovered in Plains Woodland sites in Nebraska and Kansas. Blades —Fragments of eight blades were recovered. Although none is sufficiently complete to indicate the style, it is possible that they may have been pyriform in shape. White, cream-colored, tan, and gray Flint Hills cherts are the materials. Knife—One nearly complete beveled knife was recovered (pl. 57, 6). It is straight along one edge, shouldered on the other. The straight edge has been chipped from one side of the piece, the shouldered edge from the other. The stem may have been utilized for hafting. The knife measures 54 mm. in length, has a maximum width of 17 mm., and a maximum thickness of 6 mm. End scrapers—Seven end scrapers were found. They are all of the familiar planoconvex type, with a rather abruptly chipped edge at the thick end. Three are of cream-colored chert and four are of gray Flint Hills chert. Lengths vary from 40 to 68 mm., widths from 28 to 48 mm., and maximum thicknesses from 11 to 16 mm. TORONTO RESERVOIR AREA, KANS.—HOWARD 359 Riv. Bas. Sur. Pap. No. 38] *SUOT]LALDX SUIMOYS ‘OTZYOFT OS JO ue[g—*sg TUN 334 > —TVANSLNI YMOLNOS 1333 09 o Ce) YIOAYSSSY OLNOYOL Oley vl SASAYNS NISWE USAIN NOJLALILSNI NVINOSHLIWS & Lid Is3 7 4 360 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Flake scrapers.—There were 18 objects of this sort found, all with one or more edges showing secondary chipping. They exhibit no uniformity in either size or shape. Materials are tan and gray Flint Hills cherts (pl. 57, 7-0). Chopper.—One heavy chopper, manufactured of gray Flint Hills chert with white inclusions, was found. It is pyriform in shape, and shows no evidence of secondary chipping. It measures 118 mm. in length and has a maximum width of 65 mm. and a maximum thickness of 35 mm. Manos.—Seven manos were found at the site (pl. 57, m). All are made of local sandstone. Six have two abrading surfaces, the re- maining specimen only one. The bifacial specimens are subrectangu- lar in shape, while the unifacial specimen is loaf shaped. Lengths of the complete specimens vary from 15 to 83 mm., widths from 76 to 104 mm., and thicknesses from 39 to 50 mm. The unifacial specimen has a maximum thickness of 83 mm. Hammerstones—Three hammerstones, two made from river peb- bles and one from gray Flint Hills chert, were recovered. All are of a size to fit conveniently in the hand, and all show evidence of bat- tering on at least two surfaces. Unidentified objects—Two unidentified objects may be manos or abrading tools. Both are made of Cottonwood limestone. One is rectangular when viewed from above, with the outline of a parallelo- gram when viewed from the side. It is 148 mm. in length, 75 mm. in width, and 23 mm. thick. The other piece is wedge shaped if viewed from the side, ovoid if viewed from above. It has a sharp edge at one end, perhaps used for cutting. It measures 87 mm. in length and 73 mm. in width, and has a maximum thickness of 22 mm. Hematite—Two pieces of soft hematite show evidence of having been rubbed to produce paint. The smaller of the two pieces has been rubbed on three surfaces. The pieces measure 62 and 31 mm. in length, 52 and 30 mm. in width, and 29 and 9 mm. in maximum thickness, respectively. CULTURAL AFFILIATION Artifacts recovered from the site suggest that it was occupied dur- ing the Archaic period. Projectile points and blade fragments from 14GR210 bear a closer resemblance to artifacts of the Archaic of the Southeast than to those of a corresponding time period further north and west in the Plains. Srtre 14GR216 This is a rather extensive site located on a slight rise on the west side of the Verdigris River (fig. 59). Brazell Creek flows into the Riv. Bas. Sur. Bap. No 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 361 Verdigris River just north of the site on the opposite side of the river. The area of occupation is approximately 1,000 feet in length (NW-SE) by 350 feet in width (NE-SW). There seems to be a greater concentration of artifacts at the northwest end of the site (fig. 66). Although the site has been plowed for some time, it is nevertheless quite rich in surface materials. EXCAVATIONS A trench 65 feet in length and 5 feet in width was run through the area of greatest surface concentration at the north end of the site. This was marked off into thirteen 5-foot squares. Near the center of this trench, part of a circular hearth was uncovered at a depth of 0.5 foot (pl. 58,a@). The trench was then widened to 15 feet in this area in order to expose the remainder of the hearth and to locate possible post molds or other features. Although this widened section, 15 feet square, was shaved down to a depth of 2.5 feet, no additional features were discovered. The remainder of the trench was excavated to a depth of 1.5 feet. Since no cultural stratification could be found, material from the surface of the site has been grouped with that from the excavations for purposes of description. FEATURES Hearth.—At a depth of 0.5 foot from the surface, a circular hearth was discovered. It measured 2.0 feet in diameter and had a maximum depth of 0.3 foot at the center. The fill was white ash with occasional lumps of charcoal. Beneath the hearth was a lens of burned earth approximately 0.5 foot in thickness at the center and tapering to 0.2 or 0.3 foot at the outer perimeters. The hearth was lined with limestone slabs, some of which were set on edge, while others were laid flat. Although fiecks of charcoal and bits of burned earth appeared throughout the test trench at approximately the same level as the hearth, no additional features appeared. If this hearth was the central fireplace of a dwelling, the structure must have been of a rather temporary sort. Associated with this hearth were three artifacts: a bone awl, a sec- tion of antler (possibly a flint-working tool), and a small bone bead. All of these objects seem to have been preserved by the ash of the hearth, and were the only objects of bone or antler secured at the site. SPECIMEN DESCRIPTIONS Rim sherds.—The three rim sherds recovered were apparently from vessels of Aksarben affiliation. They are tempered with fine grit. [Bull. 189 BUREAU OF AMERICAN ETHNOLOGY 362 *SUOIBAVIXS SUIMOYS ‘Q[ZYOFI os Jo ue[g—"99 AUNDI,J HOLIG J9UNIVYO #S AVMHSIH ‘SN Q10 HOLIO JOVNIVEC 234 S—WANSLNI YNOLNOD 1333 oe OO OF YIOAYASSY OLNOYOL 912 HOH} SASAUNS NiSVa YSATY NOILALILSNI NVINOSHLIWS NO|LVYLNS9NO9 SQ VWauV Pete Tae - ae ‘. ‘. ~ ss. OO Riv. Bas. Sur. kilos Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 363 The first is large enough to indicate that it came from a vessel with a constricted neck (pl. 58,0). The rim is straight, with a flat lip. The outer surface bears vertical cord impressions, the inner surface is smooth. The color is dark brown throughout. Thickness is 5 mm. at the lip and 8 mm. at the point of breakage. The second sherd is also straight, with a rounded lip (pl. 58, n). Decoration consists of a series of parallel, horizontal, incised lines and notching on the outside of the lip. The color is dark brown throughout. The lip thickness is 3mm. and the lower rim thickness, at the point of breakage, is 4 mm. The third rim is too fragmentary to be described meaningfully. It is orange buff in color on its exterior surface, dark brown on the interior. Body sherds——A total of 15 body sherds was recovered. Five of these are much thicker than the rest and have coarser cord marking. They have been assigned to the Plains Woodland component of the site. The remainder are of the Aksarben Complex. The five Woodland sherds are grit tempered, cord marked on the outer surface, and smooth on the inner surface. They have buff exteriors, with blue-gray interiors flecked with white specks. These specks are probably the remains of small fossils present in the clay at the time it was secured, although at first glance they might be taken for tempering material. Thicknesses of the sherds vary between 8 and 9 mm. (pl. 58,7, 8). The Aksarben sherds are also grit tempered. In color they vary from buff to black. They are cord marked on the outer surface, smooth on the inner. Thicknesses range from 4 to 7 mm. (pl. 58, p,q). Projectile points——Twenty-two projectile points sufficiently com- plete for classification were recovered at the site, as well as twenty-five unclassifiable fragments, such as tips or the section between the tip and the shoulders. Those points that are classifiable fall into five distinct styles. The first style, represented by six specimens, is a large, thick form that has been assigned to the Archaic component of the site (pl. 58, b-d). Shoulders are not pronounced and stems are contracting, straight, or slightly convex at the base. The only complete specimen is of tan chert. It is 50 mm. long, with a maximum width of 23 mm. and a maximum thickness of 8 mm. Maximum widths of the other points are 19, 18, 21, 380, and 29 mm., and maximum thicknesses range between 7 and 9mm. Materials are dark gray chert and tan chert. Points of the second style, represented by 10 specimens, have straight stems and definite shoulders, sometimes terminating in barbs (pl. 58, e, 7). The three complete specimens measure 58, 55, and 43 mm. in length. Maximum widths vary between 20 and 32 364 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 mm., and maximum thicknesses between 3 and 8mm. Materials are white, tan, and gray cherts, and quartzite. The third style is represented by a single complete specimen (pl. 58, g). This point has an expanding stem with a convex base. It is 47 mm. long, has a maximum width of 22 mm., and a maximum thick- nessof7mm. The material is tan chert. The fourth style is represented by two specimens, one of which is incomplete (pl. 58, h, 7). These points are small, with expanding stems and either straight or convex bases. The incomplete speci- men has a serrated blade. Length of the complete point is 20 mm., and its maximum width 11 mm. The serrated point is 9 mm. wide. Both points have maximum thicknesses of 3mm. Materials are gray Flint Hills chert and tan chert. These two points may, with some assurance, be assigned to the Plains Woodland component of the site, and suggest an identity with the Keith Focus (Marvin F. Kivett, personal communication, 1957). The fifth style, represented by three specimens, is the “late horizon” triangular form (pl. 58, 7-2). Two of the points are unnotched, while a third, probably a reworked specimen, has four side notches, two to a side, plus a basal notch. Length of the only complete unnotched point is 16 mm., and its maximum widthis9 mm. It is 2 mm. thick. The incomplete unnotched point is 15 mm. wide at the base and 4 mm. thick. The notched specimen is 14 mm. long, 16 mm. wide at the base, and 2 mm. thick. Materials of these points are tan and gray cherts. Spatulate object—This small object, which may be a knife, is manu- factured of gray Flint Hills chert. The base is rounded, but the ex- act shape of the complete specimen is not known, since a large por- tion of the tip is missing. It is very neatly flaked, has a maximum width of 17 mm. and a maximum thickness of 8 mm. Blades.—Portions of 14 blades were recovered. Judging from the larger fragments, they were pyriform in shape. Maximum widths range between 26 and 37 mm., thicknesses between 9 and 15 mm. Materials are cream-colored, tan, and gray Flint Hills cherts (pl. 58, v). Turtleback scrapers——¥our objects of this sort were recovered. Two are flat on one side and rounded by chipping on the upper sur- face. The other two have been chipped on both faces, but are much flatter on one face than on the other. The first specimen measures 54 mm. in length, has a maximum width of 30 mm., and a maximum thick- ness of 12mm. The second (pl. 58, uw), a fragment from one end of a scraper, has a maximum width of 18 mm. and a maximum thickness of 7 mm. They are of tan and cream-colored cherts, respectively. The larger of the bifacial specimens is 50 mm. in length, 41 mm. in Riv. Bas. Sur. vans Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 365 width, and has a maximum thickness of 20 mm. It is manufactured of gray-brown chert. The smaller specimen is 68 mm. in length, 26 mm. in width, and has a maximum thickness of 16 mm. It is of tan chert. Flake scrapers.—Fourteen artifacts of this description were found at the site. They are distinguished by one or more edges that show secondary chipping. There seems to be no uniformity in size or shape. Materials employed are tan and gray chert and, in one in- stance, red jasper. Choppers.—Hight specimens assigned to this category were re- covered at the site. Six are of tan chert and two of gray Flint Hills material. ‘They tend to be ovoid in shape and are worked to a cutting edge by percussion flaking on two faces. The largest specimen is 83 mm. in length and 20 mm. in maximum thickness. The smallest is 50 mm. long and has a maximum thickness of 13 mm. Celt.—One celt was recovered. It is made of Cottonwood lime- stone and measures 94 mm. in length, 52 mm. in width, and 17 mm. in thickness. The bit shows considerable wear. Aw.—One chipped ax of Cottonwood limestone was recovered. It was apparently double-bitted originally, but one end has been broken off. The bit measures 97 mm. in width, the groove 74mm. It has a maximum thickness of 31mm. (pl. 58, w). Pick.—tThe site yielded one pick, also made of Cottonwood lime- stone. The bit is slightly curved. The pick is 114 mm. in length, has a maximum width of 69 mm., and a maximum thickness of 35 mm. Manos.—Seven manos were recovered, six of which were fragments. The fragments are all of sandstone. When complete, the sandstone manos were probably subrectangular in shape, with two smooth grind- ing surfaces. The only complete specimen is of Cottonwood lime- stone. It has the outline of a parallelogram when viewed from above. It is 183 mm. long, 59 mm. wide, and 23 mm. thick. Cupstone.—A single cupstone was recovered (pl. 58,¢). When com- plete it was apparently ovoid, with two smooth faces on opposite sides. In the center of each of these faces is a small depression. The piece measures 74 mm. in diameter and has an average thickness of 43 mm. The depressions are 22 and 27 mm. in diameter, and 8 and 4 mm. in depth. Hammerstones—Four battered hammerstones were recovered. Three are of tan jasper, one of reddish-brown jasper. They are of a size to fit conveniently in the hand and show no evidence of artificial shaping. Core.—Only one core was recovered. It is made of a rather gran- ular gray chert. 366 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 Paint.—The site yielded two paint stones. One is yellow limonite, the other reddish-brown hematite. Both have been rubbed and scratched to secure pigment. The limonite specimen is 32 mm. in length and the hematite specimen is 24 mm. in length. Awl.—A single bone awl, worked from a deer metapodial, was re- covered (pl. 58, m). It rested on the surface of the hearth, Feature 1. Itis74mm. long and has a maximum width of 15 mm. Flaking tool—A section of antler, possibly a flaking tool, in a very poor state of preservation, was also found on top of the hearth. It measured 90 mm. in length and 15 mm. in diameter. The specimen crumbled when removal was attempted. Bone bead.—A small tube of bird or small mammal bone was prob- ably usedasabead. Itis15 mm. long and 3.5 mm. in diameter. This, too, was recovered in the hearth ash. CULTURAL AFFILIATIONS On the basis of typology, three components have been isolated at this site. In probable order of sequence, late to early, they are: Aksarben; Plains Woodland, perhaps the Keith Focus of the Orleans Aspect; and Archaic. SUMMARY AND CONCLUSIONS The archeological investigations at the 10 sites in the Toronto Res- ervoir area that have been reported in this paper reveal an aboriginal occupation of the locality beginning about 5000 B.C. and extending into the historic period. Although no Paleo-Indian material was re- covered by the 1957 party, the presence of many Paleo-Indian points in local collections indicates that worthwhile sites of this period may be present somewhere in the vicinity. The materials described are attributable to five cultural groupings, each on aslightly different time level, as follows: (1) The Archaic Culture (ca. 5000 B.C. to 200 B.C.) is represented by site 14GR210 and by components at 14W 0209, 14WO215, the Wall- eye Rockshelter (14W0O222), the Dry Creek Rockshelter (14W0O224), and 14GR216, with a possible representation at the Possum Point site (14W0228). At the multicomponent sites the Archaic components have been defined on admittedly shaky evidence, namely by the pres- ence of large projectile points and by the absence of ceramic materials that would identify these points with the cultural complexes. At 14GR210, which appears to be predominantly Archaic, large unnotched and side-notched points are accompanied by large chipped blades, a beveled knife, end scrapers, flake scrapers, a chopper, and manos. The inventory suggests that these people were hunters and gatherers of BD. ae: Sv TORONTO RESERVOIR AREA, KANS.—HOWARD 367 wild food. They lacked horticulture, as well as the bow and arrow and the ceramic arts. (2) The Kansas City Hopewell Culture (ca. 200 B.C.—A.D. 500) is represented by one site, 14WO203. It is identified by smoothed, grit- tempered pottery, decorated with dentate- and rocker-stamping, to- gether with large, expanding-stem projectile points, drills, gravers, pyriform blades, and scrapers, flake scrapers, choppers, and pieces of worked hematite. The Kansas City Hopewell people practiced maize horticulture and made pottery but were apparently not familiar with the bow and arrow. Settlements were larger and more permanent than in Archaic times. (3) The Plains Woodland Culture (ca. A.D. 500-900) is represent- ed at one multicomponent site, 14GR216, where it was identified by five thick, cord-marked, grit-tempered sherds. Small expanding-stem points, with or without serrated blades, are also assignable to this com- ponent of the site, and suggest specific identification with the Keith Focus of the Orleans Aspect. Since 14GR216 is an unstratified multi- component site, it is not known which of the other artifacts recovered, such as scrapers, choppers, manos, etc., should be assigned to the Plains Woodland component. The Plains Woodland people (representing the Middle Woodland time level) possessed a culture very similar to that of the Kansas City Hopewell groups, and the artifacts of the two cultures show a close relationship. Like the Hopewell people, the Plains Woodland groups practiced horticulture, supplementing the food secured in this manner with wild vegetal products and game. The presence of both large and small projectile points at Plains Woodland sites indicates that both the bow and arrow and the atlatl were in use. (4) Components assignable to the Aksarben Aspect (ca. A.D. 1100- 1500) were present at five sites, all of which had more than one com- ponent. These are: 14W0O209, 14WO215, the Walleye Rockshelter (14W 0222), the Dry Creek Rockshelter (14W0224), and 14GR216. The Aksarben components are identified by moderately thin, cord- marked, grit-tempered pottery with either plain flaring or incised collared rims. At site 14GR216 the small side-notched and unnotched projectile points can be assigned to this component, which is the latest occupation present there. The Aksarben people, like the Kan- sas City Hopewell and Plains Woodland folk, were dependent upon hunting, gathering, and maize horticulture. Villages tended to be larger and more numerous than in earlier periods. Apparently the bow and arrow had completely supplanted the atlatl. (5) Four sites had components of the Great Bend Aspect (ca. AD. 1500-1700). These are: 14WO215; the Walleye Rockshelter (14WO 222) ; the Dry Creek Rockshelter (14WO224) ; and the Possum Point 368 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 site (14W0O228), with a possible component at 14WO209. The Possum Point site seems to be predominantly Great Bend. Compo- nents of this aspect were identified by heavy, smooth, grit- or shell- tempered pottery, gray, buff, or orange-buff in color. The presence of this material in the Toronto Reservoir area extends the known distribution of this aspect, hitherto known only from sites in Central Kansas. The Great Bend people continued the hunting, gathering, and maize horticulture subsistence pattern established and developed in the area by the earlier Kansas City Hopewell, Plains Woodland, and Aksarben people. Great Bend sites are apparently smaller than those of the Aksarben Aspect, a fact which may reflect the unstable conditions that seem to have characterized this area in early historic times. APPENDIX TABULATION OF TORONTO RESERVOIR SITES Site Type Culture Fieldwork 14GR1 Minor camp-_-_-__-- ? Visited. 14GR2 Minor village _-_-__- Upper Republican_.--__- Do. 14GR3 Minor camp-_-_---- ne Do. PAGE oO ete nS ote) Si es ? Do. 14GR202 Village-camp____-- Kansas City Hopewell Do. and Keith Focus. 14GR203 Minor camp---_-_-- i Do. 14GR204_ |, |_.-_- ro es Bae ? Do. 14GR205 Nallaoeo ss a. eee Upper Republican--_-- Do. 14GR206 Minor camp------ ? Do. 14GR207 jesse oz 2h). nen as ? Do. TAG R208 py) foe geet Fit ? Do. 14GR209 Walleme A?) 2 ? Tested in 1956. 14GR210 Major camp--_---- Archaic Complex_-_--_-- Tested in 1957. 14GR211 Minor camp-_-_-__-_- ? Visited. 14GR212 Village (?) Kansas City Hopewell - Do. 14GR213 Minor camp--_-_.-- ? Do. 14GR214—sJ____e 5.) a ee Do. TAG R21G ii o) acne co RE es ? Do. 14GR216 Village-camp-__--- Aksarben, Keith Focus, | Excavated in 1957. and Archaic. 14GR217 Minor camp-_-_-_-_-_- Keith Focus__-.------ Visited. LAG R218 pool oesee rt Ps ere v Do. 14GR219 Minor village___-- Woodland and a later Do. component. 14GR220 Minor camp---_-_-_- 4 Do. 14GR221_—sj__--- davtve sy: ais Woodland <1tc1o0 368 Do. 14WOLS (a tlesese dak. euihies JS igo Gaels WI b P's pres ae Tested in 1956. BAW? poset ASE at AG aw Jaa ? Visited. 1l4WOZ0K jet eee UOss. ae eee is Do. 14W0202_sjse---_-- 2 a eae ae ? Do. Riv. Bas. Sur. Pap. No. 38] TORONTO RESERVOIR AREA, KANS.—HOWARD 369 TABULATION OF TORONTO RESERVOIR SITES—Continued Site Type Culture Fieldwork 14W0203 Villages eS Kansas City Hopewell | Tested in 1956 and 1957. 14W0204 Rockshelter___---- ? Visited. 14W0205 Minor camp------ ? Do. 14W0O206_sf_-_--- Cc (oa es ly Bee ? Do. 14WO207_~—Ss_---- dos 2. ieee ? Do. 14WO0208_~si|__--- C0 ra pe fei fy Do. 14W0209 Village-camp -___-- Aksarben and Archaic_ | Tested in 1957. 14W0210 Minor camp-_----- ? Visited. TAWORZTL Yee 2 Co (aera eee ? Do. VAW O22 © E22 e" (3 (oe eS ees is Do. 14 WO2TS .. L222 doses. eA 3522)) Woodlands en et yaa oe Do. 14WO214 = |____- 0 £0 a pee Dit reetopeley ? Do. 14W0O215 Village-camp-_---_-- Great Bend, Aksarben, | Tested in 1957. and Archaic. 14W0O216 Minor camp------ ? Visited. 14W0217 Mound nie 20. ? Do. 14W0218 Minor camp-_.----- 6 Do. 14W0219 Villages2= 42 oes ? Do. 14W0220 Minor camp------ ? Do. T4W O22 ses £6 (a penne Aang peta t Do. 14W0222 Rockshelter___- ~~~ Historic, Great Bend, Excavated in 1957. Aksarben, and Archaic. 14W0223 Willage gout ek ? Visited. 14W0224 Rockshelter-_-__ ~~ - Historic, Great Bend, Excavated in 1957. and Aksarben. 14W0225f___-- aos ee wee ? Recorded in 1957. 14W0226 Petroglyph: 2 2-2 ? Recorded in 1957. 14W0227 Wilfageo occ cos Aksarben <2 0... Siva Visited. TAWO2ZZS) (ete Goes se ses aM Great Bend, Archaic___| Tested in 1957. 14W0229 Historic Fort_-_-_-_- White (Fort Belmont) _| Visited. 14W0230 Minor camp-_----- ? Do. 14WO231-~ Sf _=_-- GOs heme ais v¢ Do. LITERATURE CITED Anpreas, A. T. 1883. A history of the state of Kansas. Western Historical Company, Chicago. BELL, RoBert E., and BAERREIS, DAvip A. 1951. A survey of Oklahoma archaeology. Bull. Texas Archaeol. and Paleont. Soc., vol. 22, pp. 7-100. CoMMITTEE FOR THE RECOVERY OF ARCHAEOLOGICAL REMAINS. 1958. The Inter-Agency Archeological Salvage Program after twelve years. Univ. Missouri Press. FENENGA, FRANKLIN. 1953. The weights of chipped stone points; a clue to their functions. Southwestern Journ. Anthrop., vol. 9, No. 3, pp. 309-823. 370 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189 JAMES, EDWIN, EDITOR. 1956. an ~e “eg iP a a % he” 4 i m Oo p Dry Creek Rockshelter, site 14WO224, Toronto Reservoir, Kansas. Upper, view of exca- vator at work in Rockshelter; a,‘‘Harahey” blade. Lower, Artifacts from the Dry Creek Rockshelter. -h, Projectile points; i-j and n, end scrapers; k, flake knife; /, 0, blades; m, drill; p, bone needle; g, shaft smoother; 7, iron kettle fragment; s—t, Aksarben rim sherds, BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 55 a, b, Photographs of petroglyphs on large roof rock at the Dry Creek Rockshelter (14WO224); c, petroglyph at the Outlaw Rockshelter (14WO225); d ¢ photograph and drawing of “*Little-Water-Man” petroglyph at site 14WO226 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 56 Possum Point site (14WO228) Toronto Reservoir, Kansas. a, Initial excavations underway; b-e, projectile points; f, cupstone; g—k, rim sherds. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 57 Site 14GR210, Toronto Reservoir, Kansas. a, View of site; b, knife; c-1, projectile points; j-l, flake scrapers; m, mano. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 189 PLATE 58 Site 14GR216, Toronto Reservoir, Kansas. a, View of a portion of the excavated area showing hearth in place in right center; b—/, projectile points; m, bone awl; n-s, pot- sherds; ¢, cupstones; u, scraper; v, blade; w, ax. LIST OF REPORTS, ARTICLES, AND NOTES RELATING TO THE SALVAGE PROGRAMS PUBLISHED IN OTHER SERIES (As of August 1, 1963) ADAMS, WILLIAM RICHARD. 1946. Archeological survey of Martin County. Indiana Hist. Bull., vol. 23, No. 6, pp. 195-224. 1949. Archeological notes on Posey County, Indiana. Indiana Hist. Bur. Indianapolis. ADAMS, WILLIAM Y., and NETTIE K. 1959. An inventory of prehistoric sites on the Lower San Juan River, Utah. Mus. Northern Arizona Bull. 31, pp. 1-54. ANDERSON, Harry H. 1956. A history of the Cheyenne River Indian Agency (South Dakota) and its military post, Fort Bennett, 1868-1891. South Dakota Hist. Coll., vol. 28, pp. 890-551. 1961. The Fort Lookout Trading Post sites—A reexamination. Plains Anthrop., vol. 6, No. 14, pp. 221-229. 1962. Additional descriptive material on Fort Pierre II, an historic site on the Oahe Dam area, South Dakota. Plains Anthrop., vol. 7, No. 17, pp. 193-196. Basy, RAYMOND §S. 1946. Survey of Delaware Reservoir area. Mus. Echoes, vol. 19, No. 8, Ser. No. 208. 1949. Cowan Creek Mound exploration. Mus. Hchoes, vol. 22, No. 7, Ser. No. 2438. 1949. Stratton-Wallace site. Ann. Rep., Ohio State Archeol. and Hist. Soc., pp. 11-12. 1951. Explorations of the Cordray and Goldsmith Mounds. Mus. Hchoes, vol. 24, No. 12, Ser. No. 272. Bass, WILLIAM M. 1962. A preliminary analysis of burial data on 255 individuals from the Sully site, 39SL4, Sully County, South Dakota. Plains Anthrop., vol. 7, No. 16, pp. 77-78. BASs, WILLIAM M., and BIRKBY, WALTER H. 1962. The first human skeletal material from the Huff site, 32MO11, and a summary of putative Mandan skeletal material. Plains Anthrop., vol. 7, No. 17, pp. 164-177. BELL, RosBert BH. 1948. Recent archeological research in Oklahoma. Bull. Texas Archeol. and Paleont. Soc., vol. 19, pp. 148-154. 1949. Recent archeological research in Oklahoma, 1946-1948. Chronicles of Oklahoma, vol. 27, No. 3, pp. 303-312. 661-—932—64——_27 371 372 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] BELL, Ropert E.—Continued 1949. 1949. 1950. 1958. 1959. Archeological research in Oklahoma during 1947. Proce. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 6-7. Archeological excavations at the Harlan site, Fort Gibson Reservoir, Cherokee County, Oklahoma. Plains Archeol. Conf., Newsletter, vol. 3, No. 1, pp. 3-15. Lincoln. Notes on work at the Harlan site. Archeol. Newsletter, vol. 1, No. 4, pp. 1-2. Norman. . Notes on work in Eufaula and Tenkiller Reservoirs. Archeol. News- letter, vol. 2, No. 2, pp. 1-2. Norman. . Notes on Tenkiller excavations. Archeol. Newsletter, vol. 2, No. 3, pp. 1-2. Norman. . Notes on work at the Vanderpool sites. Archeol. Newsletter, vol. 2, No. 4, pp. 1-2. Norman. . Notes on Tenkiller and Eufaula excavations. Archeol. Newsletter, vol. 2, No. 5, pp. 1-4. Norman. . Notes on Keystone Reservoir survey. Oklahoma Anthrop. Soe. News- letter, vol. 1, No. 1, p. 8. . Prehistoric Oklahomans: or the Boomers came lately. Oklahoma Quart., vol. 1, No. 3, pp. 33-35. . Keystone archeological survey. Oklahoma Anthrop. Soc. Newsletter, vol. 1, No. 2, p. 4. . Archeological field work, Morris site. Oklahoma Anthrop. Soc. News- letter, vol. 1, No. 3, pp. 1-3. . Digging for Indian history. The Indian Sign, vol. 4, No. 2, pp. 9-12. . The Scott site, Le Flore County, Oklahoma. Amer. Antigq., vol. 18, No. 4, pp. 314-331. . Excavations at Lake Texoma, Marshall County, Oklahoma. Okla- homa Anthrop. Soc. Newsletter, vol. 3, No. 3, pp. 2-3. . Radiocarbon date from the Harlan site, Cherokee County, Oklahoma. Oklahoma Anthrop. Soc. Newsletter, vol. 5, No. 3, pp. 6-7. . Radiocarbon dates from the Harlan site, Cherokee County, Oklahoma. Oklahoma Anthrop. Soc. Newsletter, vol. 5, No. 6, p. 2. . Notes and news—Harlan site excavations. Oklahoma Anthrop. Soc. Newsletter, vol. 7, No. 3, pp. 1-2. . Archeological investigations at the Boat Dock site, Ma-1, in the Lake Texoma area, Marshall County, Oklahoma. Oklahoma Anthrop. Soc. Bull. VI, pp. 37-48. Radiocarbon dates from Oklahoma. Oklahoma Anthrop. Soc. News- letter, vol. 7, No. 3, pp. 3-4. Radiocarbon dates from Oklahoma sites. Oklahoma Anthrop. Soc. Newsletter, vol. 8, No. 4, pp. 2-3. BELL, Rosert E., and BAERREIS, DAvID A. 1951. A survey of Oklahoma archeology. Bull. Texas Archeol. and Paleont. Soc., vol. 22, pp. 7-100. (General study which contains information obtained by salvage program. ) Bet, Rosert E., and DALE, CHARLENE. 1953. The Morris site, Ck-39, Cherokee County, Oklahoma. Bull. Texas Archeol. and Paleont. Soc., vol 24, pp. 69-140. REPORTS IN OTHER SERIES ata BELL, Rosert E., and FRASER, RICHARD. 1952. Archeological discoveries at the Morris site, Cherokee County, Okla- homa. Chronicles of Oklahoma, vol. 30, No. 2, pp. 216-235. BIRKBY, WALTER H. 1962. A preliminary report on the dentition of the skeletal population of the Sully site. Plains Anthrop., vol. 7, No. 16, p. 79. BLiss, WESLEY L. 1949. Archeological reconnaissance in Wyoming and Montana, 1946-1947. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 7-12. 1949. Early Man in the Northwestern Plains. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 121-126. 1950. Birdshead Cave, a stratified site in the Wind River Basin, Wyoming. Amer. Antiq., vol. 15, No. 3, pp. 187-196. Bray, Rosert T. 1956. Some outstanding finds from Table Rock Reservoir area. Missouri Archeol. Soc. Newsletter, No. 99, pp. 5-7. BRETERNIRZ, DAvip A. 1957. Heltagito rock shelter (NA6380). Plateau, vol. 30, No. 1, pp. 1-16. Flagstaff. 1957. A brief archeological survey of the Lower Gila River. Kiva, vol. 22, Nos. 2-8. Tucson. Brew, J. O. and OTHERS. 1947. Symposium on river valley archeology. Amer. Antiq., vol. 12, No. 4, pp. 209-225. Brves, ALICE M. 1959. Skeletal material from the Morris site, Ck-39. Bull. Oklahoma An- throp. Soc., vol. 7, pp. 63-70. Buck, ARTHUR DEWEY, JR. 1959. The Custer Focus of the Southern Plains. Bull. Oklahoma Anthrop. Soc., vol. 7, pp. 1-82. BULLEN, RIPLEY P. 1950. An archeological survey of the Chattahoochee River Valley in Florida. Journ. Washington Acad. Sci., vol. 40, No. 4, pp. 101-125. BurGH, ROBERT F. 1949. Archeological field work of the University of Colorado Museum in 1947. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 12-14. BUBROUGHS, JOHN. 1958. Last look at a forgotten civilization. Popular Mechanics, pp. 114-116. May. Butter, B. ROBERT. 1955. The Wakemap Project—“Operation Last Chance.” Mazama, vol. 37, No. 13, pp. 33-38. 1957. Dalles Reservoir prehistory: A preliminary analysis. Washington Archaeol., vol. 1, No. 8, pp. 4-7. 1958. Ash Cave (45 WW 61): A Preliminary Report. Washington Ar- chaeol., vol. 2, No. pp. 3-10. 374 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] CALDWELL, JOSEPH R. 1948. Palachacolas Town, Hampton County, South Carolina, Journ. Wash- ington Acad. Sci., vol. 38, No. 10, pp. 321-824. 1950. A preliminary report on excavations in the Allatoona Reservoir. Early Georgia, vol. 1, No. 1, pp. 5-22. 1955. Cherokee pottery from northern Georgia. Amer. Antiq., vol. 20, No. 3, pp. 277-280. 1958. Trend and tradition in the prehistory of the eastern United States. Mem. No. 88, Amer. Anthrop. Assoc. Sci. Pap., vol. 10; Amer. Anthrop., vol. 60, No. 6, pt. 2. CALDWELL, JOSEPH R.; THOMPSON, CHARLES H.; and CALDWELL, SHEILA K. 1952. The Booger Bottom mound; A Forsyth Period site in Hall County, Georgia. Amer. Antiq., vol. 17, No. 4, pp. 319-828. CALDWELL, SHEILA KELLY. 1950. Reconstruction of the Woodstock Fort. Early Georgia, vol. 1, No. 2, pp. 22-30. CALDWELL, WARREN W. 1956. The archeology of Wakemap: A stratified site near The Dalles of the Columbia. Abs., microfilmed, Publ. No. 17, 119, University Microfilms, Ann Arbor. 1957. Cultural traditions in the Columbia Plateau and Southern Northwest Coast. Abs., Proc. Nebraska Acad. Sci., p. 4, April. 1957. Archeological salvage in the Missouri Basin. Progress, Missouri River Basin. Interior Missouri Basin Field Committee, October— December, pp. 47-55. Billings, Mont. 1958. Important archeological discoveries in the Arkansas Valley. Ozarks Mountaineer, vol. 6, No. 8, p. 7. 1960. The Black Partizan site (891LM218), Big Bend Reservoir, South Dakota: A preliminary report. Plains Anthrop., vol. 5, No. 10, pp. 53-57. 1960. Firearms and related materials from Ft. Pierre II (89ST17), Oahe Reservoir, South Dakota. Missouri Archaeol., vol. 22, December. Columbia. 1961. Excavations at certain La Roche and Thomas Riggs sites in the Big Bend and Oahe Reservoirs, 1960. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 57. CALDWELL, WARREN W.; McNutt, CHarites H.; and SmiryH, G. HUBERT. 1960. Fort Randall Reservoir. Corps of Engineers, Omaha, Nebraska, Spec. Publ. December. CASON, JOE F. 1952. Report on archeological salvage in Falcon Reservoir, season of 1952. Bull. Texas Archeol. and Paleont. Soc., vol. 23, pp. 218-259. CHAMPE, JOHN L. 1949. White Cat Village. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 285-292. CHAPMAN, CARL H. 1949. Archeological field work in Missouri, 1946-1947. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 17-19. 1954. Preliminary salvage archeology in the Pomme de Terre Reservoir area, Missouri. Missouri Archaeol., vol. 16, Nos. 3-4. Columbia. REPORTS IN OTHER SERIES 375 CHAPMAN, Car H., and Bray, Ropert T. 1956. Preliminary salvage archaeology in the Table Rock Reservoir area, Missouri. Missouri Archaeol., vol. 18, Nos. 1-2. Columbia. CHAPMAN, CARL H., ET AL, 1957. Table Rock salvage archeology. In A report of progress, archeological research by the University of Missouri, 1955-1956. Special Publ. Missouri Archeol. Soc., February, pp. 5-37. Columbia. CHAPMAN, CARL H.; MAxwELut, THOMAS J., JR.; and KozLovicu, EUGENE. 1951. A preliminary archeological survey of the Table Rock Reservoir area, Stone County, Missouri. Missouri Archeol., vol. 13, No. 2, pp. 8-38. Columbia. CoALE, GEORGE L. 1956. Archeological survey of the Mt. Sheep and Pleasant Valley Reservoirs. Davidson Journ. Anthrop., vol. 2, No. 1, pp. 1-27. Coogan, ALAN H., and Irvine, WM. N. 1959. Late Pleistocene and recent Missouri River terraces in the Big Bend Reservoir, South Dakota. Abs., Iowa Acad. Sci., vol. 66, pp. 317-327. CooLrey, MAURICE EH. 1958. Physiography of the Glen-San Juan Canyon area, pt. 1. Mus. North- ern Arizona, Plateau, vol. 31, No. 2, pp. 21-33. Flagstaff. 1958. Physiography of the Glen-San Juan Canyon area, pt. 2. Mus. Northern Arizona, Plateau, vol. 31, No. 3, pp. 49-60. Flagstaff. Cooper, PAUL L. 1949. Recent investigation in Fort Randall and Oahe Reservoirs, South Dakota. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 300-811. 1949. An archeological survey of the Fort Randall Reservoir, South Dakota. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab Anthrop., Univ. Nebraska, pp. 19-20. 1955. The archeological and paleontological salvage program in the Mis- souri Basin, 1950-1951. Smithsonian Misc. Coll., vol. 126, No. 2. CoRBETT, JOHN M. 1949. Salvage archeology in river basins. Planning and Civic Comment, Quart. Amer. Planning and Civic Assoc., July—Sept. 1954. New dates from old data. Reclamation Era, vol. 40, No. 3, pp. 56-58. CRAMPTON, C. GREGORY. 1959. Outline history of the Glen Canyon region, 1776-1922. Univ. Utah, Anthrop Pap. No. 42. CREER, LELAND HARGROVE. 1958. Mormon towns in the region of the Colorado. Univ. Utah, Anthrop. Pap. No. 32. 1958. The activities of Jacob Hamblin in the region of the Colorado. Univ. Utah, Anthrop. Pap. No. 33. CrESSMAN, L. S. ' 1960. Cultural sequences at The Dalles, Oregon. Trans. Amer. Philos. Soc., vol. 50, pt. 10. December. CrEessMAN, L. 8.; Core, Davin L.; Davis, Witspur A.; Newman, THOMAS M.; and SCHEANS, DANIEL J. 1959. Cultural sequences at The Dalles, Oregon. Univ. Oregon. 376 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] CuMMINGs, THOMAS §. 1953. A preliminary report on the Blue Stone Focus, White Rock Aspect, Abs. Proc. Nebraska Acad. Sci., May. DANSON, EDWARD B. 1958. The Glen Canyon Project. Plateau, vol. 30, No. 3, pp. 75-78. Flag- staff. DAUGHERTY, RICHARD D. 1952. Archeological investigations in O’Sullivan Reservoir, Grant County, Washington. Amer. Antiq., vol. 17, No. 4, pp. 374-383. 1956. Archeology of the Lind Coulee site, Washington. Proc. Amer. Philos. Soe., vol. 100, No. 8, pp. 224-278. 1956. An archeological survey of Rocky Reach Reservoir. Northwest Arche- ology Number, Research Stud. State College of Washington, vol. 24, No.1, pp. 1-16. 1959. Early Man in Washington (Archeology of Lind Coulee site). Div. Mines and Geol., Inf. Cire. No. 32. Davis, E. Mort. 1950. The present status of the study of ‘“Harly Man” in Nebraska. Abs. Proc. Nebraska Acad. Sci., May. 1951. “Marly Man” sites in the Medicine Creek Reservoir area. Abs. Proc. Nebraska Acad. Sci., April. 1953. Early human occupation of the Plains area. Abs. Proc. Nebraska Acad. Sci., May. 1953. Recent data from two paleo-Indian sites on Medicine Creek, Nebraska. Amer. Antiq., vol. 18, No. 4, pp. 380-386. 1954. The Bridger Basin, Wyoming: An area of archeological promise. Abs. Proc. Nebraska Acad. Sci., April. 1956. Archeological survey of the Big Sandy Reservoir area, southwestern Wyoming. Notebook No. 2, Lab. Anthrop. Univ. Nebraska. 1961. Archeological sequence at the Ferrell’s Bridge Reservoir, northeastern Texas. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 61. 1962. Archeological salvage investigation of the Harling Mound in Fannin County, Texas. Year-Book of the Amer. Philos. Soc., pp. 487-489. 1962. Archeology of the Lime Creek site in Southwestern Nebraska. Spec. Publ. Univ. Nebraska State Museum, No. 3, pp. 1-106. Davis, E. Mort, and Scuuttz, C. BERTRAND. 1952. The archeological and paleontological salvage program at the Medicine Creek Reservoir, Frontier County, Nebraska. Science, vol. 115, No. 2985, pp. 288-290. Davis, HESTER A. 1961. 1960 fieldwork in Beaver Reservoir, Arkansas. Plains Anthrop., vol. 62No,; 12. pte 72: Davis, WILLIAM A., and Davis, E. Mort. 1960. The Jake Martin site: An Archaic site in the Ferrell’s Bridge Reser- voir area, Northeastern Texas. Univ. Texas, Archeol. Ser., No. 3. DE BAILLOU, CLEMENS. 1962. Archeological salvage in the Morgan Falls Basin. Univ. Georgia Lab. Archeol. Ser., No. 4, pp. 1-18. DEJ ARNETTE. Davin L., and HANSEN, ASAEL T. 1960. The archeology of the Childersburg site, Alabama. Florida State Univ. Notes Anthrop., No. 6. REPORTS IN OTHER SERIES Ov’ DIBBLE, CHARLES PB). 1959. Ecological studies of the flora and fauna in Glen Canyon. Univ. Utah, Anthrop. Pap., No. 40. Drse.e, Davin S., and Day, KEnt C. 1962. A preliminary survey of the Fontenelle Reservoir, Wyoming. Univ. Utah. Anthrop Pap., No. 58. (Upper Colorado Ser., No. 7). Di Prso, CHARLES C. 1953. The Sobaipuri Indians of the Upper San Pedro River Valley, South- eastern Arizona. The Amerind Foundation, Ine., Dragoon, Ariz., No. 6. DittEert, ALFRED E., JR. 1957. The salvage archeology program is resumed at the Navajo Dam. Hl Palacio, vol. 64, Nos. 7-8, pp. 245-246. Santa Fe. 1958. Salvage archeology and the Navajo Project: A progress report. El Palacio, vol. 65, No. 2, pp. 61-72. Santa Fe. 1958. Recent developments in Navajo project archeology. El Palacio, vol. 65, No. 6. pp. 201-211. Santa Fe. DirtTertT, ALFRED E., JR.; HESTER, JIM J.; and Eppy, FRANK W. 1961. An archaeological survey of the Navajo Reservoir district North- western Mexico. School Amer. Res. and Mus. New Mexico, No. 23. Santa Fe. DraAcoo, Don W. 1951. Archeological survey of Shelby County, Indiana. Indiana Hist. Bur., Indianapolis. DUFFIELD, LATHEL F, 1961. The Limerick site at Iron Bridge Reservoir, Rains County, Texas. Bull. Texas Archeol. Soc., vol. 30, pp. 51-116. 1962. Archeology of the Sanford Reservoir in the Texas Panhandle. Plains Anthrop., vol. 7, No. 16, p. 79. DUFFIELD, LATHEL F., and JELKS, EDWARD B. 1961. The Pearson site: A historic Indian site at Iron Bridge Reservoir, Rains County, Texas. Univ. Texas, Archeol. Ser., No. 4. Eppy, FRANK W. 1961. Excavations at Los Pinos Phase sites in the Navajo Reservoir dis- trict. Mus. of New Mexico Pap. in Anthrop., No. 4, pp. 1-107. Santa Fe. EVANS, OREN F. 1958. Analysis of flint materials from the Lee and Lacy sites. Oklahoma Archeol. Soc. Newsletter, vol. 6, No. 7, pp. 2-8. Norman. FarrRBANKS, CHARLES H. 1954. 1953 excavations at Site 9HL64, Buford Reservoir, Georgia. Florida State Univ. Stud. Anthrop., No. 16, pp. 1-25. FENENGA, FRANKLIN. 1952. The archeology of the Slick Rock Village, Tulare County, California. Amer. Antiq., vol. 17, No. 4, pp. 339-347. 1958. Rediscovering the past in the Missouri Basin. Progress, Missouri River Basin, Interior Missouri Basin Field Committee, June, pp. 202-209. Billings, Mont. 1953. The ice-glider game, an 18th-century innovation in Northern Plains culture. Abs. Proc. Nebraska Acad. Sci., May. 378 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] FENENGA, FRANKLIN—Continued 1954. The interdependence of archeology and ethnology as illustrated by the ice-glider game of the Northern Plains. Plains Anthrop., No. 1, pp. 31-88. May. FLOWERS, SEVILLE, ET AL. 1960. Ecological studies of the flora and fauna of Flaming Gorge Reservoir Basin, Utah and Wyoming. Univ. Utah, Anthrop. Pap., No. 48 (Upper Colorado Series No. 3). GAINES, XERPHA M. 1957. Plants in Glen Canyon. Mus. Northern Arizona, Plateau, vol. 30, No. 2, pp. 31-34. Flagstaff. GANT, ROBERT. 1962. The Big Bend burials, 39BF221, Buffalo County, South Dakota. Museum News, W. H. Over Museum, State Univ. South Dakota, vol. 23, Nos. 4-5, pp. 3-7. GARRETT, JOHN W. 1952. Preliminary investigations of an aboriginal occupation site near Sanish, North Dakota. Montana State Univ. Anthrop. and Sociol. Pap., No. 18. GARTH, THOMAS R. 1951. Historic sites in the Fort Randall Reservoir area. Abs. Proc. Nebraska Acad. Sci. April. GILBERT, ELIZABETH X. 1961. A pithouse village on the San Juan River, N. Mex. Southwestern Lore, vol. 27, No. 1, pp. 9-16. GROVE, FRED. 1949. Archeologists uncover real Sooners. Univ. Oklahoma News of the Month, vol. 2, No. 3, pp. 2-3. GUNKEL, ALEXANDER. 1961. The archeology of the Rocky Reach Reservoir. (A comparative cul- tural analysis of four archeological sites in the Rocky Reach Reser- voir region, Washington.) Washington State Univ., Dept. Sociol. and Anthrop. GUNNERSON, JAMES H. 1958. Archeological survey of the Kaiparowitz Plateau. A preliminary re- port. Utah Archeology Newsletter, vol. 4, No. 3, pp. 9-20. Salt Lake City. 1959. 1957 excavations, Glen Canyon area. Univ. Utah, Anthrop. Pap., No. 43. HANDLEY, CHARLES O., JR. 1953. A new South Dakota locality for the kangaroo rat, dipodomys. Journ. Mammalogy, vol. 34, No. 2, p. 264. May. Harris, R. K. 1951. Plainview point from site 18C7-8. The Record, Dallas Archeol. Soc., vol. 10, No. 1, p. 2. HASKELL, Horace 8. 1958. Flowering plants in Glen Canyon—late summer aspect. Mus. Northern Arizona, Plateau, vol. 31, No.1, pp.1-3. Flagstaff. REPORTS IN OTHER SERIES 379 HEDDEN, MARK. 1958. Surface printing as a means of recording petroglyphs. Amer. Antigq., vol. 23, No. 4, pt. 1, pp. 485-489. April. HELDMAN, DONALD P. 1961. Archaeological work within the proposed Joanna Reservoir. Plains Anthrop., vol. 6, No. 12, pt. 2, pp. 110-111. HENNING, DALE R. 1961. Archaeological research in the proposed Joanna Reservoir, Missouri. Missouri Archaeol., vol. 23, pp. 133-183. 1962. The Joanna Reservoir. Plains Anthrop., vol. 7, No. 16, pp. 84-86. HENSLEY, G. S. 1952. Story of first midwest farmers unfolds in White River Valley. Missouri Farmer, pp. 6-9. September. Columbia. HEWES, Gorpon W. 1949. The 1947 summer field session in archeology, University of North Da- kota. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 21-24. 1949. Pottery from the sites excavated by the 1947 North Dakota field ses- sion. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska. pp. 58-67. 1949. Burial mounds in the Baldhill area, North Dakota. Amer. Antiq., vol. 14, No. 4, pp. 322-328. 1950. Sheyenne River, Proc. Sixth Plains Archeol. Conf. (1948), Univ. Utah, Anthrop. Pap. No. 11, p. 9. HockKMAN, STEVEN. 1952. The history of Fort Sully (South Dakota). South Dakota Hist. Coll., vol. 26, pp. 222-227. HOoLpeErR, PRESTON, and WIKE, JOYCE. 1949. The frontier culture complex, a preliminary report on a prehistoric hunter’s camp in southwestern Nebraska. Amer. Antiq., vol. 14, No. 4, pp. 260-266. 1950. The Allen site (FT-50): Archeological evidence of an early hunter’s camp on Medicine Creek, Frontier County, Nebraska. Proc. Sixth Plains Archeol. Conf. (1948), Univ. Utah, Anthrop. Pap., No. 11, pp. 105-107. HOWARD, JAMES H. 1962. Report of the investigation of the Huff site, 32M011, Morton County, North Dakota, 1959. Univ. North Dakota, Anthrop, Pap., No. 2. Howarp, LYNN E. 1951. Archeological survey in the Bull Shoals region of Arkansas. Missouri Archaeol., vol. 18, No. 1, pp. 3-17. Columbia. HucHEs, JAcK T. 1949. Investigations in western South Dakota and northeastern Wyoming. Amer. Antiq., vol. 14, No. 4, pp. 266-277. 1950. An experiment in relative dating of archeological remains by stream terraces. Bull. Texas Archeol. and Paleont. Soe., vol. 21, pp. 97-104. Hurt, AMY PASSMORE. 1958. The race for treasure. Denver Post, Sunday Empire Mag., Aug. 31, 1958. 380 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Hurt, WESLEY R., JR. 1951. Report of the investigations of the Swanson site, 39BR16, Brule County, South Dakota, 1950. South Dakota Archaeol. Comm., Ar- chaeol. Stud., Cire. No. 6. 1952. Report of the investigation of the Scalp Creek site, 39GR1, and the Ellis Creek site, 89GR2, South Dakota, 1941, 1951. South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 4. 1952. House types of the Over Focus, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 4, pp. 51-52. 1953. Report of the investigation of the Thomas Riggs site, 39HU1, Hughes County, South Dakota, 1952. South Dakota Archaeol. Comm., Ar- chaeol. Stud., Cire. No. 5. 1954. Report of the investigations of the Spotted Bear site, 39HU26, and the Cottonwood site, 89HU48, Hughes County, South Dakota, 1953. South Dakota Archaeol. Comm., Archaeol. Stud., Cire. No. 6. IRVING, WILLIAM N. 1958. The chronology of early remains at the Medicine Crow site, 39BF2, South Dakota. Abs. Proc. Nebraska Acad. Sci., p.3. April. JELKS, EDWARD B. 1952. The River Basin Surveys archeological program in Texas. Texas Journ. Sci., vol. 4, No. 2, pp. 1381-138. 19538. Excavations at the Blum rockshelter. Bull. Texas Archeol. and Pa- leont. Soc., vol. 24, pp. 189-207. 1959. Archeologists add new data on Texas’ past. Univ. Texas, Engineering Science News, vol. 7, No. 1, pp. 1-4. 1962. The Kyle site, a stratified Central Texas Aspect site in Hill County, Texas. Univ. Texas Archaeol. Ser., No. 5. JELKS, HDWARD B., and TUNNELL, CurRTIS D. 1959. The Harroun site: A Fulton Aspect Component of the Caddoan area, Upshur County, Texas. Univ. Texas Archaeol. Ser., No. 2. JENNINGS, JESSIE D. 1947. An atlas of archeological sites within the reservoirs located in the Missouri River Basin. U.S. Dept. Interior, Reg. Two, Nat. Park Serv. Missouri Basin Recreation Survey. Omaha. 1948. Plainsmen of the past, a review of the prehistory of the Plains. U.S. Dept. Interior, Reg. Two, Nat. Park Serv., Missouri Basin Recrea- tion Survey. Omaha. 1948. Saving a segment of history. Reclamation Era, vol. 34, No. 10, pp. 192-193. October. 1955. The archeology of the Plains: an assessment. U.S. Dept. Interior, Reg. Two., Nat. Park Serv. and Univ. Utah, Dept. Anthrop. (Super- sedes Jennings, 1948.) 1961. Salvage and scholarship. Northwestern Univ. Tri-Quart., vol. 3, No. 2, pp. 48-47. Winter. 1963. Administration of Contract Emergency Archaeological Programs. Amer. Antiq., vol. 28, No. 3, pp. 282-285. JOHNSON, FREDERICK. 1951. The Inter-Agency Archeological Salvage Program in the United States. Archaeology, pp. 25-40. Spring. REPORTS IN OTHER SERIES 381 JOHNSON, LEROY, JR. 1961. The Devils Mouth site: A river terrace midden, Diablo Reservoir, Texas. Bull. Texas Archeol. Soc., vol. 30, pp. 255-285. JoHNSON, LERoy, Jr., and JELKS, HDwArp B. 1958. The Tawakoni-Ysconi village, 1760: A study in archeological site iden- tification. Texas Journ. Sci., vol. 10, No.4. December. JOHNSON, LERoy, Jr., SUHM, DEE ANN, and TUNNELL, CuRTIS D. 1962. Salvage Archeology of Canyon Reservoir: The Wunderlich, Foot- bridge, and Oblate sites. Bull. Texas Mem. Mus., No. 5, pp. 1-26. KEEFE, JIM, and SAULTS, DAN. 1953. Out of the dust of yesterday. Missouri Conservationist, vol. 14, No. 1. January. KELLAR, JAMES H., KEtty, A. R., and McMIcHAEL, Epwarp V. 1962. The Mandeville site in southwest Georgia. Amer. Antiq., vol. 27, No. 38, Dp. 8386-855. KELLER, CHARLES M. 1961. Activities in the Kasinger Bluff Reservoir. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 67. KELLY, A. R., and NEITZEL, R. S. 1961. The Chauga site in Oconee County, South Carolina. Univ. Georgia Lab. Archaeol. Ser., Rep. No. 3. KELLY, A. R., NoNAS, RICHARD, ET AL. 1962. Survey of archaeological sites in Clay and Quitman Counties, Georgia. Univ. Georgia Lab. Archaeol. Ser., Rep. No. 5. KIEHL, Mary. 1953. The Glen Hider and White Rock sites in north central Kansas. Abs. Proc. Nebraska Acad. Sci. May. KIveTtT, MARVIN F. 1948. Mechanized archeology. Plains Archeol. Conf. News Letter, vol. 1, No. 4, pp. 16-17. 1949. Archeological investigations in Medicine Creek Reservoir, Nebraska. Amer. Antiq., vol. 14, No. 4, pp. 278-284. 1949. An archeological survey of the Garrison and Baldhill Reservoirs. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 24-25. 1950. An Archaic horizon? Plains Archeol. Conf. News Letter, vol. 3, No. 4, pp. 4-7. 1952. Woodland sites in Nebraska. Nebraska State Hist. Soc. Publ. in Anthrop., No. 1. (While some of the material in this report was obtained during earlier investigations, a considerable portion of the publication is based on the results of salvage projects.) 1954. Notes on the burial patterns of the Central Plains Indians. Abs. Proc. Nebraska Acad. Sci. April. 1962. Archaeological field report, 1961. Plains Anthrop., vol 7, No. 16, p. "ile Kivert, Marvin F., and Hirt, A. T. 1949. Archeological investigations along Medicine Creek. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 25-26. 382 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] LEHMER, DONALD J. 1951. Pottery types from the Dodd site, Oahe Reservoir, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 2, pp. 1-15. 1952. The Turkey Bluff Focus of the Fulton Aspect. Amer. Antigq., vol. 17, No. 4, pp. 313-318. 1952. The Fort Pierre Branch, central South Dakota. Amer. Antiq., vol. 17, No. 4, pp. 329-336. ; 1952. Animal bone and Plains archeology. Plains Archeol. Conf. News Let- ter, vol. 4, No. 4, pp. 53-55. 1954. The sedentary horizon of the northern Plains. Southwestern Journ. Anthrop., vol. 10, No. 2, pp. 189-159. LINDSAY, ALEXANDER J., JY. 1961. The Beaver Creek agricultural community on the San Juan River, Utah. Amer. Antiq., vol. 27, No. 2, pp. 174-187. October. LIprk, WILLIAM D. 1960. 1958 excavations, Glen Canyon area. With appendices by Christy G. Turner II, Lyndon L. Hargrave. Univ. Utah, Anthrop. Pap., No. 44, (Glen Canyon Ser., No. 11). Lipe, WILLIAM, and Fow ter, Don. 1958. Archeological excavations and survey in Glen Canyon: Preliminary reports of 1958 work. Utah Archeol., vol. 4, No. 4, pp. 416. Salt Lake City. Lipk, WILLIAM D.; SHARROCK, FLOYD W.; Dissie, Davin S.; and ANDERSON, KeitH M. 1960. 1959 excavations, Glen Canyon area. With appendix by Christy G. Turner II, and addendum by Dee Ann Suhm. Uniy. Utah, Anthrop. Pap., No. 49. (Glen Canyon Ser. No. 13.) LIsTER, Ropert H. 1957. Salvage archeology in the Southwest. Southwestern Lore, vol. 23, No. 3, pp. 834-35. Boulder. 1957. The Glen Canyon survey in 1957. Univ. Utah, Anthrop. Pap., No. 30. 1958. A preliminary note on excavations at the Coombs site, Boulder, Utah. Utah Archeol.: A Newsletter, vol. 4, No. 3, pp. 48. Salt Lake City. 1959. The Coombs site, with a chapter, Pottery, by Florence C. Lister. Univ. Utah, Anthrop. Pap., No. 41. 1962. Archaeological Survey of the Blue Mesa Reservoir, Colorado. South- western Lore, vol. 28, No. 3, pp. 41-60. LIsTER, Rosert H.; AMBLER, J. RICHARD; and LISTER, FLORENCE L. 1960. The Coombs site. Part II. Univ. Utah, Anthrop. Pap., No. 41. LISTER, RopeRT H., and LISTER, FLORENCE C. 1961. The Coombs site. Part III, summary and conclusions. Univ. Utah., Anthrop. Pap., No. 41, October. McDovucaLi, WALTER B. 1959. Plants of the Glen Canyon area in the Herbarium at the Museum of Northern Arizona (mimeographed). McKusrick, M. B., and Watson, R. S. 1959. Grinding implements from Vaquero Reservoir, San Luis Obispo and Santa Barbara Counties. Univ. California, Ann. Rep., Archeol., Surv., Dept. Anthrop.-Sociol., pp. 13-14. Los Angeles. REPORTS IN OTHER SERIES 383 MocMicHarEL, Epwarp V., and KELuar, JAMES H. 1960. Archeological salvage in the Oliver Basin. Univ. Georgia Lab. Archeol. Ser., Rep. No. 2. McNutt, CHARLES H. 1958. La Roche ware and relative chronology. Abs., Proc. Nebraska Acad. Sci., April, pp. 3-4. McNutt, CHARLES H., and WHEELER, RICHARD P. 1959. Bibliography of primary sources for radiocarbon dates. Amer. Antiq., vol. 24, No. 3, pp. 323-824. MALOourF, CARLING. 1950. The archeology of the Canyon Ferry region, Montana, 1950. Univ. Montana Anthrop. and Sociol. Pap. No. 11. 1951. Archeological studies of aboriginal occupation sites in northwestern North Dakota. Montana State Univ. Anthrop. and Sociol. Pap., No. 7. MARSHALL, RICHARD. 1962. Archaeological survey of Jasper County, Missouri. Plains Anthrop., vol. 7, No. 16, pp. 82-84. Mattes, MERRILL J. 1947. Historic sites in Missouri Valley reservoir areas. Nebraska Hist., vol. 28, No. 3, pp. 1-15. 1949. Historic sites in the Fort Randall Reservoir area. South Dakota Hist. Coll., vol. 24. 1952. Revival at old Fort Randall. Military Engineer, vol. 44, No. 298, pp. 88-93. 1952. Salvaging Missouri Valley history. The Westerners Brand Book, vol. 9, No. 3, pp. 17-19, 22-24. 1954. Under the wide Missouri [historic sites in Missouri Basin reservoirs]. North Dakota Hist., vol. 21, No. 4, pp. 146-167. Mattison, Ray H. 1951. Old Fort Stevenson, a typical Missouri River military post. North Dakota Hist., vol. 18, Nos. 2-3, pp. 2-40. 1954. The Army post on the northern Plains, 1865-1885. Nebraska Hist., vol. 35, No. 1, pp. 1-27. 1954. Report on historical aspects of the Oahe Reservoir area, Missouri River, South and North Dakota. South Dakota Hist. Coll. and Rep., vol. 27. 1955. The Indian reservation system on the upper Missouri, 1865-1890. Ne- braska Hist., vol. 36, No. 3, pp. 141-172. 1955. Report on historic sites in the Garrison Reservoir area, Missouri River [North Dakota]. North Dakota Hist., vol. 22, Nos. 1-2, pp. 5-73. 1956. The military frontier on the upper Missouri. Nebraska Hist., vol. 37, No. 3, pp. 159-182. 1956. Report on historic sites adjacent to the Missouri River, between Big Sioux River and Fort Randall Dam, including those in the Gavins Point Reservoir area (South Dakota and Nebraska). South Dakota Hist. Coll., vol. 28, pp. 22-98. 384 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189) MAYER-OAKES, WILLIAM J. 1953. An archeological survey of the proposed Shenango River Reservoir area in Ohio and Pennsylvania. Ann. Carnegie Mus., vol. 33, art. 3, pp. 115-124, Anthrop. Ser. No. 1. 1962. Archaeological work at Grand Rapids, Manitoba. Plains Anthrop., vol. 7, No. 16, p. 88. MELEEN, BE. E. 1949. A preliminary report on the Thomas Riggs village site. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 310-821. (This paper includes material from investigations in the Oahe Reservoir area prior to the salvage pro- gram but also reports on work done in 1947.) METCALF, GEORGE. 1956. Additional data from the Dodd and Phillips Ranch sites, South Dakota. Amer. Antiq., vol. 21, No. 3, pp. 305-309. MILLER, CARL F. 1948. Early cultural manifestations exposed by the archeological survey of the Buggs Island Reservoir in southern Virginia and northern North Carolina. Journ. Washington Acad. Sci., vol. 38, No. 12, pp. 397- 399. 1949. The Lake Spring site, Columbia County, Georgia. Amer. Antiq., vol. 15, No. 1, pp. 38-51. 1949. Appraisal of the archeological resources of the Buggs Island Reservoir in southern Virginia and northern North Carolina. Quart. Bull., Archeol. Soc. Virginia, vol. 4, No. 1. 1949. Early cultural manifestations exposed by the archeological survey of the Buggs Island Reservoir in southern Virginia and northern North Carolina. Quart. Bull., Archeol. Soc. Virginia, vol. 4, No. 2. 1950. Early cultural horizons in the southeastern United States. Amer. Antiq., vol. 15, No. 4, pp. 273-288. (A general article but contain- ing data collected during survey work.) 1950. An analysis and interpretation of the ceramic remains from site 38 Mc6 near Clarks Hill, South Carolina. Journ. Washington Acad. Sci., vol. 40, No. 11, pp. 350-354. 1956. Burin types from southern Virginia: A preliminary statement. Amer. Antiq., vol. 21, No. 3, p. 311. 1959. Physical structure of Rock Mound at 9ST3, Georgia. Southern Indian Stud., Archeol. Soc. North Carolina and Res. Lab. Anthrop., Univ. North Carolina, vol. 11, pp. 16-19. MItteRr, E. O., and JELKS, EDWARD B. 1952. Archeological excavations at the Belton Reservoir, Coryell County, Texas. Bull. Texas Archeol. and Paleont. Soc., vol. 23, pp. 168-217. MILiLrer, Love H. 1957. Bird remains from an Oregon Indian midden. The Condor, vol. 59, No. 1, p. 59-63. January-February. MILLER, WILLIAM C., and BRETERNITZ, Davin A. 1958. 1957 Navajo Canyon survey, preliminary report. Plateau, vol. 30, No. 3, pp. 72-74. Flagstaff. 1959. 1958 Navajo Canyon survey, preliminary report. Plateau, vol. 31, No. 1, pp. 3-7. Flagstaff. REPORTS IN OTHER SERIES 385 MILLS, JOHN H.., and OSBoRNE, CAROLYN. 1952. Material culture of an Upper Coulee rockshelter. Amer. Antiq., vol. 17, No. 4, pp. 352-359. MOosIMAN, JAMES, and RAss, GEORGE B. 1952. The herpetology of Tiber Reservoir area, Montana. Copeia, No. 1, pp. 23-27. June 2. MULLOy, WILLIAM. 1954. The McKean site in northeastern Wyoming. Southwestern Journ. Anthrop., vol. 10, No. 4, pp. 432-460. NEUMAN, ROBERT W. 1957. Supplementary data on the White Rock Aspect. Abs., Proc. Nebraska Acad, Sci., pp. 3-4. April. 1960. The Truman Mound site, Big Bend Reservoir area, South Dakota. Amer. Antiq., vol. 26, No. 1, pp. 78-92. July. 1961 Excavations at four mound sites in the Oahe Reservoir. Plains An- throp., vol. 6, No. 12, pt. 1, pp. 57-58. 1962. Field season, 1961. Plains Anthrop., vol. 7, No. 16, p. 81. 1962. Historic Indian burials, Fort Thompson, South Dakota. Plains Anthrop., vol. 7, No. 16, p. 95. 1963. Check-stamped pottery on the Northern and Central Great Plains. Amer. Antiq., vol. 29, No. 1, pp. 17-26. OSBORNE, DOUGLAS. 1950. An archeological survey of the Benham Falls Reservoir, Oregon. Amer. Antiq., vol. 16, No. 2, pp. 112-120. 1953. Archeological occurrences of pronghorn antelope, bison, and horse in the Columbia Plateau. Sci. Month., vol. 77, No. 5, pp. 260-269. OSBORNE, DoUGLAS ; CALDWELL, WARREN C.; and CRABTREE, ROBERT H. 1956. The problem of Northwest coastal-interior relationships as seen from Seattle. Amer. Antiq., vol. 22, No. 2, pp. 117-128. OsBORNE, DOUGLAS, and CRABTREE, ROBERT H. 1961. Two sites in the Upper McNary Reservoir. Tebiwa, Journ. Idaho State College Mus., vol. 4, No. 2, pp. 19-36. OSBORNE, DOUGLAS ; CRABTREE, ROBERT; and BRYAN, ALAN. 1952. Archeological investigations in the Chief Joseph Reservoir. Amer. Antigq., vol. 17, No. 4, pp. 360-373. OSMUNDSON, JOHN, and HULSE, CHRISTOPHER. 1962. Preliminary report on an archeological survey of the Bruces Eddy Reservoir, north-central Idaho, 1961. Tebiwa, Journ. Idaho State College Mus., vol. 5, No. 1, pp. 11-29. PECKHAM, STEWART. 1958. Salvage archeology in New Mexico, 1957-58: A partial report. El Palacio, vol. 65, No. 5, pp. 161-168. Santa Fe. PENDERGAST, Davin M., and MEIGHAN, CLEMENT W. 1959. The Greasy Creek site, Tulare County, Arizona. Univ. California, Ann. Rep. Archeol. Surv. Dept. Anthrop.-Sociol., pp. 1-9. Los Angeles, Pierson, Luioyp. 1957. A brief archeological reconnaissance of White Canyon, southeastern Utah. El Palacio, vol. 64, Nos. 7-8, pp. 222-230. Santa Fe. 386 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Proctor, CHARLES C. 1953. Report of excavations in the Hufaula Reservoir. Bull. Oklahoma Anthrop. Soce., vol. 1, pp. 48-59. PuRDY, WILLIAM M. 1959. An outline of the history of the Flaming Gorge area. Univ. Utah Anthrop. Pap., No. 37. ; ROBERT, FRANK H. H., JR. 1948. A crisis in U.S. archeology. Sci. Amer., vol. 179, No. 6, pp. 12-17. 1952. River Basin Surveys: The first five years of the Inter-Agency Archeo- logical and Paleontological Salvage Program. Ann. Rep. Smith- sonian Inst. for 1951, pp. 351-383. 1955. The Inter-Agency Archeological and Paleontological Salvage Program in the United States. Pro Natura, vol. 2, pp. 213-218. (Series pub- lished by International Union for the Protection of Nature, Brussels.) 1955. The Inter-Agency Archeological and Paleontological Salvage Program. Missouri Archeol. Soc. News Letter No. 96, pp. 7-10, November. Rupy, Jack R., and STIRLAND, RosBert D. 1950. An archeological reconnaissance in Washington County, Utah. Univ. Utah, Anthrop. Pap., No. 9. ScHOLTz, JAMES A. 1962. Salvage archaeology in Arkansas. Plains Anthrop,, vol. 7, No. 16, p. 84. ScHULTz, C. BERTRAND, and FRANKFORTER, W. D. 1948. Early man. Amer. Antiq., vol. 13, No. 3, pp. 279-280. 1948. Preliminary report on the Lime Creek sites: new evidence of early man in southwestern Nebraska. Bull. Univ. Nebraska State Mus., vol. 8, No. 4, pt. 2. 1949. The Lime Creek sites. Jn Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 182-134. SCHWARTZ, DoucLas W. 1961. The Tinsley Hill site. (A late prehistoric stone grave cemetery in Lyon County, Ky.) Univ. Kentucky, Stud. Anthrop., No. 1. 1962. The Driskill site; a Late Woodland occupation in the Lower Cumber- land River Valley. Trans. Kentucky Acad. Sci. vol. 23, Nos. 1-2; 1-18. SEARS, WILLIAM H. 1950. Preliminary report on the excavation on an Etowah Valley site. Amer. Antiq., vol. 16, No. 2, pp. 137-142. SHANE, RALPH M. 1956. A short history of Fort Berthold [Indian Reservation]. 22 pp. Fort Berthold Indian Agency, Newtown, N. Dak. SHARROCE, FLoyD W., ET AL. 1963. 1961 excavations, Glen Canyon area. Univ. of Utah Anthrop. Pap., No. 63 (Glen Canyon Series No. 18). SHARROCK, FLoyD W.; ANDERSON, Kerra M.; Fow rer, Don D.; and D1sste, Davin §. 1961. 1960 excavations, Glen Canyon area. Univ. Utah, Anthrop. Pap., No. 52, May. (With appendices by Don D. Fowler, and Christy G. Turner, II.) REPORTS IN OTHER SERIES 387 SHARROCK, FLoyp W.; Dissiz, Davin S.; and ANDERSoN, KzuitH M. 1961. The Creeping Dune Irrigation site in Glen Canyon, Utah. Amer. Antiq., vol. 27, No. 2, pp. 188-202. October. SHarRockK, FLoyp W., and Kranz, Epwarp A. 1962. Carnegie Museum collection from southeast Utah. Univ. Utah, Anthrop. Pap., No. 57. (Glen Canyon Ser., No. 160.) SHiner, Jorn L. 1952. The 1950 excavations at site 45BN6, McNary Reservoir, Washington. Amer. Antiq., vol. 17, No. 4, pp. 348-351. Surpeer, J. M. 1953. A Folsom fluted point from Marshall County, Kansas. Plains Archeol. Conf. News Letter, vol. 5, No. 4, p. 54. Smita, CARLYLE 8. 1949. Archeological investigations in Ellsworth and Rice Counties, Kansas. 1949. 1949, 1950. 1951. 1959. 1959. Amer. Antiq., vol. 14, No. 4, pp. 292-300. Archeological research at the University of Kansas, 1946-1947. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 29-380. Field work in Kansas, 1949. Plains Archeol. Conf. News Letter, vol. 2, No. 4, pp. 5-6. Climate and archeology in Kansas. Proc. Sixth Plains Archeol. Conf. 1948, Univ. Utah, Anthrop. Pap., No. 11, pp. 98-99. Pottery types from the Talking Crow site, Fort Randall Reservoir, South Dakota. Plains Archeol. Conf. News Letter, vol. 4, No. 3, pp. 82-41, . Digging up the Plains Indian’s past. Univ. Kansas Alumni Bull. December. . Cartridges and bullets from Fort Stevenson, North Dakota. Plains Anthrop., No. 1, pp. 25-29. . An analysis of the firearms and related specimens from Like-a-Fish- hook Village and Fort Berthold I. Plains Anthrop., No. 4, pp. 3-12. Reconstructing a Plains Indian earth lodge. In Robert F. Heizer, “The Archaeologist at Work,” pp. 181-133. New York. The temporal relationships of coalescent village sites in Fort Randall Reservoir. Actas del XXXIII Congreso Internacional de Ameri- canistas, tomo II, pp. 111-123, figs. 1-3. San Jose. SauirH, G. Huperr. 1958. 1954. 1954. 1957. 1957. 1960. 1961. Indian trade beads from Fort Berthold, North Dakota. Central Texas Archeol. No. 6. Waco. Excavations at Fort Stevenson, 1951. North Dakota Hist., vol. 21, No. 3, pp. 127-1385. July. Archeological work at 32MUL2 (Like-a-Fishhook Village and Fort Berthold), Garrison Reservoir area, North Dakota. Plains Anthrop., No. 2, pp. 27-32. Archeological salvage at historic sites in the Missouri Basin. Prog- ress, Missouri Basin Field Committee, pp. 49-50. January—March. Billings. The present status of research on early historic sites of the Missouri Basin. Abs., Proc. Nebraska Acad. Sci., p. 3. April. Historical archeology in Missouri Basin Reservoir areas: Current investigations. Plains Anthrop., vol. 5, No. 10, pp. 58-64. Historie sites in the Oahe and Big Bend Reservoir areas. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 57. 661—-932—64_28 388 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] SM1TH, JAcK H.; LAF AVE, JACQUELINE M.; and Wik, Marcia V. V. 1961. Archeological resources of Vaquero Reservoir. Univ. California, Dept. Anthrop. and Sociol., Los Angeles. SOLECKI, RALPH S. 1949. An archeological survey of two river basins in West Virginia. West Virginia Hist., vol. 10, No. 3, pp. 189-211, and No. 4, pp. 319-482. 1952. Photographing the past. Progress, Missouri River Basin, Interior Missouri Basin Field Committee, pp. 1-9. September. Billings. 1953. A Plainview point found in Marshall County, Kansas. Plains Archeol. Conf. News Letter, vol. 5, No. 4, pp. 52-53. SPERRY, JAMES E., and Krause, RicHarp A. 1962. 1961 excavations at the Leavenworth site, 39CO9. Plains Anthrop., vol. 7, No. 16, p. 80. STALLARD, BRUCE. 1957. Report on talk by Jim Garner concerning removal of Nespelem Indian burials from Chief Joseph Reservoir. Washington Archeol., vol. INO. 8, D2 1c STANTON, ROBERT B. 1961. The Hoskaninni Papers: Mining in Glen Canyon, 1897-1902. Univ. Utah, Anthrop. Pap. No. 54 (Glen Canyon Ser. No. 15). STEEN, CHARLIE R. 1956. The archeological salvage program today. Archaeol., vol. 9, No. 8, pp. 175-181. STEPHENSON, RosBeErtT L. 1947. Archeological survey of Whitney Basin. Bull. Texas Archeol. and Paleont. Soc., vol. 18, pp. 129-142. 1948. Archeological survey of McGee Bend Reservoir. Bull. Texas Archeol. and Paleont. Soc., vol. 19, pp. 57-73. 1949. A note on some large pits in certain sites near Dallas, Texas. Amer. Antiq., vol. 15, No. 1, pp. 538-55. 1949. Archeological survey of the Lavon and Garza-Little Elm Reservoirs. Bull. Texas Archeol. and Paleont. Soc., vol. 20, pp. 21-62. 1949. A survey of the Whitney Basin in Hill and Bosque Counties, Texas. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 31-382. 1952. The Hogge Bridge site and the Wylie Focus. Amer. Antiq., vol. 17, No. 4, pp. 299-3812. 1954. Salvage archeology. Bible Archeol. Digest, vol. 9, No. 2, pp. 2-11. 1954. Taxonomy and chronology in the Central Plains—Middle Missouri River area. Plains Anthrop., No. 1, pp. 15-21. May. 1954. The Stansbury site: An historic site in the Whitney Reservoir, Texas. Abs. Proc. Nebraska Acad. Sci. April. 1957. Some research problems emerging out of Missouri Basin salvage. Ne- braska Acad. Sci., p.4. April. 1958. The Missouri Basin chronology program. Abs. Proc. Nebraska Acad. Sci., p.4. April. 1961. 1960 progress report, Missouri River Project. Plains Anthrop., vol. 6, INO: 125 pt. Alp. oD: 1962. Three Smithsonian salvage sites. Plains Anthrop., vol. 7, No. 16, pp. 80-81. 1963. Administrative problems of the River Basin Surveys. Amer. Antiq., vol, 28, No. 8, pp. 277-281. REPORTS IN OTHER SERIES 389 Stoney, GEORGE. 1950. Georgia’s archeologists before the camera. THarly Georgia, vol. 1, No. 2, pp. 18-21. Srrone, WM. DuncAN; JOHNSON, FREDERICK ; and WEBB, WILLIAM §. 1945. National archeological resources. Science, vol. 102, No. 2637, p. 44. July 13. SwaANson, Hart H., Jr.; TouHy, DONALD R.; and Bryan, ALAN L. 1959. Archeological explorations in central and south Idaho, 1958. Idaho State College Mus., Occas. Pap., No. 2. TREGANZA, ADAN HI. 1952. Archeological investigations in the Farmington Reservoir area in Stanislaus County, California. Univ. California Archaeol. Surv. Rep., No. 14. 1954. Salvage archeology in Nimbus and Redbank Reservoir areas, central California. Uniy. California Archaeol. Surv. Rep., No. 26. 1958. Salvage archaeology in the Trinity Reservoir area, northern California. Univ. California Archaeol. Surv. Rep., No. 43, pt. 1. 1959. Salvage archaeology in the Trinity Reservoir area, northern California, 1958 field season. Univ. California Archaeol. Surv. Rep., No. 46. TREGANZA, ADAN E., and HEICKSEN, MARTIN H. 1960. Salvage archeology in the Whiskeytown Reservoir area and the Wintu Pumping Plant, Shasta County, California. San Francisco State College, Occas. Pap. Anthrop., No. 1. TUNNELL, CURTIS D. 1961. Evidence of a Late Archaic Horizon at three sites in the McGee Bend Reservoir, San Augustine County, Texas. Bull. Texas Archeol. Soc., vol. 30 (for 1959) pp. 1238-158. TuoHy, DonaLp R., and Swanson, Haru H. 1960. Excavation at Rockshelter 10-AA-15, southwest Idaho. Tibiwa, Journ. Idaho State Coll. Mus., vol. 3, Nos. 1 and 2, pp. 20-24. WALLACE, WILLIAM J. 1960. Archaeological resources of the Buena Vista Watershed, San Diego County, California. Univ. California, Ann. Rep., Archaeol. Surv., Dept. Anthrop.-Sociol., pp. 277-294. Los Angeles. WEAKLY, Harry 9H. 1961. Current developments in Plains dendrochronology. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 59. WEAKLY, WARD F’. 1961. A site in the Fort Randall Reservoir, Brule County, South Dakota. Piains Anthrop., vol. 6, No. 14, pp. 230-241. 1961. 1960 excavations at the Leavenworth site, 39CO9. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 58. WEDEL, WALDO R. 1947. Prehistory and the Missouri Valley Development Program; summary report on the Missouri River Basin Archeological Survey in 1946. Smithsonian Misc. Coll., vol. 107, No. 6. 1947. The Missouri Basin Archeological Survey. Nebraska Hist., vol. 28, No. 1, pp. 32-40. 1948. Prehistory and the Missouri Valley Development Program: summary report on the Missouri River Basin Archeological Survey in 1947. Smithsonian Mise. Coll., vol. 111, No. 2. 390 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] WEDEL, Watpo R.—Continued 1949. Some provisional correlations in Missouri Basin archeology. Amer. Antiq., vol. 14, No. 4, pt. 1, pp. 328-329. 1949. A summary of recent field work in central Plains archeology. Proc. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Nebraska, pp. 3-5. 1950. Missouri River Basin Survey 1948 season. Proc. Sixth Plains Archeol, Conf., Univ. Utah, Anthrop. Pap., No. 11, pp. 3-8. 1951. The use of earth-moving machinery in archeological excavations. In Essays on Archeological Methods, Univ. Michigan, Mus. Anthrop., Anthrop. Pap., No. 8, pp. 17-28. 1961. Historic man on the Great Plains. Univ. Oklahoma. (Contains in- formation gathered by salvage operations. ) WEDEL, WALDO R., and GRIFFENHAGEN, GEORGE B. 1954. An English balsam among the Dakota aborigines. Amer. Journ. Pharmacy, vol. 126, No. 12, pp. 409-415. WEDEL, WALDO R., and Kivett, Marvin F. 1956. Additional data on the Woodruff Ossuary, Kansas. Amer. Antiq., vol. 21, No. 4, pp. 414-416. WENDORF, FRED; LUEBBEN, RALPH A.; Bruaer, Davin; and SCHROEDER, ALBERT H., with appendices by IRENE EMERY; Ear H. Morris; and Erik K. REep. 1953. Salvage archeology in the Chama Valley, New Mexico. Monogr. School Amer, Res., No. 17. WHEAT, JOE BEN. 1947. Archeological survey of the Addicks Basin: A preliminary report. Bull. Texas Archeol. and Paleont. Soc., vol. 18, pp. 143-145. WHEELER, RICHARD P. 1950. Archeological investigations in Angostura Reservoir, Cheyenne River Basin, South Dakota. Abs. Proc. Nebraska Acad. Sci. May. 1951. The archeology of the Boysen Reservoir, Fremont County, Wyoming. Abs. Proc. Nebraska Acad. Sci. April. 1952. A note on the McKean lanceolate point. Plains Archeol. Conf. News Letter, vol. 4, No. 4, pp. 45-50. 1952. Plains ceramic analysis: a check list of features and descriptive terms. Plains Archeol. Conf. News Letter, vol. 5, No. 2. 1953. The distribution, cultural relationships and chronology of mounds and earthworks in the Dakotas. Abs. Proc. Nebraska Acad. Sci. May. 1954. Selected projectile point types of the United States: II. Bull. Okla- homa Anthrop. Soc., vol. 2, pp. 1-6. March. 1954. Two new projectile point types: Duncan and Hanna points. Plains Anthrop., No. 1, pp. 7-14. 1954. Check list of Middle Missouri pottery wares, types, subtypes. Plains Anthrop., No. 2, pp. 3-21. 1954. New contributions to the archeology of Oahe Reservoir. Abs. Proc. Nebraska Acad. Sci. April. 1955. Recent archeological salvage operations in the Missouri Basin. Prog- ress, Missouri River Basin, Interior Missouri Basin Field Com- mittee, October-December, pp. 65-73. 1957. Archeological field data and their interpretation. Abs. Proc. Ne- braska Acad. Sci., p.4. April. 1958. A bibliography of the Indian archeology of the Central and Northern Plains. Abs. Proc. Nebraska Acad. Sci. pp.4-5. April. REPORTS IN OTHER SERIES 391 WHEELER, RICHARD P., and Smiru, G. HUBERT. 19538. The prehistory and early history of the Niobrara River basin. Mis- souri River Basin Project, Niobrara River Basin Development Plan, U.S. Dept. Int., Bur. Reclam., Reg. 7, Denver. WHITE THEODORE E. 1952. Observations on the butchering technique of some aboriginal peoples: I. Amer. Antiq., vol. 17, No. 4, pp. 337-338. 1952. Preliminary analysis of the vertebrate fossil fauna of the Boysen Reservoir area. Proc. U.S. Nat. Mus., vol. 102, No. 3296, pp. 185-208. 1952. Suggestions on the butchering technique of the inhabitants at the Dodd and Phillips Ranch sites in the Oahe Reservoir area. Plains Archeol. News Letter, vol. 5, No. 2, pp. 22-28. 1953. Studying osteological material. Plains Archeol. Conf. News Letter, vol. 6, No. 1, pp. 8-16. 1953. Bison steaks and venison chops—the Flint-age way. Montana Farmer- Stockman, p. 8. September 15. 1953. A method of calculating the dietary percentage of the various food animals utilized by aboriginal people. Amer. Antiq., vol. 18, No. 4, pp. 296-398. 1953. Collecting osteological material, or how to get a block plastered. Plains Archeol. Conf. News Letter, vol. 6, No. 1, pp. 3-7. 1954. Preliminary analysis of the fossil vertebrates of the Canyon Ferry Reservoir area. Proc. U.S. Nat. Mus., vol. 103, No. 3326, pp. 395-438. 1954. Observations on the butchering technique of some aboriginal peoples, Nos. 3, 4, 5, and 6. Amer. Antiq., vol. 19, No. 3, pp. 254-264. 1955. Observations on the butchering techniques of some aboriginal peoples, Nos. 7, 8, and 9. Amer. Antiq., vol. 21, No. 2, pp. 170-178. 1956. The study of osteological materials in the Plains. Amer. Antiq., vol. 21, No. 4, pp. 401-404. 1960. Plains quotes, “The Big Dogs of the Earth Lodge People.” Plains Anthrop., vol. 5, No. 9, p. 35. WILLIAMS, Bos. 19538. The Ward site, Le Flore County, Oklahoma. Oklahoma Anthrop. Soe. Newsletter, vol. 1, No. 9, pp. 2-9. WILMETH, ROSCOE. 1956. The Payne site. Mus. News, W. H. Over Mus., Univ. South Dakota, vol. 18, Nos. 11-12, pp. 18-21. 1958. Appraisal of the archeological resources of the Pomona and Melvern Reservoirs, Osage County, Kansas. Kansas State Hist. Soc. 1958. Report of the investigation of the Payne site, 39WW302, Walworth County, South Dakota, 1956. Archeol. Stud., Cire. No. 8. Pierre. WITHERS, ARNOLD. 1950. Survey in eastern Colorado—University of Denver. Proc. Sixth Plains Archeol. Conf., Univ. Utah, Anthrop. Pap., No. 11, pp. 10-11. Witty, THOMAS A. 1961. Excavations in the Wilson Reservoir area, Russell County, Kansas. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 64. 1962. Archeological salvage in the Milford and Council Grove Reservoirs. Plains Anthrop., vol. 7, No. 16, pp. 79-80. 661-932—64—_29 392 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Woop, W. RAYMOND. 1958. Additional data on the La Roche Focus. Abs. Proc. Nebraska Acad. Sci., May. 1954. Kipp’s Post, 32MNi1, Garrison Reservoir, North Dakota. Interim Report. State Hist. Soc., North Dakota. 1958. Excavations of a village site in the Table Rock Reservoir. Missouri Archeol. Soc. News Letter, No. 119, p. 3. February. 1961. 1960 field work at the Huff site, North Dakota. Plains Anthrop., vol. 6, No. 12, pt. 1, p. 56. 1961. The Pomme de Terre Reservoir in western Missouri prehistory. Mis- souri Archaeol., vol. 28, pp. 1-132. December. Woopzury, Ancus M., and Starr Mempsers of the Division of Biological Sciences, University of Utah. 1958. Preliminary report on biological resources of the Glen Canyon Reser- voir. Univ. Utah, Anthrop. Pap., No. 31. Woopzsury, ANGUS M.; DURRANT, STEPHEN D.; and FLOWER, SEVILLE. 1959. Survey of vegetation in the Glen Canyon Reservoir Basin. Univ. Utah, Anthrop. Pap., No. 36. 1962. A survey of vegetation in the Curecanti Reservoir Basins. Univ. Utah, Anthrop. Pap. No. 56 (Upper Colorado Ser., No. 6). WoopBpuRy, ANGUS M., ET AL. 1961. Ecological studies of the flora and fauna of Navajo Reservoir Basin, Colorado and New Mexico. Univ. Utah, Anthrop. Pap., No. 55 (Upper Colorado Ser., No. 5). Woo.wortH, ALAN R. 1954. A search in the past [River Basin salvage in North Dakota]. North Dakota Outdoors, vol. 16, No. 9, pp. 6-7. (North Dakota State Game and Fish Dept.) Bismarck. 1956. Archeological investigations at site 832ME59 (Grandmother’s Lodge). North Dakota Hist., vol 23, No. 2, pp. 22-86. WoORMAN, FREDERICK C. V. 1959. 1957 archaeological salvage excavations at Los Alamos, New Mexico: A preliminary report. Hl Palacio, vol. 66, No. 1. Santa Fe. INDEX Abraders, bone, 32 cancellous tissue, 120 (fig.) faceted, 32, 59 grooved, 32, 57, 59, 118, 114 (fig.) pumice, 209, 228 sandstone, 209, 228, 356 scoria, 60 stone, 113, 305 Adenauer, Frank J., 327 Agate, 25, 27, 106 Agatized wood, 31 Agricultural Complex, trait list, 226 Akaska Focus, 130, 181 Aksarben culture complex, 328, 334, 337, 3388, 342, 344, 347, 350, 353, 3638, 366, 367, 368 Amphibia (Rana sp.), 233 Anderson, John, 146 Anderson, Sidney, 327 Andreas, A. 'T., 847, 348 Animals, 283-234 effigies of, in baked clay, 24-25, 36 (figs.), 58 hides of, 210, 223, 225 remains of, 125 (list) sinews of, 223 wild, 223 See also Bones; names of specific animals. Antelopes, 118, 125, 154, 212, 218, 234 Antler, bands of, 57, 59 cylinder, 117 (fig.), 128-124, 133 fragments, 219-220, 229 miscellaneous shaped, 59 sections, 219, 361, 366 tines, 59, 60, 219 work in, 45-48 Anvil, used in pottery making, 90, 161, 192, 224 Anvil stones, 208, 228, 300, 337 Archaic Component, 326, 328, 335, 337, 357, 360, 368, 366, 367 Archaic horizon, 223, 339, 347, 366 “Archaic Mandan” Period, 52, 58, 61 sites, 57, 62 Archaic sites, 195 Archambault, George H., 73 Archer, figure of, 344 Architectural Complex, Hosterman site, 226 Arikara Indians, 208 basketry, 125 historic, 152 sites, 56, 126, 132, 193 Arrowhead, 330 Arrow points, 34, 60, 61 Hosterman site, 193-194 miscellaneous, 194 Arrows, 348 Arrowshafts, 209, 314, 352 smoothers, 286 straighteners, 209-210, 228, 340 wrenches, 2238, 276, 287 Artifacts, 14-51, 78, 88-126, 147, 275, 283, 284-287, 288, 296 antler, 117 (figs.), 119 (figs.), 123-— 124, 133 basketry, 111 (figs.) bone, 45, 111 (figs.), 116-124, 132, 212-220, 229, 274, 284, 286-287, 305, 306, 353 chipped stone, 105-112, 193-208. 304, 324 ground stone, 109 (figs.), 118, 114 (figs.)-116, 208-211, 304-805 lithic, 284, 285-286 metal, 229, 277, 353 pottery, 88-105, 160-193, 299-303 306 recent, 275, 277, 279, 282 shell, 117 (figs.), 120 (figs.), 124, 221-222, 229, 284 stone, 33 (fig.), 274, 279, 284, 286. 303-304, 306, 337, 340 unidentified, 332, 341, 360 wood, 111 (figs.) Artiodactyles, 51 (list), 125, 233 Arzberger site, 96, 129, 130, 131, 132, 133, 134 Ash, white, 4, 7, 9, 22, 127, 151, 153, 154. 192, 278, 282, 283, 361, 366 Ash Hollow Cave, Nebr., 310 Atlatl, 367 Awls, bird long bones, 41 (fig.), 43, 59, 60, 119 (fig.), 121 bone, 41 (fig.), 42, 43, 44, 45, 55, 59 (list), 120 (fig.), 121, 132, 209, 2138, 217, 223, 224, 353, 361, 366 fish spine, 41 (fig.), 438, 59 mammal long bones, 40 (fig.), 42, 61, 120, 121-122 mammal ribs, 41 (figs.), 48, 59 (list), 119 (fig.), 122 miniature, 119 (fig.), 122 neutral spine, 119 (figs.), 122, 132 rough splinter, 218, 229 scapula splinter, 121, 122, 213 split metapodial bone, 219 grooved, 54, 208, 209, 223, 228, 3386 stone, 34, 58, 365 Azure, Lee, 295 AX, Badger (Tavidea tarus), 125, 226, 233, 234 Badger Component (39LM238), 297, 808-314, 316 (list), 317 discussion, 314-815 Badger Creek, see Good Soldier Creek. 393 394 Badwater site, 57 Baerreis, David, and Dallman, John E., 133 Bag, buckskin, 125, 134 Balls, cancellous bone, 218, 229 stone, 210 Barka, Norman, 146 Barr, Robert P., x1u1, 73 Bartos, Stanley, Jr., 309 Basalt, 112, 312 Baskets, 111 (fig.), 125, 184 Bastian, Tyler, 146 Bastions, 273, 274, 275, 276, 277 Beads, bone, 59, 212, 353, 361, 366 ceramic, 25, 58 copper, 222, 229 disk, 48, 49 (figs.), 57, 59,120 (fig.), 124, 229 long bone, 38, 39 (figs.), 353 shell, 12-18, 57, 120 (fig.) stone, 32, 59 Beans, charred, 127, 222, 223, 307 kidney (Phaseolus vulgaris), 226, 232 red, 125, 134 Bear, black, 226, 233 head, model of, 25 jaw, 147 Beaver, 51, 226, 233, 234 incisor, 220, 229 model of, 24 Beef, dried, 210 Belemnite, 212, 228 Bell, Robert E., and Baerreis, David A., 356, 357 Benthall, Joseph, 295 Berries, 208, 209, 210, 222 Bienville, Jean Baptiste Le Moyne, Sieur de, 344 Biesterfeldt Cheyenne site, 1383, 134 Big Bend Reservoir, Lyman County, S. Dak., 271, 295 Birds, 51 (list), 125, 257 bones of, 212, 217, 218 sternum of, scored, 220 Bismarck, N. Dak., 1, 72 Bison, 51, 159, 226, 233, 234 bones, 12, 32, 35, 38, 44, 45, 50, 56, 125) “151, 15, LoS, oo; 2b2, 223, 235, 297, 299, 305 bone tools, 111 (figs.), 116, 133 chips, used for fuel, 151 fat, 224 incisor, grooved, 59 kills, 158 long bones, 216 metapodial, 287 radius, 219 ribs, 214, 215, 218, 287, 296 scored, 117 (fig.), 121 scapulae, 212, 213, 214, 217, 223, 282, 287, 299, 305 skulls, 11, 157, 158, 159, 220 vertebrae, 299 Black Eagle Creek, 73 Black Hoop, David, tribal chairman, XIV BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Black Partizan site (89LM218), 273 Blades, 342, 352, 356, 358, 364 fragments of, 342, 351, 360, 366 pyriform, 331, 336, 339, 340, 358, 364, 367 trapezoidal, 276 triangular, 352 Blake, Leonard, 231 Blum, Mark F., 73 Boatstones, limestone, 336 Bodkins, bone, 223 Boley, 57 Bone, fragmented, 78, 147, 308 human, 85 (list) mineralized, 219, 229 unmodified, 50, 51 (table) worked, 35-45, 37 (figs.), 39 (figs.), 41 (figs.), 48 (figs.), 59, 61, 353 See also Artifacts; Tools; and names of animals. Bones, animal, 12, 75, 212, 219, 223, 224, 225, 229, 216,277, 279;-281, 283, 342 mammal, 217, 306, 353 Bow and arrows, 158, 226, 342, 344, 367 Bower, John, 231 Bowers, Alfred W., 62, 63, 127 Boxelder, 125 Bracelets, antler, 119 (fig.), 124, 183 bone, 119 (figs. ) Brazell Creek, 360 Brother-of-All, Eugene, 295 Brown, W. L., and Anderson, Edgar, 231 Brule County, 8S. Dak., 312 Buckbrush, 74, 75 Buffalo, see Bison. Bull Creek, Lyman County, 8S. Dak., 306 Bulldozer, x111, 3, 74, 78 Bullroarer, 215 Bureau of Indian Affairs, 295 Burials, 74, 315 description of, 261-263 human, 12-14 primary, 310 secondary, 309 Burin, angle, 205, 206, 228 Burinlike instruments, 205 (fig.), 205- 206, 228 Burkitt, Miles Crawford, 206 Burned earth (F94), 7, 9, 151, 278, 281, 282, 288, 299, 308, 361 Caddoan group, 344, 350 Calcite, 34, 116, 310 Caldwell, Warren W., x11, 73, 271, 281, 295 Canids, 125 Canis sp., 216 Cannonball Focus, 62 Cannonball River, 55 Carbon-14 age determination, 50, 63, 131, 306, 308, 309, 311, 314, 315 Carbon stains, 281 Carnivores, 51 (list), 125, 233 Carter, Hugh, 231 Cass Lake, Minn., 354 INDEX Catfish, 51, 125, 236, 237, 316 blue, 236 bones, 299 spine, 42, 236 Catlinite, 211-212, 228 Cattail Creek, 3 (map) Celts, 34, 58, 223, 336, 341, 342, 365 diorite, 118, 114 (fig.), 286 ground, 276 Cemetery, 12 Central Plains, 55, 56, 57, 71, 1383, 161, 350 Cephalopod, extinct, 212 Chalcedony, 25, 60, 105, 110, 194, 196, 201, 203, 204, 240, 241, 242, 288, 285, sie Gilg: Badlands, 106 vein, 110, 112, 204 Chalk, 35, 115 Chamberlain, 8. Dak., 62, 272 Champe, John L., 62, 309, 310, 315 Charcoal, 7, 9, 50, 78, 80, 276, 281, 282, 2838, 288, 306, 308, 315, 361 Charles Mix County, 8. Dak., 309 Chase County, Kans., 324 Chatfield, D. William, 327, 354 Chert, black, brown, or gray, 25, 27, 29, 80, 31, 105, 106, 110, 112, 193, 194, 196, 201, 204, 207, 240, 241, 303, 324, 330, 331, 332, 335, 336, 337, 338, 339, 340, 341, 346, 351, 352, 356, 358, 360, 364, 365 cream-colored, 356, 364 Flint Hills, 330, 331, 332, 335, 336, 337, 338, 339, 340, 341, 345, 346, 351, 352, 356, 358, 360, 364, 365 pink, 331, 335, 336, 337, 339, 340, 345, 346, 351, 352, 356, 358 red, 335 white, 60, 332, 340, 352, 358 Cheyenne River, 193 Chipped-stone Complex, site, 227-228 Chippewa Indians, 56 Chisels, bone, 39 (fig.), 40, 59 stone, 200, 223 ulna, 111 (figs.), 123 Chokecherries, 222, 226 Choppers, 340 biface, 203, 228, 286 ovoid, 352, 365 pyriform, 360 stone, 30, 58, 112, 331, 336, 352, 360, 366, 367 uniface, 286 Chordates, 51 (list) Choutean Aspect, 129, 131, 1382, 134, 307, 315 Clay, miscellaneous objects of, 24-25, 58 (list) Clear Lake site, 311 Cleaver, scapula, 116, 117 (fig.), 132, 213-214, 228, 229 Cobbles, waterworn, 151 Cobblestone line, 299, 306 Collins, H. R., 205 112, 207, 287, 193, 239, 304, Hosterman 395 Committee for Recovery of Archeologi- cal Remains, 324 Concretion, tubular, 286, 314 Coogan, Alan H., 296 Coogan, Alan H., and Irving, William INespe Oo Cooper, Paul L., 71, 96, 180, 145, 192, 309, 311 Copper, 222 Cord, used on paddles, 92, 97, 99, 166, 170, 177 Cores, 340-341, 342, 346, 365 pyviform, 337 Corn, 348 braided, 134 kernels, charred, 50, 124, 125, 134, 222, 2238, 231 (table Northern Flint, 2381 stalks, 125 Cornecobs, charred (Zea mays), 50, 125, 231, 282 Corson County, 8. Dak., 71, 73 Cosner, Aaron J., 314 Coteau du Missouri, 245 Cottonwood, 151, 152, 328 Cottonwood formation, 324 Coyote (Canis latrans), 226, 233, 234, 316, 317 skull, 299 Crane (Grus canadensis) , 51, 125 Crane, H. R., and Griffin, James B., 310 Crow (Corvus brachyrhynchos), 51, 125 Crow Creek Indian Reservation, 295 Crystalline rock, 208 Cucurbita, 231, 232 Cupstones, 337, 838, 342, 346, 357, 365 Curtis, Norman, 353 Cylinders, antler, 133 Cyprinidae, 236 Dakota County, Nebr., 309 Dakota Indians, 208 Dalton gang, 353 Daly, Robert, 347 Danaczko, Edward A., 327 Dart points, stone, 54 Dawdles, clay, 191-192 Day, Adolph Burns, 295 Deer (Odocoileus), 51, 233, 234, 516, 317 antlers, miscellaneous objects of, 46, 47 (figs.), 118, 219 bones, 125, 212, 217, 218, 226 footprints, pecked, 344 leg bone, 353 tines, worked, 46 Demery Component, 185 pottery list, 105, 128, 129 Demery home, 74, 126 Densmore, Frances, 56 Dentry, Gordon, 146 Depressions, circular, 273, 276, 279 Deuder, Francis, 309 Dillenger culture, 311 Dinosaur National Monument, 316 Diorite, 34, 118, 209, 286 Disks, bone, 38, 39 (fig.), 56, 59 sandstone, 114 (fig.), 115 shell, 48, 49 (figs.), 57, 59, 61 396 Ditch, fortifying, 74, 146, 147, 148-149 (fig.), 225, 272, 273, 275, 276, 288 Dodd site, 55, 56, 182-133, 197, 213, 217, 220, Zola des Dog (Canis familiaris) , 233, 234 Canis sp., 51 domestic, 226 prairie (Cynomys ludovicianus), 226, 233, 234, 316, 317 Double Ditch (Bourgois) Mandan site, ING Dake. al, 56, 57, 15s 1% Drills, expanding base, 106, 304, 329, 352 stone, 29, 58 (list), 109 (figs.), 304, 329, 330-331, 335, 340, 342, 345, B52, 367 tapered, 109 (fig.), 115 Dry Creek, 347, 353 Dry Creek Rockshelter 347-353, 366, 367 Ducks (size of teal and gadwill), 51 Dunkle, David H., 316 Dunlevy, Marion L., 133 (14W 0224), Eagle, bones of, 118 flesh eaten, 220 tail feathers, 220 Hagle, golden, 51, 220, 233, 237 HKagle, Hobart, 295 Early Man sites, 195 Early Woodland cultures, 339 Earth lodge, 281 Effigies, baked clay, 36 (fig.) Elk (Cervus canadensis) , 51 antlers, 219 bones, 35, 125, 1383, 305 footprints, pecked, 344 ribs, 215 teeth, perforated, 46, 47 (fig.), 48, 56, 59 Ellsworth, Kans., 325 Elm trees, 323 Emetics, 56 End scrapers, 195-200, 249-255, 286, 312-313 earinated, 200, 247-249 chipped stone, 106, 109 (figs.), 331, 332, 336, 337, 342, 358 irregular, 286 keeled, 195, 197 (fig.) , 244-246 (list) stone, 27, 28 (figs.), 29, 58 (list), 60, 61, 287, 303-304, 312-313 teardrop, 250-251 (list) See also Scrapers. Entryway, covered, 127 Epidemic, 288 “Face,” miniature shell, 78, 120 (fig.), 124, 134 Fairbanks, Charles H., 226 Farming, 223 Faunal remains, 126 (list), 233-237 Feldspar, 90 Fence line, 283 Fenceposts, 279 Fenenga, Franklin, 54, 330 Fetish or trophy skull, 220, 229 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Figures, human, 348-344, 346, 347, 350 Finger Creek, 241, 354 Fire Heart Butte, 2, 73 Firepit, 282, 283, 306, 307, 308 Fireplaces, 3, 4, 6, 7, 8 (map), 9, 12, 75, 78, 79, 80, 82, 85, 86, 127, 222, 361 Mish; 51,5125, 2267 235, «236,n2okemela bones, 316 Fisher, Richard, 327 Fishhook, bone, 38, 39 (fig.), 59, 118, 119 (figs.), 216, 229 unbarbed, 216, 223 Fishing, 223 Flakers, bone, 214-215, 223, 229 tine, 124 Flakes, stone, 31, 112, 338, 342 worked, 314, 357 Flesher, metapodial, 282, 283, 286, 287 serrated, 35, 37 (fig.), 55, 59, 133 Flint, Richard Foster, 145 Flint Hill region, Kans., 324 Flints, 196, 239, 308, 313, 334 Floyd, Edwin, 146 Fort Berthold Reservation, §. Dak., 56, 125 Forth, Michael, 295 Fortifications, 276-279 Fort Lincoln site (32SI4), 177 Fort Randall Reservoir, 130 Fort Thompson, 295 Fort Thompson Focus, 272 Fort Yates, 2, 71, 73 Fort Yates Focus, 62 Fort Yates site, x11, 2, 3 (map), 62, 177 Fossils, shell, 48, 49 (fig.), 50 (table) Four Bears site (89DW1), 186 Four Mile Creek, 3 (map), 50 Fowl, 316 Fowler, Melvin L., 311 Fox, 226 gray (Urocyon cinereorgentus), 51, 226, 233 kit (Vulpes velox), 220, 234, 316, 317 model of, 25 red (Canis lupus), 226, 233 Fraser, Dorothy E., 145, 146 Friedmann, Herbert, 237 Fritz, George, 327 Fruits, wild, 223, 226 Gaines Ranch, 57 Game pieces, stone, 109 (figs.), 115 Gannon, Craig, 73 Gastropods, 51 (list) Gathering, 222, 223, 366, 367, 368 Gavins Point site (89YK203), 309 Geesefoot, 226 Gilmore, Melvin R., 125 Gipp, Robert F., x11 Gipp, William C., 73 Glass fragment, 281, 282 Golden, Bernard, 271 Good Soldier Component (39LM238), 297-305, 306, 307, 315, 316 (list), 317 INDEX Good Soldier Creek, 296 Goose (Branta canadensis), 51 Gopher, pocket (Geomys bursarius), 226, 233 Gouges, 200, 201 Gourd, bottle (Lagenaria) , 232 Grand River, S. Dak., 55, 89, 135 Grandmother’s Lodge site (382MH59), 52, 54, 62 Grainers, bone, 224, 229 Granite, 30, 31, 34, 58, 90, 118, 115, 163, 192, 208, 209, 224, 286, 308 Gravers, 207, 256 (list), 331, 367 tips of, 206, 208 Great Bend Culture Complex, 328, 334, 346, 347, 350, 353, 357, 367, 368 Great Plains province, 145, 306 Green stem fragments, 226 Greenwood County, Kans., 323, 324, 326, 328 Greenwood County sites 218), 324 Grinders, hide, 215 paint, 208 Ground stone, 113-116 spheres, 31, 33 (figs.), 59 tools, 208 Hackberry, 323 seeds (Celtis occidentalis), 232 Hadleigh-West, Frederic, 71, 73 Hagen site, 177 Hairdress, figure of, 350 Haiser, George, 12 Hall, Robert L., 309 Hammer, 34, 113, 220 antler, 229 (14GR201- 397 Hidatsa Indians, 133 earth lodges, 126, 127 sites of, 132 Hide grinders, 215, 229 Hide perforators, bone, 40, 45, 121 Hide serapers, 212, 213, 214, 224, 229 Hill, A. T., and Wedel, Waldo R., 133 Hoard, Lyon J., 130 Hoes, scapula, 35, 36 (figs.), 44, 45, 55, 59 (list), 111 (fig.), 116, 132, 212— 2138, 216, 223, 229, 282 stone, 341, 342, 346, 357 Hoffman, C. W., 56 Hold County, Nebr., 302 Homer, Dakota County, Nebr., 309 Hones, stone, 223 Hook and line, 226 Hopewell Complex, 325 Horse, figure of, 344, 346, 348, 353 Horticulture, 367, 368 Hosterman, John B., 145 House: depressions, 2, 11, 12, 60, 85, 146, 271, 2738 entrance, 4, 7, 9, 51, 78, 81, 85, 1 127, 152, 1538, 154; 156, 222; 223, 306 fill, 3, 12, 127, 128 floors, 2, 3, 4, 7, 59, 81, 84, 127, 148, 152, 153, 154, 156, 160, 225, 279, 306 midline, 51 pits, 127, 147, 148 (map), 153, 222 sites, 74, 306 superstructure, 153 walls, 4, 7, 12, 51, 52, 84, 128, 152, 153, 154, 155 128, 223, Hammerstones, 31, 58, 113, 210, 223, 331, | Houses, 51, 61, 74, 78-85, 126, 151-154, 226 337, 342, 346, 360, 365 discoidal, 113 ovoid, 286 pebble, 113, 208 Hand County, S. Dak., 309 Handles, 184-186, 188 (figs.) lug, 183, 184-186, 192 strap, 183-184, 186, 192 Handstones, 210, 228 Harding, Lester, 327 Havens site, 57, 60-61, 62 Hawk, 51, 125 Heart Butte campsite (82GT1), 175, 177 Heart Butte rockshelter (82GT5), 177 Hearth, 148 (map), 152, 153, 154, 306, 361 Heart River, 61, 62 Heath hen (Tympanuchus cupido), 51 Hecker, Thad. C., 1, 61 Hematite, 208, 211, 296, 299, 308, 360, 367 and other paint substances, 211-212, 228 Java Brown (8110), 35, 115 red, 192, 308, 332, 366 siliceous, 341 Hensler site, 57 Hewes, Dr. Gordon W., x11, 1, 4, 11, 12, 14, 19, 54, 68, 100, 175, 177 Hickey Ranch, 272 circular, 75, 126, 151, 158, 215, 223 conical, 127 Demery site, 75, 78, 79 (map), 80— 85, 92, 105, 124-128 Havens site, 60 oval, 128 Paul Brave site, 2-11, 22, 51, 52, 58 (list) rectangular, 154, 222, 271, 283, 287, 306, 309 Thomas Riggs site, 51, 58 (list) Howard, J. Dan, 73 Howard, James H., 62, 124, 134, 288, 326, 327, 328, 330, 331 Howard, Mrs. James H., 327 Howe, Donald, 295 Hrdlicka, Ales, 261 Huff Focus, 129, 130, 131, 134, 135 Huff site (32MO11), 51, 63, 125, 129, 132, 133, 134, 288 Hughes County, 8. Dak., 306 Human remains, 85-86, 221 Hunting, 222, 223, 226, 235, 366, 367, 368 camp, 159 Hurt, Wesley R., Jr., 2, 35, 51, 52, 53, 56, 62, 96, 127, 130, 181, 132, 138, 171, 175, 179, 284, 287, 309 Huscher, Harold A., 295 398 Huscher, Harold A. and McNutt, Charles H., 295 Husks, 125 Ictaluridae, 236 Illinois River, 311 Incisor, grooved, 47 (fig.), 48 Indian dwelling, figure of, 353 Indian Knoll skeletons, 260 Insect remains, 238 Inter-Agency Archeological Salvage Program of the Smithsonian In- stitution, 295, 324 Irving village, 311 James, Edwin, 348 James, Jesse, 353 Jamison, Fred, 327 Jasper, 194, 196, 199, 201, 204, 239, 240, 241, 308, 318, 340, 351, 365 John Grass, Dakota chief, 73 John Grass Creek, 73, 76 (map), 127 Johnson, Alfred E., 324, 325, 326, 327, 330, 331, 334 Johnson, Elden, 309 Johnston, Francis E., 260 Jones, Hugh Carl, 146 Juniper, 152 Kanapolis Reservoir, 325 Kansa Indians, 323, 344 Kansas City Hopewell culture, 326, 328, 332, 367, 368 Kearney, Milo, 295 Keeler, Charles L., 259 Keith Focus, 310, 325, 326, 364, 366, 367 Kenel, S. Dak., 57 Kenel to Fort Yates road, 73, 76 (map) Kettle, iron, 353 Kidder, A. V., 40, 121, 132 Kivett, Marvin F., 56, 310, 339, 364 Kleinsasser, Glenn, 14, 59, 284 Knife handles, bone, 37 (fig.), 44, 56, 59, 110, 111 (fig.), 118, 125, 134, 215, 229 wood, 110, 111 (fig.), 118, 125 Knife River flint (chalcedony), 25, 27, 2930, sols) GUS Glen 105. JO) 112: 303 Knives, 41 (fig.), 44, 213, 283, 285-286, 287, 304, 313-314, 339-340, 352, 358 asymmetrical, 108 (figs.), 112 Badger Component, 313-314 beveled, 366 bifaced, 203, 204 bone, 44, 56, 59, 116, 117 (fig.), 132, 212, 223, 224 broad, 29-80, 58 (list), 110 chipped stone, 108 (figs.), 224, 228 flake, 30, 58, 61, 112, 204, 352 four-edged, 203, 204 Harahey, 340, 352 hook-bladed, 56 leaf-shaped, 285 narrow, 30, 48, 54, 58 (list), 110 oval, 110, 116 ribbon flake, 204 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Knives—Continued scrapula, 116, 117 (fig.), ‘182, 213, 229 skinning, 196 squash, 56, 116, 132 stone, 28 (figs.), 54, 203-204 thin-bladed, 204 trapezoidal blade, 286 ? vein chalcedony, 108 (figs.), 112, 203, 204 Knox County, Nebr., 302 Krogman, W. M., 258, 261 Lagomorpha, 233 Landes, Mr. and Mrs. Kenneth, 327 La Roche Focus, 130, 131 La Roche site, 97, 128, 129, 130, 133, 134, 179 Larson culture, 177 Larter site, 203 Later Heart River Period, 61 Lawrence, Russell B., x1It Leaners, 52 Leather work, 134 Le Beau Focus, 171 Lehmer, Donald J., 1, 15, 44, 55, 56, 132, 152, 161, 197, 215, 284, 302, 317 Libby, Orin G., 56 Like-a-Fishhook Village, 127 Limestone, 314, 332, 341, 342 Cottonwood formation, 336, 341, 346, 357, 360, 365 Permian, 324 slabs, 361 Limonite, 332, 366 Lincoln, Nebr., 162 Lindenmeier site, 206, 256 (list) Lithie materials, 285-286 Little Missouri River, 54, 62 Little-Water-Man, supernatural being, 354 petroglyph (14W0226), 347, 353- 354 Livingston, Jerry, 295 Lodge, circular, 74, 126 eagle-trapping, 127 earth-covered, 52, 61, 282, 297, 305, 307 Log, charred, 281 Love, August, 327 Lower Brule, reservation town, 272 Lower Loup Focus, 133 Lyman County, S. Dak., 62, 214, 271, 272, 296, 306 MacNeish, Richard S8., 196, 203 McDermott, John Francis, editor, 323 McKern, Thomas W., and Stewart, T. D., 260-261 McLaughlin, Jack, xm MeNutt, Charles H., x11, 301 Madison, Lee G., 271, 281 Maerz, A., and Paul, M. Rea, 35 Maize (Zea mays), 226 Mandan culture, 61, 63, 127, 131, 133, 134, 152, 220, 231 sites, 55, 56, 61, 126, 132 INDEX Mandan-Hidatsa culture, 177 Mano, stone, 34, 55, 58, 115, 8336-337, 341, 360, 365, 366 Maples Mill culture, 311 Marion County, Kans., 324 Marshno, Joseph, 327 Mason County, IIl., 311 Mauls, grooved, 31, 58 (list), 113, 208, 209, 224, 228 Mealing slab, 33-34, 55, 58, 115 Meat, dried, 208, 209, 223 Medicines, 337 Meleen, Elmer B., 130, 183, 134 Metcalf, George, 213, 214 Meyer, Winton, 231 Mica, 90, 116 Microblades, 194, 204, 227, 243-244 Middens, refuse, 60, 151, 152, 213, 223, 225,205 Middle Missouri area, 1, 14, 15, 57, 62, (il eo Byer es {yey bs Middle Tent, Frederick, 295 Middle Woodland time level, 367 Miller, Carl F., 259 Miller, Loye, 50, 125 Miller, Ruth, 146 Mississippi Valley, 231 Missouri Basin Chronology Program, Missouri Basin Project, Smithsonian Institution, 62, 71, 73, 131, 145, 146, 271, 273, 295, 327 Missouri Botanical Gardens, St. Louis, 231 Missouri Plateau section, 145 Missouri River, 2, 3 (map), 50, 52, 55, 60, 61, 63, 71, 73, 112, 127, 128, 129, 130, 184, 185, 145, 153, 222, 259, 271, 272, 295, 296, 306, 307, 311, 315 drainage basin, 327 flood plain, 2, 78 terrace, 1, 296 Missouri River Valley, 1, 61, 62, 245 Mitchell Creek site, 56 Moat, protective, 147 Mobridge, S. Dak., 56, 132, 133 Mole, meadow, 233 Mollusks, 50, 51 (table) Mongoloids, 258 Mooney, James, 344 Moorman, Edward H., 324, 326, 327 Morris County, Kans., 324 Morrison, Joseph P. E., x11, 50, 221 Mortars, wooden, 152, 154 Motsiff site, 56, 57 Mullers, 115 Mulloy, William, 177 Muskrat, 51, 226, 233 Myers site, 130 Nakasa tribe (Caddoan group), 344 National Park Service, Dept. of Interi- or, XIU, 71, 295, 324, 327 Nebraska culture, 337 Needlelike tool, 215, 229 Needles, bone, 59, 352-353 Nelson, Wayne, 295 399 Neosho Focus, 356, 357 Neuman, Robert W., 73 Newell, H. Perry, and Krieger, Alex D., 89 Nickerson, Norton H., 316 Nickerson, N. H., and Hou Ding, 231 Niobrara River, Nebr., 89 North Cannonball sites, 57 North Dakota State Legislature, 71 Northern Plains tribes, 1, 56 Nuts, 209 Oahe Reservoir, N. Dak., x1, 61, 71, 130, 259 Oak, blackjack, 347 trees, 323 Obsidian, smoky, 112, 193, 194, 239, 241 Ocher, red, 34, 38, 166, 211 Ojibwa Indians, 354 Oldham site (89CH7), 309 Omaha Indians, 231, 310 On-a-Slant Mandan village, 133 One Mile Creek, 3 (map) Oneota site, 133, 338 Orleans Aspect, 366, 367 Osage Indians, 323, 344 Osterholt, Dale, 146 Otter (Lutra canadensis), 226, 233 Outlaw Rockshelter (14W0O225), 3538 Over, H. R., and Thomas, C. S8., 237 Over Focus Swanson site (89BR16), 51, Paddle, 90, 128, 161, 165, 178, 192, 224, 300 cord-wrapped, 53, 90, 92, 128, 161, 165, 177, 192 grooved, 53, 128, 161 Paint, 332, 366 brushes, 211 red, 344 stones, 208, 228 See also Hematite. Paint Creek culture, 334, 357 Paleo-Indian material, 366 Palisade, 150-151, 225, 272 posts, 277 wall, 275, 277 Pashe, H. W., 327 Paul Brave estate, xIv Paul Brave site (32814), xu, 1, 2, 3 (map), 4, 53, 54, 56, 57, 60, 61, 62, 63, 100, 104, 125, 132, 1383 See also Fort Yates site. Pawnee Indians, 208 Payne site, 130, 131, 183, 224 Pebbles, unmodified, 34 utilized, 210-211, 228 Pecking stones, 208 Pelecypods, 50, 51 Pelican, white, 125 Pemmican, 210, 224 Pendant, bone, 120 (fig.), 123 rib, 59 shell, 48, 49 (fig.), 59, 221, 222 wapiti tooth, 229 Perforators or gravers, stone, 206-207, 228 400 Perfumes, 337 Petrified wood, 27, 30, 31, 106, 110, 112 Petroglyph, Little-Water-Man (14WO 226), 847, 353-354 Petroglyphs, 341, 342-344, 346, 347, 348- 350, 353 sites of, 326, 341-342 Phillips, George, 327 Phillips, Richard, 327 Phillips County, Kans., 326 Phillips Ranch site (89ST14), 55, 133, 231 Picks, bison radius, 111 (fig.), 123, 132, 219 bone, 229 stone, 365 ulna, 111 (fig.), 123 Pierre, 8. Dak., 133, 192, 257 Pigments, 35, 115, 189 Pine, scrub, 347 Pins, bone, 182 Pipe bowl, 211, 286 Pipes, baked clay, 58 eatlinite, 114 (fig.), 115, 211 eatlinite disk-bowl, 109 (fig.), 114 (fig.), 115 fragments, 276 Pisces, 233 Pi-Sunyer, Oriol, xm Pits, 3, 78, 83, 297 aboriginal, 283 ash, 148 (map) basin-shaped, 4, 7, 9, 52, 75, 86, 127, 151, 1538, 154, 157 (fig.), 158, 299, 308 bell-shaped, 11, 52, 75, 80, 82, 8&4, 85, 86, 87, 124, 127, 148 (map), 152, 154, 156, 157 (fig.), 160, 223, 307 borrow, 86, 156, 159 eache, 3, 52, 148 (map), 151, 154, 156-157 (fig.), 160, 190, 209, 211, 212, 219; 220; 12238, 279, 283, 307 circular, 151, 152, 159 fire, 148 (map), 151, 152, 158, 156 irregular, 9, 52 midden, 159-160, 192, 212, 214, 219, PPA PPE PLATA OS UIT oval, 80, 86 quasi-rectangular, 159 rectangular, 9 rubbish, 52, 60, 75, 82, 152, 154, 156, 276 saucer-shaped, 156, 159 storage, 52, 156, 160, 223 test, 11-12, 59, 74, 75, 78, 146, 276, 283, 296, 329, 330, 331, 332, 337, 348, 354, 357, 361 undercut, 152 Plains Indian groups, 220, 348, 360 Plains Woodland Period, 315, 316, 328, 351, 358, 363, 364, 366, 367, 368 Plums, wild (Prunus americana), 222, 232, 342 Point Peninsula Focus, 221 Pomme blanche, 125 152, 205, 238, 154, BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Possum Point site (14W0228), 327, 354-357, 355 (map), 366, 367, 368 Postholes, 3, 4, 6 (map), 7, 9, 12, 75, 79, 85, 127, 148, 151, 152, 155, 159, 160, 276, 277, 281, 297, 299, 306 Post molds, 277, 281, 282, 288, 297, 361 Posts, 78, 79 (map), 80, 82, 84, 125, 127, 154, 281, 283, 297 cedar, 7, 11, 299 charred, 9, 85 Paul Brave site, 4, 52 stockade, 150 (map) wall, 152, 222 Potsherds, 271, 277, 284, 296, 297, 299, 334, 346, 350, 354, 357 Potter County, 8S. Dak., 145, 259 Pottery, 92-105, 1638-165, 211, 226, 284— 285, 287, 299-303, 306, 308-311 Aldren cord impressed, 55 Anderson, 287, 288 Anderson flared rim ware, 287 Anderson low rim, 279, 284, 285 appendages, 15, 17, 20, 21, 53, 90, 95, 97, 98, 100, 101, 102, 103, 161, 300, 302, 303 Archaic Mandan, 20, 52, 58 Arikara, 193 Arzberger Group, 130, 131 Arzberger horizontal incised, 130 Arzberger opposed diagonal, 130 Badger Component, 308-311, 314 bases, 224, 300 biconical perforations in, 24 black painted, 188-189 body sherds, 23-24, 60 (list), 89, 90, 91-92, 100, 162, 163 (list), 165-167, 285, 308, 329, 330, 334, 338, 344-845, 350-351, 356, 363 brushed, 162, 163, 166, 171, 173, 177, 178, 224, 285 Cadotte collared, 299, 301-302, 307 Campbell Creek types, 132, 307 Chamberlain incised triangle, 309 check-stamped, 24, 53, 55, 58, 89, 92, 130 color, 53, 90, 160, 165, 192, 224, 284, 285, 300, 301, 302, 309, 310, 330, 334, 338, 350, 351, 354, 363, 368 cord-impressed, 103-105, 162, 166, 171, 173, 174, 175, 177, 178, 185, 192, 224-225, 308, 309, 310, 3380, 334, 3388, 342, 344, 347, 350, 363, 367 cord-paddled, 315 cord-roughened, 53, 55, 58, 60, 89, 92, 97, 99, 180, 131, 276, 285, 287, 288, 347 corncob-impressed, 163, 167, 190 Cowley plain type, 345, 350 curvilinear-stamped, 162 decoration, 17 (fig.), 18 (figs.), 19, 20, 21, 22, 23, 58, 55, 89, 92, 96, 97-98, 99, 100, 101, 102, 103, 104, 129, 162, 165, 167-171, 172 (fig.), 175, 224, 284, 285, 299, 300, 301, 302, 308, 356, 363, 367 Demery site, 90-105, 128 dentate stamping, 330, 332, 367 INDEX Pottery—Continued Hvans incised, 97 Evans indented, 302 fabric-impressed, 162, 167 finger-indented, 170, 171, 172 (figs.), 177, 2838, 284 fingernail-pinched, 163, 192 firing of, 160 forms, 15, 16, 19, 20, 21, 22, 53, 89, 90, 98, 96, 98, 99, 100, 101, 102, 103, 104, 174, 175, 177, 300 Fort Rice cord-impressed, 100, 101, 105, 129 Hort Rice trailed, 101, 105, 129 Fort Yates cord-impressed, 20-21, 54, 55, 58, 59, 60, 100 Fort Yates Cross-Hatched, 19, 21, 54, 55, 58, 60 Fort Yates fine incised, 19 Fort Yates plain, 19 Fort Yates ware, 16, 20-21, 24, 53 fragments, 55 (list) Geneseo plain type, 338, 345, 354 Geneseo red filmed, 334 geometric designs on, 182 grass-marked, 345, 356 Great Bend, 350, 357 Great Oasis, 309, 315 Grey Cloud horizontal-incised, 97, 131, 299, 300-301, 307 grit-tempered, 89, 163, 192, 300, 301, 302, 303, 309, 310, 330, 334, 338, ore 344, 345, 346, 350, 363, 367, 36 handles, 24, 53, 90, 97, 161, 170, 183— 186, 188 (figs.), 334 hardness, 90, 284, 285 Harlan cord-roughened, 310 herringbone incised, 168 (fig.), 170, 173, 174, 175, 177, 178, 179, 184, 185, 191, 299 horizontal parallel line patterns on, 179, 184, 190, 300, 301 horizontal stab and drag design on, 184 Hosterman cord-impressed, 177 Hosterman site, 160-193 Hosterman vertical rims, 177-178 Huff, 101 Hughes Group, 130, 131 incised or trailed, 58, 60, 89, 90, 91, 92, 162, 168, 168 (fig.), 170, 171, 178, 174, 178, 182, 192, 224, 302, 330, 363 Iona diagonal-incised, 307 Iona horizontal-incised, 97, 98, 132, 307 Iona indented, 296, 299, 300, 307 Iona S-rim, 307 Iona ware, 132 Le Beau cord-wrapped rod designs, AT Le Beau finger-indented, 171 Le Beau horizontal cord-impressed, 175, 178 line-block zonal decorations, 176, 179, 186 (fig.), 187 (fig.) 401 Pottery—Continued lip form, 90, 167, 168 (figs.), 169 (figs.),; 171, 174, 175, 177, 284, 285, 300, 302, 309, 310, 350, 356 lip treatment, 167-171, 224, 284 making, technique of, 128, 224, 300 Mandan, 192 mat-impressed, 162 miniature vessels, 22-23, 161, 189- 190 (figs.), 224 modeling tools, 44, 56, 59, 61, 119 (fig.), 122-123 Nordvold horizontal-incised, 131, 166 (fig.), 179 notched-stick decorations on, 310 painted, 53, 1638, 182, 188-189 paste, 15, 24, 53, 89, 90, 99, 103, 104, 160, 163, 174, 177, 300, 311 Paul Brave, 14, 15, 53, 58 (list) pinched ridged, 163 plain, 162, 163, 165, 168 (fig.), 171, 174, 175, 177, 178, 182, 192, 276, 279, 285, 299, 330 Plains, 301 polished, 90, 350, 351 punctated, 163, 168 (figs.), 170, 175, 177, 179, 182, 224, 301, 302 recurved rims, 101-103, 105 red-filmed, 89, 92, 166 Riggs cross-hatched rim, 19, 21, 55, 58, 60, 104, 105, 128, 129 Riggs flared rim, 279, 283, 284 Riggs incised rim, 19-20, 58, 55, 58, 60 Riggs pinched rim, 20, 55, 58, 60 Riggs plain rim, 16, 17 (figs.), 18 (figs.), 58, 54, 55, 58, 59, 60 Riggs punctate, 54, 55, 58, 59, 281, 282, 283, 284 Riggs straight rim, 104, 105, 128, 129 Riggs ware, 16-20, 24, 53, 58 (list), 284, 287, 288 Riggs wavy rim, 54 rim decoration on, 175-180 (figs.), 224 pubis aay, als} ey, Bh val Is Wal (figs), 92, 129, 162, 173, 174, 177, 284, 285, 300, 356 rim sherds, 128, 162, 283, 284-285, 287, 300, 301, 302, 306, 307 (list), 808, 310, 329-330, 334, 338, 350, 354, 361-3638 rim types, 89, 91 (figs.), 93 (figs.). 94 (figs.), 129, 173, 174 (table), 363 Rio Grande glaze paint, 334 rocker-stamped, 330, 332, 367 Russel plain rim, 301 Russell ware, 301 Rygh rainbow corded, 175, 177, 178 S-shaped rim decorations, 172 (figs.), 173, 174 (table), 175, 177, 185 (fig.), 192 sand-tempered, 354 shapes, 53, 161 shell-tempered, 89, 342, 344 shoulder decorations on, 183, 184 (fig.) , 224 402 Pottery—Continued shoulder elements on, 93 (figs.), 94 (figs.), 95 (figs.), 97, 99, 100, 101, 161 shoulder treatment, 182-183, 300 simple-stamped, 23, 53, 55, 58, 59, 60, 89, 90, 91, 92, 98, 128, 130, 161, 162, 163, 165, 173 (fig.), 177, 182, 183 (fig.), 192, 224, 285, 297, 299, 306 smoothed, 58, 59, 60, 90, 92, 128, 161, 334, 338, 344, 345, 346, 367 Stanley braced rim ware, 302, 307 Stanley tool impressed, 299, 302, 307 surface finish, 15, 58, 89, 90, 97, 98, 99, 100, 101, 102, 103, 104, 165, 174, 177, 192, 284, 285, 300 Talking Crow brushed types, 179 Talking Crow straight rim, 91 (figs.), 92, 96, 105, 129, 131, 132, 299, 301, 307 Talking Crow ware, 301 tempering, 53, 90, 160, 224, 284, 285, 800, 334, 354 texture, 90, 160, 164, 284, 285, 300 ticked-line, 180-182 (fig.) tool impressed, 171, 182 (list), 301, 302 BUREAU Twelve Mile black-on-gray, 55, 58 unclassified, 16, 21-22, 58, 60, 61 Valley cord-roughened, 310 vertical rim sherds, 178 (table) vessel shape, 15, 177, 186-188 Weaver ware, 311 Wheeler horizontal-incised, 96, 131, 132, 179 Wheeler incised-triangle, 132 Wheeler plain rim, 96 Wheeler ridged, 96 with flaring rims, 96, 97, 99, 105, 130, 297 Woodland, 100 Prairie chicken (Tymfanuchers ameri- canus), 24 Pre-Arikara group, 208, 226 Preceramic remains, 325 Prehistoric Omaha, 309, 315 Pre-Mandan group, 226 Projectile points, 2389-242 (list), 276, 285, 297, 303, 311-3138, 329, 330, 332, 334-836, 338-339, 340, 342, 845, 351, 354, 356, 358, 360, 363- 364, 366 chipped stone, 54, 105-106, 107 (figs.), 193-194, 223, 225, 227 expanding stem, 342, 345, 347, 351, 356, 357, 358, 367 fragments, 106, 330, 332, 335, 352, 357, 358, 363 large point, 335, 339, 345, 352, 356, 357, 367 late horizon, 335, 339, 356 plain lanceolate, 25, 27 plain triangular, 27 small point, 335, 345, 352, 356, 367 stone, 25, 26 (figs.), 27, 58 (list), 224, 357 OF AMERICAN ETHNOLOGY [Bull. 189] Pottery—Continued triangular, basal notched, 303 triangular, corner-notched, 312 triangular, straight base, 303, 311- 312 triangular blade, concave base, 106, 285, 303 triangular blade, straight base, un- notched, 106, 303, 351 triangular blade, two side notches, 106, 285 Pronghorns (Antilocapra americana), 212, 216, 226, 233 metapodial bone, 216, 217, 219, 229 Prunus americanus, 317 Pumice, 209, 223, 228 Pumpkin, sugar, 232 Pumpkin seeds, charred, 50, 231 Punches, antler, 219 bone, 45, 59, 215-216, 224, 229 Quartz, 30, 31, 90, 303, 304, 312 Quaritzite, 30, 31, 106, 110, 112, 113, 193, 194, 196, 201, 204, 207, 208, 239, 240, 241, 242, 286, 308, 312, 313, 314, 331, 341, 364 Bijou Hills, 30, 112, 282, 283, 285, 304, 314 Quill flatteners, 44, 282, 287 Rabbit, 226, 317 cottontail (Sylvilagus floridanus), 51, 125, 233, 234, 316 jack (Lepus townsendii), 51, 125, 238, 234 Radiocarbon dating, 63 Rafters, 222 Redbird Focus, 97, 130, 133 Reed, Erik K., 327 Ree Heights, Hand County, 8. Dak., 309 Ree Heights Buffalo Kill site (89HD3), 309 Reid, Russell, x11, 71 Reptiles, 51, 125 Ribs, cut, 215, 229 fragments of, perforated, 229 notched, 215, 229 tip of, slotted, 45, 59, 117 (fig.), 119 (fice) seb22 el oe Rice County, Kans., 334 Ridgepole, 52 Rifle, 344 Rim sherds, 59, 60 Ritchie, William A., 221 River Basin Survey, x11, 324, 326, 327 Roberts, Frank H. H., Jr., 206, 327 Robert Zahn site (32813), 57-60, 62 Robinson, Stephen W., xu, 73 Rock fragments, 297 Rocks, fire-cracked, 308 Rockshelter, Dry Creek, (14W0O224), 347-353, 349 (map), 366, 367 Outlaw (14W0225), 353 Walleye (14W0222), 326, 341-347, 343 (map), 366, 367 Rodentia, 233 INDEX Rodents, 51 (list), 125 Roof, dirt, 222 poles, 127, 152 supports, 127, 151, 154, 222 Roots, 125, 223 Rough-stone Complex, Hosterman site, 227 Rubbing stones, 208, 228 Rubbing tools, bone, 45, 59 Rutherford Mound, 311 Ryan site (25DK2), 309 Sample, Mark, 327 Sandstone, 31, 32, 33, 55, 57, 113, 114, 115, 209, 286, 328, 336, 337, 341, 342, 344, 346, 348, 352, 354, 357, 360, 365 Sdtwa, Arikara work basket, 125 Scalp Creek, 97 Scalp Creek site, 130, 132, 133, 179 Scapula, worked, 216, 229 Scapula spines, 2138, 214, 229 Scheans, Daniel J., 62 Schist, gray micaceous, 336, 341 Scoops, horn core and frontal, 59 split metapodial, 37, 38, 59 Scoria, 32, 60, 113, 304 Seraper-burin combination, 205 Scrapers, 195-203, 227-228, 286, 329, 367 bone, 212, 223 bowshaft, 340 circular, 195 elliptical, 195 end, 195, 196, 197-201, 205, 207, 228, 244-247 (list), 248-255, 282, 283, 286, 314, 340, 345-346, 366 flake, 112, 195, 198, 282, 331, 336, 337, 338, 340, 342, 346, 352, 360, 365, 366, 367 ovate, 195 ovoid, 203 planoconvex, 195, 198 pointed oval, 196 scapula, 214 shell, 117 (fig.), 124, 134, 221 side, 194, 195, 198, 201-203, 228, 244 (list), 314 single edged, 195 snub-nosed, 223 stone, 225 subtriangular, 195 teardrop, 195, 198 thumbnail, 195, 199 turtleback, 364-865 unifaced, 195 Sereen, 153 Seeds, 208, 222, 226 Setzer, H. W., 233 Shaft scrapers, 340 Shaft smoothers, 32, 33 (fig.), 59, 114 (fig.), 115, 223, 282, 304, 342, 352 Shaft wrenches, 59, 117 (fig.), 118, 215, 229 Shale, 113 Shannon Focus, 129, 130, 131, 307, 315 Sharpening stone, 352 Sha-we, Mandan women’s game, 56 403 Shells, 184, 221 (table)—-222, 277 Anodonta grandis plana, 51, 221, 226 deposits of, 146 “Goniobasis” (fig.), 50 Lampsilis cardium, 51 DL siliquoidea, 51 Lasmigona complanata, 51, 221 DES Vg PY Lioplaw nebrascensis, 48, 49 (fig.) modified, 120 (fig. ) Nucula planimarginata, 50 Olivella biplicata, 221 Oyztrema insculpta, 49 (fig.), 50 snail, 57, 59 Succinea grosvenori, 221 Tancredia americana, 50 unmodified, 50, 51 (table) unworked, 221, 229 Viviparus leidyi, 49 (fig.), 50 V. retusus, 49 (fig.), 50 work in, 48-50, 59 (list) Sherds, 12, 147, 161, 224, 274, 277, 278, 281, 283 See also Pottery. Sheyenne-Cheyenne site, 56 Shotwell, J. Arnold, 50, 125 Sickles, bone, 212, 223 scapula, 214, 229 Siltstone, 31, 55 Siouan tribes, Thegiha-speaking, 323 Skeletons, analysis of, 261 Skulls, 220, 229, 235 Skunk (Mephitis hudsonica), 51, 226, 233, 234 Slabs, grooved sandstone, 57 limestone, 361 Slant site (82M026), 56, 57, 133 Slaughtering or butchering areas, 157- 159 Slow, Horace, 295 Smith, Carlyle S., 62, 96, 129, 132, 162, 180, 301, 306, 307, 325, 327 Smith, Carlyle S., and Grange, Robert T., Jr., 96, 98, 129, 130, 131, 133, 296, 299, 300, 301, 306, 307 Smith, G. Hubert, 295 Smoky Hill River, 325 Smoothers, stone, 223 Solecki, R., 205 Southern Cult items, 134 Sower, J. E., 327 Spain site (89LM301), 96, 98, 130, 131, 133, 300, 306, 307 Spaulding, Albert C., 89, 96, 130, 132, 133, 134, 197, 337 Spatula tip, bone, 38, 39 (fig.), 59 Spears, 158 Sperry culture, 177 Spinden, Herbert J., 1 Spokeshaves, 218 Spoons, brass, 353 shell, 221 Spotted Bear site, 127, 133 Squash, 226 seeds, charred, 50, 222, 223, 231 nebrascensis, 49 404 BUREAU OF AMERICAN ETHNOLOGY [Bull. 189] Squirrel, ground (Citellus sp.), 51, 125, | Tools—Continued 226, 233, 234. Standing Rock Indian Agency, 71 Standing Rock Indian Reservation, xiv, 73 Standing Rock Tribal Council, x11, x1v, 73 Standing Soldier site (82818), 62 Stanek, James J., 295 Stanton, N. Dak., 57 Stanton Ferry, 57 State Historical Society of North Da- kota, xIII, xIv, 53, 56, 57, 59, 60, (Ga Wey fs aang (2 Be ss Steamboat Creek site (39PO1), 186 Steen, Charlie R., 327 Steinbrueck, Emil R., 57 Stephenson, Robert L., x111, 73, 131, 295, 327 Stewart, Evelyn B., 295 Stewart, T. D., 261 Stockade trench, 146, 147, 279 Stone, fragments of, 332 problematical object of, 314 unworked, 116, 276, 277, 281, 333, 341 worked 25-35, 58 (list), 59 (list), 60-61 Stringers, wooden, 127 Strong, William D., 56, 192, 193 Structures, ceremonial, 12 discussion of, 51-52, 126-128 Swan, whistling (Cygnus columbianus), 51 Swan Creek site (39WW7), 130, 131, 171, 186, 217, 224 Swanson site, 309 Swanton, John R., 344 Talking Crow site (89BF3), 96, 132, 301 Tanner, Lloyd, 327 Tattoo designs, 344, 350 Taylor, William, 236 Tazwell County, Il., 311 Teeth, beaver, 220 elk, 46, 47 (fig.), 48, 56, 59 human, 85, 221, 229 wapiti, 221 worked, 220-221, 229 Tefft, Gaylord S., 327 Thomas Riggs Focus, 62, 63, 85, 89, 104, 105 (list), 128, 129, 180, 131, 133, 135, 288 Thomas Riggs site (39HU1), 2, 14, 51, 52, 53, 54, 55, 57, 62, 63, 182, 284, 287, 306 Thong stretchers, bone, 35, 59, 212, 213, 214, 216, 229 Timbers, charred, 75 milled, 281 Tines, deer, 46 Tipsina (Psorales esculenta), 125, 134 Tissue abraders, cancellous, 118 Tony Glas site (832EM3), 62 Tools, antler, 352 bone, 12, 212, 229, 352 digging, 123, 212, 213 flake, lamellar, 207-208, 228 flaking, 366 flint, 12, 223 ground and pecked stone, 208-211, 228 lamellar flake, 207-208, 228, 255 (list) needlelike, 215 pentagonal-shaped, 216-217 scapula, 1382 sharpening, 346 wooden, 352 Toronto, Woodson County, Kans., 323, 347 Toronto Dam, 323, 324 Toronto Reservoir, 323, 324, 325, 326, 327, 328, 347, 366, 368 Toronto Reservoir sites, 368 (tabula- tion ) Trade goods, European, 334, 347 southwestern, 334 White, 315 Tree ring dating, 63 Trench, exploratory, 147, 148, 275, 2838 Paul Brave site, 11, 52 stockade, 146, 147 test, 275-276, 361 Tribal Council, Standing Rock Indian Reservation, 73 Trotter, Mildred, and Gleser, Goldine C., 258, 261 Tubers, 223 Tubes, bone, 37 (fig.), 40, 56, 59, 118, 119 (fig.), 212, 217, 229 Tulsa District, Kans., 823 Turkey, 24, 226 bones, 212 Turtle, 51, 125 Twelve Mile Creek site, 56 Two Teeth site (39BF204), 301, 307 Underwater Panther, mythical figure, 348 University of California, 125 University of Kansas, 62, 306, 307, 324, 327 University of Michigan, 63, 308, 309 radiocarbon laboratory, 63, 181 University of Minnesota, 3809 University of Nebraska, 131, 309 laboratory of anthropology, 309 State Museum of, 327 University of North Dakota, xIII, xIv University of Oregon, 125 Upper Plains, 191 Upper Republican, 56, 133, 222, 325, 326, 337 Upper Sanger, 57 Valley Focus, 310 Vegetal remains, 50, 231-232, 317, 342 Verdigris River, 323, 332, 337, 341, 360, 361 Vertebrae, 235 Vessels, bowl shaped, 338 exotic, of possible Caddoan extrac- tion, 190-191 globular, 189, 297, 306 miniature, 9, 22-238, 92, 189-190 (figs.), 191 (table) INDEX Vessels—Continued Plains Woodland, 310 pottery, 12, 80 soot deposit on, 187, 224 two-spouted, 191 (fig.) See also Pottery. Villages, area of, 74, 129, 226 farming-hunting, 306 Paul Brave site, 58 (list) Thomas Riggs site, 58 (list) unfortified, 61 Vole, meadow, 226 Wallace, Wayne O., 327 Walleye Rockshelter (14W0224), 326, 341-347, 343 (map), 366, 367 Walls, 222, 281 posts for, 11, 52 wattle-and-daub, 52 Wapiti tooth (Cervus canadensis), 221, 229 Waterfowl, 226 Watermelons, 146 Webb, Albert, 327 Webb, George, 327 Wedel, Waldo R., 44, 55, 56, 132, 133, 208, 332, 334, 338, 345, 350, 357 Wedel, Waldo, and Kivett, Marvin F., 310, 326, 354 Weeks, Richard, 257 Weinreich, Bernard, x11, 72 Wellington, J. W., xiv, 71 Wheeler, Richard P., 44, 56, 146, 161 Whetstones, 209 Whistle, bone, 38, 39 (fig.), 59 405 White, Theodore, 159, 233, 235, 316, 317 White River, 62 Whitlocks Crossing, 145, 257 Wichita culture, 334 Wichita Indians, 344, 357 Wilford, Lloyd A., 309, 315 Will, George F., 1, 61, 63, 187 Will, George F., and Hecker, Thad. C., 1, 2, 20, 51, 52, 57, 60, 61, 62, 63, 74, 175 Will, George F., and Spinden, H. J., 55, 57, 61, 133) 175, 225 Will, Oscar H., 125 Willey, Gordon R., 129 Willow, black, 125 Wilmeth, Roscoe, 130, 131, 133, 179 Wilson, Gilbert L., 127 Wilson, Ione, 295 Winona, 3 (map) Wood, W. Raymond, 50, 56, 71, 97, 125, 129, 130, 133, 184, 302 Wood, charred, 342 See also Petrified wood. Woodland Indians, 348 Woodland Pattern, 326 Woodland remains, 325, 347, 357, 863 Woodruff Ossuary, Kans., 310, 326 Woodson County, Kans., 323, 326, 328, 347 Woodson County site, 324 Woolworth, Alan R., 52, 54, 62, 71 Missouri River, Yankton County, S. Dak., 309 Yellow Back, Junior, 295 O rR ay tena Se. he, eo Sst: he rectage "eB ousstir mee Vig aes TaEAS daca eae thy " OS ROR LA Bait ge: Swag Gua eeBO RD Wwe oytos SEAT Abode? ol ier wt eat if nh — SO Deine Ge OS St} Bi ataton Pe ee ae arr ay } ‘ wis {7 Gut etal 72 Bags, Ae oras®) LEM) Peacoat ce DEL ATEIE) ARTs vi 4 we 1 75, TS, WS 'ESt 1 reso JG vrteyte f Bay i bis H aah 108) SHOW, h Bota “seanaee) . ol tee Bald wollte] eS, 3 inh haan Het costa cegrth ye [rata Seer mate a i TIC ch area roth | Lia Rt t We snokooall VA fi eit ey of paipiam Be yom vi (outa) Baton! | (Soot Satine Ly a 3 Ae tei Jie it * OVE boo! H BP) See Say 4 ‘ G05 WG, er OBE OS rine ah ty, Bet ‘hae, - * : hy be) »? ; ear tad “DOO Th | nes rah ea Soe Bek! st n on raoA ‘ Ww YAS Star ernhnl Dash ee yA | ' aes eee tT haribeo Vi hice Ghee get hee, OS ST ES. ih a itenior ba athegitn ae, pers Pe OL OIE 2und cisueeO Tihoow: 8 peer ees ov bite She Pt WES pastasl iGOU). 10 nbuove | WADA WF Ss OSE?) WER Dany) Oe | Tai | Pi SP TO ide ROG tin gavoln coabooy? [eset ST i ce: sine, ITS 2s SB foe. divowinin te } : » . 7-4 », ‘ {ve roa are Ws Paley iar pag GOS pete gi ib wry tied | GUNN) 8 889 | * gt C) S ty f F i es bin. , ' f : . ¥ * 43 } - s a) . 1 ° ~ iss 4 ‘ y ei ? * ‘sg pws * Lies | aes pea es | : ry ay" xi Tip At ¢ ( LL a ‘ 4 2 one “i ti ‘ 4 ' ,' ae e/ > ORs ; te Sep ert KS rrt' .) 1? aie, beret Sic radars q ‘cand at ps i ' «< | ~ 4 TIT LIBRARIES TI it | Mt ‘|