A \\ \\ AS s \ ONY ~ SN TO OSS RM} }Hrwowy) WY RW AX re Pigs anol SMITHSONIAN INSTITUTION BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 RIVER BASIN SURVEYS PAPERS Frank H. H. Roserts, Jr., Editor Inter-Agency Archeological Salvage Program NUMBERS 9-14 UNITED STATES GOVERNMENT PRINTING OFFICE WASHINGTON : 1958 For sale by the Superintendent of Documents, U. S. Government Printing Office Washington 25,D.C. - Price $3.25 LETTER OF TRANSMITTAL SMITHSONIAN INSTITUTION, Burrav or AMERICAN ErHNotoey, Washington, D. C., March 29, 1957. Sir: I have the honor to submit the accompanying manuscripts, en- titled “Archeological Investigations in the Heart Butte Reservoir area, North Dakota,” by Paul L. Cooper; “Archeological Investigations at the Tuttle Creek Dam, Kansas,” by Robert B. Cumming, Jr.; “The Spain Site (89LM301), a Winter Village in Fort Randall Reservoir, South Dakota,” by Carlyle S. Smith and Roger T. Grange, Jr.; “The Wilbanks Site (9CK-5), Georgia,” by William H. Sears; “Historic Sites in and Adjacent to the Jim Woodruff Reservoir, Florida- Georgia,” by Mark F. Boyd; “Six Sites near the Chattahootchee River, in the Jim Woodruff Reservoir area, Florida,” by Ripley P. Bullen, and to recommend that they be published as a bulletin of the Bureau of American Ethnology. Very respectfully yours, Frank H. H. Roserrs, Jr., Acting Director. Dr. Leonarp CARMICHAEL, Secretary, Smithsonian Institution. II Newer Gf bt kes 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 recovering archeological and paleontological remains which would otherwise 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 undertaking, the Smithsonian Institution organized the River Basin Surveys 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 pertain- ing 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 Department of the Interior appropriations. The operations of the River Basin Sur- veys, Smithsonian Institution, have been supported by funds trans- ferred 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. Frank H. H. Roserts, Jr., Director, River Basin Surveys. Tit PUBLISHER’S NOTE A separate edition is published of each paper in the series entitled “River Basin Surveys Papers.” Copies of Papers 1-14 are available at the Bureau of American Ethnology, Smithsonian Institution, and can be had free upon request. RIvER BASIN SURVEYS PAPERS PUBLISHED PREVIOUSLY No. 1. Prehistory and the Missouri Valley Development Program: Summary No. No. No. 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. . 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. . The Woodruff Ossuary, a Prehistoric Burial Site in Phillips County, Kansas, by Marvin F. Kivett. Bull. 154, pp. 103-141, pls. 16-28, figs. 2-8. 1953. . The Addicks Dam Site: I. An Archeological Survey of the Addicks Dam Basin, Southeast Texas, by Joe Ben Wheat. Bull. 154, pp. 148-252, pls. 29-47, figs. 4-28. 1953. II. Indian Skeletal Remains from the Doering and Kobs Sites, Addick Reservoir, Texas, by Marshall T. Newman— Bull. 154, pp. 253- 266, figs. 24-28. 1953. . The Hodges Site: I. Two Rock Shelters near Tucumeari, 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. 285-302, figs. 31-35. 1953. . The Rembert Mounds, Elbert County, Georgia, by Joseph R. Caldwell. Bull. 154, pp. 303-320, pls. 55-56, figs. 36-40. 1953. . Archeological Investigations in the Oahe Dam area, South Dakota, 1950-51, by Donald J. Lehmer. Bull. 158, 190 pp., 22 pls., 56 figs., 6 maps. 1954. . 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. Iv CONTENTS Poreword, \by-Prank HoH. Roberts, Jrs220 Ufo. 8 FU seeOeo alk No. 9 Archeological investigations in the Heart Butte Reservoir area, nothin Makorn, jay aul Cooper. 9 22.2 oo eo eek No. 10. Archeological investigations at the Tuttle Creek Dam, Kansas, RECOMCER Mier CUMMNITIES (OI oe oy ees Se ee eee ee No. 11. The Spain site (839LM301), a winter village in Fort Randall Reser- voir, South Dakota, by Carlyle S. Smith and Roger T. Grange, Jr____- No. 12. The Wilbanks site (9CK—5), Georgia, by William H. Sears_____-_ No. 13. Historic sites in and adjacent to the Jim Woodruff Reservoir, Hignda-Georgia, by Markl. Boyd... 222222 -2---6--6 oe decone nn cow No. 14. Six sites near the Chattahoochee River in the Jim Woodruff Reser- vor area. lorida, by Ripley P. Bullen. 2 +2022. be oe Appendix. List of reports, articles, and notes relating to the salvage Brograa Wublisned an other serieszis. 24.202! 222) Shee eS roms & re ery A it — . + te Ge ees | WF j ing Lnofpatons Lf F . ‘ * 7 trey 22 ATAU - bs i Sf.) ht RET ee ea os aa cae ee tt Dcbteoh wel ue tate iets Hit ree Pah ot di coe Pets ail enti’ Seahd) is iT ae ear ae SDV? iF ‘fake wal oy ee ol. aaa SeeeBG Gee Aral, Boeoa i Sie tf aera ed eek ee ihe ae ee cay of = os eu “ED eared Gabriel ; ye “tn f tie hee SE ateyot fi, ie a “ - , : ike Ms Re eo Siyetk hee ‘af bod ile , > a, r FOREWORD The six reports which form the contents of this volume of the River Basin Surveys Papers are based on the results of field investigations carried on as a part of the Inter-Agency Archeological Salvage Pro- gram. Three of the articles are concerned with projects in the Mis- souri Basin and three with studies made in the Georgia-Florida area. Three reservoirs were involved in the Missouri Basin and two in Geor- gia-Florida. The work at two Missouri Basin reservoirs was done by field parties under the direction of members of the staff of the Missouri Basin Project of the River Basin Surveys, Smithsonian Institution, with funds transferred from the National Park Service. The third party was from the University of Kansas and was operating under a Memorandum of Agreement with the National Park Service whereby some moneys were made available to assist the university in its investi- gations. One of the projects in Georgia was a cooperative contribu- tion on the part of the University of Georgia which bore the entire cost of the operations. The other two investigations in the Southeast were carried on under Memoranda of Agreement between the National Park Service and the Florida Historical Society and between the Serv- ice and the Florida State Museum, University of Florida, with financial assistance from the Service. In each case where a Memorandum of Agreement was concerned, the report submitted was in partial fulfill- ment of the agreement and received the approval of the Regional Di- rector for the region in which the investigations were made. The archeological investigations in the Heart Butte Reservoir area in North Dakota were under the direction of Paul L. Cooper and were carried on during the summer field season of 1948. Owing to insuffi- cient funds, it was not possible for Mr. Cooper to accomplish as much as should have been done in that reservoir basin. However, the infor- mation and materials which he obtained there do constitute a contribu- tion to the archeological knowledge of the area. The dam has long since been completed and the basin flooded. The investigations at the Tuttle Creek Dam in Kansas were under the supervision of Robert B. Cumming, Jr., a member of the Missouri Basin Project staff. The particular excavations which Mr. Cumming made were truly of a salvage nature, because the sites tested were on the dam axis and construction work was then under way. Lack of funds and sufficient time prevented more extensive investigations, but the results obtained were better than expected. The Tuttle Creek area vir VIII FOREWORD is receiving further attention, and before the dam is completed and the basin is flooded good information should be obtained from the numerous sites located in that area. The excavations at the Spain site in the Fort Randall Reservoir area in South Dakota were supervised by Dr. Carlyle S. Smith of the University of Kansas, assisted by Roger T. Grange, Jr., who at that time was associated with the Chicago Natural History Museum. The work was carried on during the summer field season of 1953 and most of the members of the party were college students. Dr. Smith and Mr. Grange completed their report in July 1954 and submitted it to the Regional Director of Region Two of the National Park Service at Omaha, Nebr. After its subsequent acceptance and approval the man- uscript of the report was submitted to the River Basin Surveys of the Smithsonian Institution for publication. In view of the delay in send- ing the report to the printer, it was returned to Dr. Smith in the autumn of 1956 so that he might make certain revisions and bring his conclusions up to date. He returned the revised manuscript to the River Basin Surveys in December 1956. The report on the Spain site provides new and useful information on certain archeological aspects in that portion of the Missouri Basin. The Fort Randall Dam was closed in the summer of 1953 and the site is now under water. When it appeared that funds for archeological excavations in the Allatoona Reservoir area in Georgia might not be forthcoming in time to salvage information from sites in the construction area for the dam, the University of Georgia provided for Some digging at one of the more important locations and a party under the direction of Dr. William H. Sears, then associated with the Department of Anthropology at the university, conducted excavations there in 1949. That site is now beneath some 90 feet of water. Subsequently funds were appropriated and transferred from the National Park Service to the Smithsonian Institution and an extensive series of excavations in the Allatoona Reservoir area got under way. ‘Two parties from the River Basin Surveys excavated 11 additional sites and tested 19 others. Originally it was contemplated to include Dr. Sears’ report in the major publication on the Allatoona Reservoir. Because of the size of the main report and the attendant delay in arranging for its printing, it was deemed best to make Dr. Sears’ results available at this time rather than to hold them indefinitely, The cooperation of the Uni- versity of Georgia was indeed helpful at a critical period in the Alla- toona basin investigations, and the results obtained add a useful chapter to the Allatoona story as a whole. The study by Dr. Mark F. Boyd of the historic sites in the Jim Woodruff Reservoir area in Florida and Georgia is the first report based mainly on historical records and evidence to be issued in the River Basin Surveys Papers. Reports of investigations at historic FOREWORD | ix sites in other areas will be issued in later bulletins. While not as much mention has been made of the historic aspects of the salvage pro- gram as of those pertaining more to the aboriginal remains, they nevertheless are an important part of the overall investigations and a number of studies of that nature are being carried forward in various reservoir basin areas. Dr. Boyd’s project was sponsored by the Florida Historical Society under a Memorandum of Agreement with Region One of the National Park Service. In commenting upon the report by Dr. Boyd when it was forwarded to the Region One office, the President of the Florida Historical Society stated that in the opinion of the society the report was an important contribution to the preserva- tion of the history of that area. The report was accepted and ap- proved by the Regional Director and was subsequently released to the River Basin Surveys for publication in the present volume. The archeological investigations pertaining to aboriginal sites made along the Chattahoochee River in the Florida portion of the Jim Woodruff Reservoir area were under the direction of Ripley P. Bullen of the Florida State Museum and were carried on under a Memo- randum of Agreement between the National Park Service and the University of Florida. The field work was done in June 1953. The report was completed and submitted to the Region One office of the National Park Service and was accepted and approved by the Regional Director in December of that year. Early in 1954 the manu- script was released to the River Basin Surveys. New information de- veloping in the interim between its receipt and scheduling for publica- tion made certain additions and revisions advisable and it was returned to Mr. Bullen who resubmitted the new draft in October 1956. Plans were to include both Mr. Bullen’s and Dr. Boyd’s reports in a general publication on the Jim Woodruff Reservoir area which would include the reports on archeological investigations in Indian sites in the Georgia portion of the basin. One series of excavations there was carried on by the National Park Service and another by a member of the staff of the River Basin Surveys. Since there has been some delay in completing those reports, it has been deemed advisable to issue the papers by Dr. Boyd and Mr. Bullen in the present volume. Frank H. H. Roserts. JRr., Director, River Basin Surveys. ae ~ anh 4 STIR Oth, SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 169 River Basin Surveys Papers, No. 9 Archeological Investigations in the Heart Butte Reservoir Area, North Dakota By PAUL L. COOPER 7 ‘ ee ‘a ter ae oe ee WNT STITSA BARMORE TRG ‘ f i 2, — ‘ i Porininis CAPISTAA 1] WRI, eee oe we ri ovr vay ! 2 Es ee oe Te RT Ne el iy Aan - a J Th aise a hos P| , = 7 Oye Le eee nk ir ook ee oe : 9 OK Mest ons of enOnRavaorn nol — CONTENTS 3 PAGE (eo Sih en's & aa Ne ee ee Ee eee 5 a ee ee ee 5 AVeTAGTO GL ULC DEON ote spe em ee EE spe ah ate esa ee En 9 Bem OE MlerrSI ben (ese Calley eae er rete eee ie yen ue ee al eel ce 12 LEDS SOD HEEN BCS HTS te RIA La de Ee ope ee pean ea se 12 PAT GIT ACLS enters een ae ee eee een eae See eed Slee 20 NO Ghernye ee er at ee Seat ee cee ee Se ee 20 WiGnkeilistOnGste = aaa. ee ee ee 2 ee eee eee le 26 Work inebonevanGramt lene ewes ae mtn eee Wee eee eh ye ee 29 aunalpremaing’ 05 en eee Mine SAR Rear a ye eS ke ae soso 31 CUE CSTE TT EST Seen ee TI eS OY UR een py «Le ee erage 32 TAT) SEE ea cPEI GN ETS bp a eel UE eat ae agile Oh te os eet lh, Spall A Ai ips Napali gee cag a 34 Appendix. Sites in the Heart Butte Reservoir area___.._.._.-_.--------- 37 ILLUSTRATIONS PLATES FOLLOWING PAGE lene Wwoton Wochlen siten(o2Gulyl p22 = pss Se ee 40 Dm Koehlemsite (2G EL) duringvexcayvation. — 22-2 a 40 3. Koehler site (32GT1). a, Profile on W15 line, N50 to approximately N62, with zones marked. b, Typical profile of N45 line, W70.3 to W75.2, with zones marked. c, Profile on W15 line, N85 to N92, showing deposit of mussel shells, Feature 29, and beneath it, shallow pit, Feature 43. d, Pottery fragments, Feature 28, in situ. e, Pottery fragments, Reature $0; in sit. 222-2. 22h eet as 40 4, Koehler site (82GT1). a, Typical profile on N47.5 line. 5, Profile of fireplace, Feature 35, on W15 line. cc, Profile of fireplace, Feature 7, COLLIN Se SEN SARI PERS Oy EP S L, b F re 40 5. Koehler site (32GT1). a, Top of cultural deposit in excavation unit 1, square N170W50, showing nature of cultural debris. 6, Discarded unidentifiable bones recovered from Feature 19. c, Top of deposit of mUSselshells; WeatUTe ZO. se te eee a ene ee een See 40 6. Pottery rim sherds from the Koehler site (82GT1)___________------ 40 7. Pottery sherds from the Koehler site (82GT1)________-_-_-_--_---- 40 8. Stone artifacts from the Koehler site (32GT1)____________-___-_-_-- 40 9. Chipped stone artifacts from the Koehler site (32GT1)_________--_- 40 10. Large blades and ‘‘choppers” from the Koehler site (32GT1)______-_- 40 11. Bone and antler artifacts from the Koehler site (32GT1)______-____- 40 12. a, Site 82GT5, rock shelter (arrow), to southeast. 6b, Site 32GT5, rock shelter, to northeast. c, Artifacts from Koehler site (82GT1) andirock shelter, site 3267 S232.) et Aces eves te hee 40 TEXT FIGURES are 1, Plan of excavation units 1 to 4 (X1—X4), Koehler site (32GT1), showing locations of fireplaces and other features___.....___.____._-------- 14 2. Characteristic profiles in Koehler site (82GT1)___________----------- 17 MAPS 1. Map of Heart Butte Reservoir showing locations of archeological sites _ 9 2. Contour map of Koehler site (32GT1) showing locations of excavations_ 12 ARCHEOLOGICAL INVESTIGATIONS IN THE HEART BUTTE RESERVOIR AREA, NORTH PAKO RAT By Paut L. Cooper FOREWORD The investigations upon which this report is based were carried out as a part of the inter-agency salvage program in the Missouri Basin and reflect the cooperation of a number of agencies and indi- viduals. The work was instigated by the plan of the Bureau of Re- clamation to construct the Heart Butte Dam, a unit of the compre- hensive water-resources development program under the Pick-Sloan plan. The dam, now completed, is on the Heart River south of Glen Ullin, N. Dak., and is designed to create a reservoir of almost 11,000 acres at its maximum elevation of 2,118.2 feet above mean sea level. In August of 1946, J. Joseph Bauxar and the writer, archeologists with the Smithsonian Institution River Basin Surveys, spent 2 days in reconnaissance of a small part of the area that would be flooded if the proposed Heart Butte Dam was constructed. Only a few sites were recorded, but one of these, the Koehler site, was deemed suffi- ciently important to require intensive investigation. Rather exhaus- tive excavation of this site and additional survey of the reservoir were recommended in an appraisal report.?, In 1947, under an agreement with the Smithsonian Institution, the University of North Dakota Field Session in Archaeology, cosponsored by the State Historical Society of North Dakota, devoted a week, June 25 to July 1, to the Heart Butte Reservoir area. The party of 8, supervised by Dr. Gordon W. Hewes, excavated 8 test pits, each 5 feet square, in the Koehler site and found a few previously unrecorded sites in the vicinity. The limited tests in the Koehler site confirmed the earlier judgment that full-scale excavation would be worth while. In 1948, with the dam already under construction, limited funds were available for further investigations there by the Smithsonian Institution. These funds were far less than the amount recommended as necessary for an ade- quate sampling of the archeological remains that were to be destroyed, and permitted only 7 weeks in the field with a party of from 2 to 4 1 Manuscript submitted July 1954. 2 Preliminary appraisal of the archeological and paleontological resources of Heart Butte Reservoir, Grant County, N. Dak., June 1947, prepared by River Basin Surveys, Smithsonian Institution, for Missouri Basin Recreation Surveys, Region T'wo, National Park Service. 5 . GNSTITUTION JAN 2 7 1989 6 BUREAU OF AMERICAN ETHNOLOGY [Buy. 169 individuals. Approximately 12 man-days were devoted to recon- naissance, 2 man-days to the investigation of a rock shelter near the dam site, and the remainder of the time to work in the Koehler site. The little more than 100 man-days spent in excavation, recording, mapping, and refilling trenches at this site represent about 20 percent of the original estimate of the time that would be required for inves- tigation on a desirable scale. The writer arrived in the Heart Butte area June 12, 1948, and was assisted in reconnaissance until June 19 by the late Thad. Hecker, of the State Historical Society of North Dakota. Robert L. Hall and Warren L. Wittry, then students at the University of Wisconsin, re- ported on June 22 and worked for the remainder of the time spent in the area. Vernon Gerving of Glen Ullin, employed from July 6 to July 30, completed the roster of paid workers. The writer’s wife, Dorothy Thurlow Cooper, was with the party after July 1 and assisted in the field. Most of the time after June 22 was devoted to the Koehler site, but between July 20 and the termination of the Heart Butte opera- tions on July 31 the investigations in the rock shelter were accom- plished and some reconnaissance was carried out. In addition to the workers in the field, all of whose services were eminently satisfactory, a number of individuals, as well as several agencies, have been of assistance in various ways. Funds to carry on the investigations were transferred to the River Basin Surveys by the National Park Service. Personnel in the Bismarck_and Glen Ullin offices of the Bureau of Reclamation freely provided information and maps. Camping space and other facilities were made available at the Government camp in Glen Ullin. W. W. Brenner, construction engi- neer, and B. L. Mendenhall, field engineer, both of the Heart Butte project office, were especially helpful. Landowners in the area of investigation and various residents of Glen Ullin were uniformly co- operative and rendered assistance of various kinds. We are especially indebted to August Koehler, owner of the site to which we have given his name, for permission to excavate on his property and for other courtesies, and to W. F. Salzer, a resident of the reservoir area, who guided us to sites and presented the River Basin Surveys with speci- mens which he had collected in the vicinity. Dr. Hewes supplied us with information relative to his investigations and made helpful sug- gestions based on his experience in the Koehler site. The State His- torical Society of North Dakota, through its superintendent, Russell Reid, not only made the services of Mr. Hecker available for several days, but also has loaned the Surveys the specimens collected and records made by Dr. Hewes in 1947. Material assistance was provided by Alan Woolworth of the Historical Society staff. Identifications of the unworked bones from the Koehler site were made by Dr. Theodore E. White, paleontologist with the River Basin lay. e's] «= HEART BUTTE RESERVOIR AREA—COOPER 7 Surveys, and of the mollusks by Dr. J. P. E. Morrison, of the United States National Museum. Dr. Charles M. Riley, Department of Geology, University of Nebraska, assisted in the identification of the stone material, and Dr. C. Bertrand Schultz, director of the University of Nebraska State Museum, identified the worked bone. Members of the laboratory staff of the River Basin Surveys headquarters in Lincoln were instrumental in the production of this report in diverse ways. Raymond S. Price prepared the final copies of the map, ground plan, . and profiles, and Nathaniel L. Dewell photographed the artifacts. 422141582 i salt gall iovrgesS otto m1adHt 16 INTRODUCTION The Heart River rises in Billings County, N. Dak., only a few miles from the northward-flowing Little Missouri River, and follows a tor- tuous course to its confluence with the Missouri at Mandan, approxi- mately 120 miles due east. Its total length les within the Missouri Plateau section of the Great Plains Province (Fenneman, 1931, pp. 71-79), which is characterized by rolling uplands relieved by isolated mountains, buttes, and sections of badlands and by terraced valleys of streams that, with their tributaries, have dissected the general surface. The area which is now covered by the Heart Butte Reservoir and is the subject of this report lies about 50 miles by air from the mouth of the Heart River, considerably less than the channel distance between the two points. The outcropping bedrock here belongs to the Paleocene Fort Union formation, which consists of a basal zone of fine sand over- lain by zones of shale, clay, and sand. Lignite deposits of considerable extent occur in this formation and there are small mines nearby which produce this material for local use. A conspicuous feature of ex- posures in road cuts and elsewhere is the frequent presence of brick- red deposits of various thicknesses, apparently the result of the heat generated by natural burning of the lignite beds. Silicified wood (probably similar or identical to so-called “Knife River flint”) is re- ported to be associated with almost all of the Fort Union beds in the immediate vicinity, though not in great quantity (Tisdale, 1941, p. 14). At several localities there are limited zones in the basal sand which are cemented and much more resistant than the surrounding materials. These concretions, typically elliptical in shape and fre- quently of considerable size, are responsible for shallow rock shelters where they outcrop on the slopes (see pl. 12, a, 6). There are a few scattered patches of extensively weathered glacial till and occasional erratics in the area (Tisdale, 1941, p. 6). In the reservoir area, the Heart River, a perennial stream, meanders in a relatively wide, flat-floored valley lying approximately 200 feet be- low the rolling, treeless uplands into which it is incised (map 1; pl. 1, a). The stream, which is easily forded by car in places except during times of flood, has an average fall of approximately 0.6 foot per mile (Leonard, 1912, p. 29). All the tributaries in the area are relatively short and are intermittent. The shallow river channel is cut into allu- vium, but where it swings against the main valley walls there are sheer bluffs composed of bedrock, ordinarily the basal sand of the Fort 9 Y WeaRT @uTTE DAM SITE 326712 9 226T!0 326TII NORTH DAKOTA BISMARCK - a i s Py | ° 1360 2-H i954 Mar 1.—Map of Heart Butte Reservoir sh wing locations of archeological site , ' ; sb ral : 4 , , Ct — ¥ i vi ky / ats : : - ati FE ae Pee ney t an, Sai F es G | un oh a} P ve i i : vi; ‘| = Pl t ‘ nu? 4) a) r - ; Alaa : ry, tee. 2 reat v ; hae we ROD m. % Aa ia ¥ j fees it a ies CIS LY : % oN A ye ; : i, 2 : “4 4 a j a ; le ww ' : ? ba he ah Os i amg fe Lh a Mis 3 a F a ie irc ia : , ’ i, ’ r i) oe M4 ie f ie ily ¥ uy v7 VaR re nV A i & i i} a ie Al it, ay ri ' ‘ i i Mf m4 fi y 7 ~~ 7 ¥ s iv Py } | : 10 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 Union formation. The predominant vegetation, even today, is grass, and trees occur only in a sparse fringe along the course of the Heart River, in the valleys of the tributaries, and occasionally in favored spots on the main valley floor and the lower slopes. The common- est trees are ash, elm, and boxelder; there are some cottonwoods along the river and chokecherries are locally abundant. Lying within the Saskatchewan biotic province (Dice, 1943, pp. 24— 26), the region has a predominant vegetation of mixed short and mid- height grasses which undoubtedly once supported abundant herds of grazing animals. While very early descriptions of the faunal re- sources of the immediate vicinity are apparently lacking, the chron- iclers of various early 19th-century expeditions along the Missouri River (e. g., the Lewis and Clark and the Atkinson-O’Fallon Expe- ditions) recorded observation of vast herds of bison and numerous other mammals such as elk, deer, bears (including grizzly bears), and wolves. The climate of the region is characterized by long, cold winters, hot summers, rather scanty rainfall, and a mean growing season of ap- proximately 120 days. During the period of record at Carson, N. Dak., some 20 miles southeast of the reservoir area, the mean annual precipi- tation has been 15.68 inches, and the maximum and minimum temper- atures have been 116° and —48° F. respectively. Average temper- atures are 10.7° for January and 70.5° for July. At Dickinson, which is more than 100 miles northwest of Carson but has a very similar climate, the extremes in growing season during the 40 years of record have been 62 and 164 days. There are favorable factors in the climate of the region that tend to counterbalance the brevity of the growing season and the scantiness of precipitation. These are, on the one hand, the long hours of daylight during the summer months and, on the other, a relatively low rate of evaporation and the fact that a large proportion of the annual precipitation occurs during the growing season (approximately 70 percent during the months of April through September). The prevailing winds at all seasons of the year are from the northwest. On the average, the climate here does not compare unfavorably with that of stations along the Missouri River, where na- tive horticulture is known to have been practiced, but, at least in re- cent times, there have been occasional years when the frost-free season was exceedingly brief. The Heart Butte vicinity would seem to have little to offer in the way of a location for an aboriginal group seeking a place to settle and raise crops, but it would provide eminently suit- able camping locations for groups engaged in the hunt. Reported archeological investigations antedating those described here are lacking for the immediate area and are rare for the region west of the Missouri River in the Dakotas. This is not surprising, in view of the spectacular nature of the remains to the east and the relatively Lav. ke's))”)=)d«XHHEART BUTTE RESERVOIR AREA—COOPER sb minor attention that has been given the archeological situation in the northern Plains generally. Previous to the current salvage program, under which the areas for investigation are dictated by the plans of the dam construction agencies, most of the limited archeological research has been understandably directed toward the more intensively occupied banks of the Missouri River and of the streams in the eastern parts of North and South Dakota. Since the confluence of the Heart River with the Missouri is not more than approximately 50 miles from the Heart Butte vicinity, it is reasonable to suppose that at least some of the remains here might relate to sites in the immediate valley of the Missouri. Much of the reported work in the latter area has consisted of surface surveys (Will, 1924; Will and Hecker, 1944), but excava- tions have been accomplished in a few sites, notably in the contact- period Slant (or Old Fort Abraham Lincoln) and Double Ditch (or Burgois) sites near the mouth of the Heart River (Will and Spinden, 1906; Strong, 1940, pp. 360-365). These two sites have been attributed to the Mandan by the investigators. West of the Missouri River in areas immediately adjacent to North Dakota, two sites showing rela- tionships to complexes on the Missouri River have been investigated and reported. These are the Hagen site, on the Yellowstone River near Glendive, Mont. (Mulloy, 1942), and Ludlow Cave, near the head- waters of the Grand River and about 20 miles from the Little Missouri River in the extreme northwestern corner of South Dakota (Over, 1936; Strong, 1940, p. 884). The excavations in the former revealed a circular house site and yielded an artifact complex similar in many respects to Mandan and Hidatsa material culture. This similarity and the location of the village led Mulloy to suggest that the Hagen site might be attributable to the Crow, who were in the Yellowstone area in historic times and are believed to have been originally a part of the Hidatsa group (Mulloy, 1942, pp. 100-103). Ludlow Cave is reported to have yielded numerous projectile points, several sherds, and other artifacts in deposits underlying a superficial layer containing glass beads and projectile points and other objects of metal (Over, 1936). Strong has stated, on the basis of an examination of the speci- mens, that the pottery rather closely resembles ceramics of “a gen- eralized Mandan-Hidatsa type” and that “perhaps the closest re- semblance is to the Old Fort Abraham Lincoln Mandan” (Strong, 1940, . 384). ‘ Returning to the Heart Butte Reservoir area, the combined investi- gations of the River Basin Surveys in 1946 and 1948 and of the North Dakota University-Historical Society party in 1947, which included inspection of almost all of the area to be flooded, resulted in the location within or immediately adjacent to the reservoir of 16 sites that either certainly or probably represented aboriginal activity. These sites, most of which appeared to be lightly occupied camps, are 12 BUREAU OF AMERICAN ETHNOLOGY [Bury. 169 listed and briefly characterized in the Appendix. Most of them are now covered by the reservoir waters. The site designations used in this report are in accordance with the standard system of the River Basin Surveys in the Missouri Basin. The first number indicates the State, and represents the numerical po- sition of the State in an alphabetical list of the United States; the two letters following the numbers are symbols for the county; and the final number designates the specific site within the county. Thus, for example, 32GT1 is the first site numbered by the River Basin Surveys in Grant County, N. Dak., and 32SK4 is the fourth site num- bered in Stark County, N. Dak. THE KOEHLER SITE (32GT1) EXCAVATIONS The Koehler site, now covered by the reservoir, lay on the north bank of the Heart River near the line between the NE14 and NW% of the NEY, of sec. 9 T136N R89W. The river in this vicinity meandered from one side to the other of a relatively wide, fiat-floored valley, and the site was at the outside of a bend which closely approached the north- ern valley wall (see map 1). It occupied a small terrace remnant, roughly triangular in shape, although somewhat altered by the en- croachment of a short, shallow ravine. The remnant was bounded on the northwest by bluffs, on the south by a narrow strip of lower terrain bordering the channel of the Heart River, and on the northeast by a shallow dry ravine (pl.land map 2). A camp at that spot would have been well sheltered from the prevailing northwesterly winds by the bluffs, which rise 40 to 60 feet above the terrace surface. The border- ing ravine was heavily timbered, and there were a few trees on the terrace at the foot of the slopes from the uplands and in the small ravine that cuts into the terrace. Trees along the course of the river in the vicinity of the site were few and of small size. The predomi- nating varieties were ash, elm, and boxelder, but there were a number of chokecherry shrubs in the two ravines adjoining the site. During the period covered by the investigations, there was a heavy cover of grass on the site, as elsewhere throughout the reservoir area where cultivation was not practiced, but the owner of the land stated that during the drouth of the 1930’s the surface was bare of vegetation except for scanty patches of sunflowers and miscellaneous weeds. The surface of the terrace on which the site lay, approximately 18 feet above the river, was basically almost level; there were considerable areas, however, where the surface was irregular, quite obviously as the result of recent disturbance. This situation was explained by Mr. Koehler, the owner of the site, who reported that about 1940 his neph- ew had dug a number of pits in search of artifacts and that several 422141 O—8&8 (Face p. 12) LEGEND 1947 EXCAVATIONS BY NORTH DAKOTA UNIVERSITY — HISTORICAL SOCIETY PARTY 1948 EXCAVATIONS BY RIVER BASIN SURVEYS PARTY 50 = FEET GONTOUR INTERVAL 1947 EXCAVATIONS BY NORTH DAKOTA UNIVERSITY — HISTORICAL SOCIETY PARTY 1948 EXCAVATIONS BY RIVER BASIN SURVEYS PARTY so FEET 422141 O—58 (Face p. 12) CONTOUR INTERVAL : | Map 2.—Contour map of Koehler site (32GT1) showing locations of excavations. as - bs > re - af TES y P a ~ ee —. ? “peer i) : \ a AS SS | \ RIAN Sy ; : \y ; \" AN WS OS ia! rat 2Vvations, cars \\ \ 2 ER sheng los - 3 8" ' \ a Wt eR ‘ \ - x \ , \ Se sntnnsetineepaetinisscasiansiineaisree : 4 * a seme iol me gato reins Riv. Bas. Sur. : eit Per ae Pal No 9] HEART BUTTE RESERVOIR AREA—COOPER 13 deep furrows had been plowed across the site to turn up these ma- terials. In 1946, when the site was first observed by River Basin Sur- veys personnel, virtually no cultural material was observable on the surface except in places which had been thus disturbed. During its week in the Heart Butte area in 1947, the Hewes party established a grid system and excavated 8 test pits, each 5 feet square, at scattered points over the site (see map 2, for locations). These pits were carried to depths ranging from 8 to 86 inches. Hewes reported finding a cultural deposit of some thickness containing pottery and other artifacts, as well as charcoal-blackened soil, animal bones, and mussel shells (Hewes, 1949 a, pp. 21-22). No cache pits, post molds, or floor lines were observed, but it was thought concentrations of char- coal might mark the locations of fireplaces. Before the River Basin Surveys excavations were initiated in 1948, Dr. Hewes provided us with a ground plan showing his grid system and the locations of his test pits. Since several of his reference points (including the zero stake) were identifiable on the site, the established grid system was adopted for recording horizontal locations in most of the new excavations. Exceptions were 2 isolated trenches, excavation units 5 and 6 (X5 and X6), whose locations were plotted directly on the topographic map of the site made in the field (map 2). The zero point was at the southeast corner of the site. One base-line extended from this point along the northeast edge of the terrace in a direction 30 degrees west of magnetic north, and the other was a perpendicular through the same point. The former was arbitrarily designated the north-south zero line, the latter, the east-west zero line. In the follow- ing description of the excavations, references to directions are in re- lation to these baselines rather than to actual compass directions. Dis- tances “north,” “east,” or “west” of the zero point were recorded as the number of feet, together with the symbol “N,” “EK,” or “W.” Thus, a point designated ““N170W40” was 170 feet from the arbitrary east- west zero line and 40 feet “west” of the arbitrary north-south zero line (see map 2 and fig. 1). The units within the grid system were 5-foot squares, which were identified by the positions of their southeast cor- ners. Thus, for example, the southeast corner of square N170W40 was 170 feet north and 40 feet west of the zero point. The surface at the zero stake served as the vertical datum point and measurements in re- lation to this elevation were made by level readings with a telescopic alidade or by tape from points established by this method. Because the time and manpower available for the investigation of the site were strictly limited, only relatively small-scale excavations could be undertaken. They consisted of trenches, usually 2.5 feet wide, through various parts of the area of occupation. Ordinarily, the trenches were not expanded to uncover the entire extent of such features as fireplaces and concentrated refuse deposits when they ex- ue BUREAU OF AMERICAN ETHNOLOGY [BuLu. 169 LEGEND Limits of excavation Feature Figure 1.—Plan of excavation units 1 to 4 (X1—X4), Koehler site (32T1), showing locations of fireplaces and other features. Riv. Bas. Sur. is O88 Pap. No. 9] HEART BUTTE RESERVOIR AREA—COOPER 15 tended beyond the trench walls, since as extensive a sampling of the site as possible was deemed essential. For convenience in recording, trenches or segments of trenches were assigned excavation unit num- bers and are designated on the site map (map 2) as X1 through X6. Since all records were categorized under feature numbers—e. g., Fea- ture 2 was the description of the superficial appearance of the site, Feature 3, the description of the horizontal and vertical control system, ete.—only a minority of such feature numbers refer to cultural phenomena encountered in the excavations. All trenches were excavated well into the undisturbed soil which underlay the cultural deposits. Except as specifically noted, all were 9.5 feet wide. Excavation unit 1, at the north end of the main seg- ment of the level terrace and at the foot of the slope to the upland, con- sisted of a trench (N172.5-175, W10-70) 60 feet long lying perpen- dicular to the axis of the terrace finger and of an intersecting trench (W50-52.5, N150-172.5) 22.5 feet long. Excavation unit 3 extended nearly the full length of the terrace finger and was intersected by excavation unit 2 (N92.5-95, 15E-55W), a trench 70 feet long, and excavation unit 4 (N45-47.5, W17.5-77.5), a trench 60 feet long. Ex- cavation unit 3 was 160 feet long (0-N160) and was 2.5 feet wide (W15-17.5) except from N60 to N75, where it was 5 feet wide (W15-20). Excavation units 5 and 6 were on small detached seg- ments of the terrace (see map 2, for locations) and were 7 and 5.5 feet long, respectively. Excavation was in horizontal layers varying usually from 0.2 to 0.5 foot in thickness, depending upon the individual situation. The proveniences of specimens were recorded by square and excavation level except that exact horizontal and vertical measurements were taken of the locations when it seemed such data might have sig- nificance. As the excavations were carried downward, an attempt was made to define the horizontal and vertical limits of all features of cultural significance. A profile of one wall of each of the trenches was sketched on graph paper, based on vertical measurements in reference to the site datum. Such drawings were made of the W15 line from 0 to N92 and N95 to N160, the N45 line from W17.5 to W77.5, the N95 line from E15 to W55, the N175 line from W10 to W70, the W50 line from N150 to N172.5, the west wall of excavation unit 5, and the east wall of excavation unit 6. All pottery sherds, regardless of size, all artifactual material, and all identifiable faunal remains were col- lected, as were considerable samples of unworked stone and of un- identifiable bone refuse. A generally similar stratigraphy was found to prevail over all por- tions of the site tested, except in excavation unit 6, which was at a higher elevation, and in areas which had been recently disturbed. The following description applies to all excavations but excavation unit 6. 16 BUREAU OF AMERICAN ETHNOLOGY [Buuw. 169 Six zones were identifiable in all other trenches and were usually pres- ent at almost every point on the profiles. Recent disturbance, how- ever, had been so extensive in the area of excavation unit 2 that the classic profile was only fragmentarily represented on the N95 line, and various of the zones on the W15 line, especially between N80 and N123, had been destroyed. Other local absences of the upper zones on the W15 and N45 lines were also apparently attributable to the deep plowing which had been done on the site. The six zones were as fol- lows, top to bottom (see fig. 2 and pls. 3, a, 6, 4, a, for characteristic profiles). Zone A, a very thin dark layer, was the latest humus zone. It al- most never exceeded 0.15 foot in thickness, averaging closer to 0.1 foot, and was often not discernible on the profile. It was unusually well represented on the part of the N45 line shown in plate 3,b. This zone was culturally sterile. Zone B was a layer of light-tan material, predominantly a very fine sand over most of the site, but having a clayey component in excava- tion unit 1, near the foot of the adjoining bluffs. This zone varied in thickness from 0.1 to 0.8 foot, the thicker deposits lying generally near the slopes to the upland. Over much of the site, the average thickness was about 0.3 foot. No cultural material was found in this layer. Zone C was a dark layer, probably an old humus zone, containing occasional lenses, usually very small, of light-colored silty material. It appeared to have a considerable organic content. It ranged in thickness from 0.15 to 0.5 foot, being thicker in excavation unit 1 (where it has apparently been augmented by slope wash), and aver- aged approximately 0.2 foot thick in most of the site. Cultural mate- rial, mainly small bone fragments and flecks of charcoal, occurred sparsely within this zone, but there were no fireplaces nor any con- centrations of refuse. Zone D was a layer of predominantly light-colored silty material containing, from top to bottom, thin, discontinuous lenses of darker earth; small lenses of burned earth and ashes; and other evidences of occupation such as flecks of charcoal, fragments of stone and bone, and artifacts. A number of fireplaces and dense deposits of bones and mussel shells were also uncovered in this zone. The thickness of the layer varied from 0.25 to 0.8 foot and averaged about 0.5 foot in most of the site. Zone E resembled zones A and C in consisting of dark, heavily organic material; it was probably an old humus zone. It was not discernible in restricted areas of the site where it may have been de- stroyed by the activities of the aboriginal occupants. This was true in at least one instance, in excavation unit 3, where an extensive, shal- low pit, apparently excavated from the base of zone D, had removed Riv. BAS. Sur. PaP. No. 9] NI75W40 LOOPS PAL TLIPTOL LAGS LLAMA ACL ALL I LELE LS, Z COCs le FAT, eee es jake 7 7 GOAPOEOR AD CLF, ZZ 7 GG NMS SL APL gees I FELIPE LIS hh PMLA, 4 A PES So Lie HEART BUTTE RESERVOIR AREA—COOPER 17 NI75W30 SAS 4S 44s z Burned earth Feet © OOOO < oS RX 2A Keg Recently disturbed | Animal burrow Bone deposit 4 4 220) SSYY Ficure 2.—Characteristic profiles in Koehler site (32GT1). a, b, N175 line in excavation unit 1. c, W15 line in excavation unit 3. d, N45 line in excavation unit 4. 18 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 zone E (Feature 43, fig. 2,c). Small quantities of cultural materials occurred in this zone. Zone F was the sterile, light tan, silty or very fine sandy material underlying the cultural deposits on the terrace. At places on the N175 and W15 lines, the upper part of this deposit was stained gray and seemed to be slightly cemented, perhaps as the result of percola- tion of water carrying certain substances from the overlying deposits. The area in which excavation unit 6 was situated had been subjected to different depositional conditions. The profile there consisted of alternate layers of clay, varying from 0.25 to 0.9 foot in thickness, which were distinguishable by differences in color. There were six of these layers between the surface and the base of the trench, which was 3.5 feet deep. Bone fragments and an occasional fleck of charcoal occurred in the fourth and fifth zones below the surface, and a small fireplace lay at the contact between these two zones. The top of the upper cultural layer was 1.6 feet beneath the surface, the base of the lower was at a depth of 2.8 feet. Cultural debris was found throughout the excavations but was not uniformly distributed in the site. Evidences of occupation were very scanty in excavation units 5 and 6, and relatively so south of N50 in excavation unit 3. While moderate quantities of bone and stone frag- ments and two fireplaces occurred in the latter area, no pottery or other artifacts were found. Artifactual materials, not recovered in profusion anywhere in the site, were more abundant in the western part of excavation unit 4, in the northern part of excavation unit 3, and especially in excavation unit 1. Bone fragments were particu- larly numerous in the latter (pl. 5, a,b). They were found from top to bottom of zone D, and scantily in zones C and FE, but did not occur above or below these limits except where disturbance by man or beast was apparent. In only a few instances were there small clusters of sherds. Pottery fragments and other artifacts usually occurred as isolated finds within the general occupational zone. Similarly, bones, stone fragments, and mussel shells were ordinarily found scattered through the site, although an occasional cluster of stones, a concen- trated deposit of bones, and a similar deposit of shells were uncovered in the excavations. No cache pits were found, nor were there post molds or other evi- dences of structures. The only pit, other than those attributable to very recent activities on the site, was Feature 43, which lay between N66 and N91 in excavation unit 3 (fig. 1; pl. 3, ¢). Extending un- known distances east of the W15 line and west of the W17.5 and W20 lines, its irregular floor lay from 0.2 to 1 foot beneath the base of zone D. The fill was a relatively homogeneous brown silt in which there were no fireplaces and no lenses of ashes or burned earth but which contained occasional bone and stone fragments and, near the Riv. Bas. Sur. id AS Pap. No. 9] HEART BUTTE RESERVOIR AREA—COOPER 19 top, a small cluster of heat-fractured stones, Feature 27 (fig. 1), and a group of approximately 50 pottery fragments (Feature 44), mostly very small, from a single vessel (figs. 1; 2,¢). The excavation of this pit had obliterated zone E, so that it was directly overlain by zone D. Other conspicuous concentrations of debris consisted of two small clusters of potsherds in addition to the deposits of bones and shells previously mentioned. Two rim sherds and 43 body sherds, some exceedingly small, apparently all part of the same vessel represented by Feature 44, lay at the base of zone D, just above the pit, Feature 43, at approximately N71 on the W17.5 line (see pl. 3, d, and Feature 28, fig. 1). At N107-109, W17-18.5, and 0.2 foot above the base of zone D were the remains of approximately one-third of a vessel, so badly fragmented that restoration is impossible (pl. 3, e). This group, Feature 30, included 11 rim sherds and 325 body sherds, many of the latter not exceeding a centimeter in greatest dimension. Feature 19 was a dense deposit of bones, mostly of bison, encountered in the western end of excavation unit 1 (fig. 1) and extending north and south of the limits of the trench. Varying in thickness from 0.1 to 0.2 foot, it lay approximately horizontal 0.1 to 0.2 foot above the base of zone D (fig. 2, 6) and extended from W57 to W68. Plate 5, } shows the unidentifiable bones from this deposit which were discarded in the field. A deposit of mussel shells (Feature 29), similar to Feature 19 in thickness, and probably in horizontal extent, was partially ex- posed in excavation units 2 and 3 (pl. 5,c). It also lay horizontally, but directly upon the base of the cultural zone, except that much of it immediately overlay the large pit, Feature 43 (pl. 3,¢). Like Feature 19, this was a dense deposit ranging from 0.1 to 0.2 foot in thickness. The other features in the site were fireplaces, of which 13 were un- covered completely or partially in the trenches. These tended to cluster in two areas, near the slopes to the upland, in excavation unit 1, and from N60 to N85 in excavation unit 3. Judging from those which were completely exposed and what observations were possible on those only partially excavated, they ranged in diameter from slightly less than 2 feet to approximately 3 feet. A size nearer the lower end of this range appears to be most frequent. Stones were never associated with the fireplaces, which at least in some instances were simply limited circular areas on the occupational surface where fires had been maintained and which were apparent as lenses of earth usually surmounted by ashes and burned toared color. In eight cases, the sur- face of the burned earth was deeper in the center than at the edges so that the ashes lay within a basin-shaped depression. These depressions varied from less than 0.1 foot to 0.25 foot in depth and some at least may have been created incidentally as the ashes were removed by the users. Ashes usually, but not invariably, lay upon the surface of the fireplaces; sometimes there were traces only but in other instances the 20 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 deposits were as much as 0.25 foot thick. The depths to which the color of the underlying earth had been affected by heat varied considerably as well, suggesting differences in the duration of the fires which had burned on the hearths. The range in maximum thickness of the burned earth lenses was from 0.05 foot in the case of Feature 33, a small fire- place in excavation unit 6, to 0.5 foot for Feature 32 at the base of zone D at N60.8W15 (fig. 2, ¢). The fire-reddened earth beneath the majority of the fireplaces, of which Feature 7 (pl. 4, c) and Feature 13 (fig. 2, a) were typical, was 0.2 to 0.25 of a foot thick. In the main part of the site, where the classic profile prevailed, fireplaces were found from top to bottom of zone D, although the majority (8 of a total of 12) were in the lower half of this deposit and 4 lay on or nearly on its base. One (Feature 35) was at the upper limits of zone D in the southern part of the site. ARTIFACTS As indicated above, artifactual materials were recovered in rather limited quantities. Furthermore, the range of forms is not great, and numerous types almost invariably present in adequate samples from sites on the Missouri River in the northern Plains are absent from the collections. Examples of such artifacts are the ubiquitous scapula hoes and shaft straighteners. No metal or other White trade items were found, nor have the collectors reported finding such materials. In the following sections, the specimens collected by the North Dakota University-Historical Society party have not been included in the artifact counts, but are mentioned when they provide informa- tion lacking in the River Basin Surveys collections. Hewes has de- scribed the few rim sherds found in his excavations (1949 b). POTTERY The pottery collected from the site by the River Basin Surveys party consists of 1,883 fragments, which range in size from less than 1 cm. to 8.5 cm. in maximum dimension. More than 75 percent of these are less than 3 cm. long, and a very small number lie near the upper end of the size range. Sixty-two sherds are from the rim of the vessel and include the lip—some of them consist of little more than this feature—and 23 additional sherds are from various deco- rated parts of the rim area, but do not extend to the lip. Several of the sherds adjoining the vessel mouth have been fitted together to reduce the number of such fragments from 62 to 39, believed to repre- sent 28 vessels. Three groups of sherds, found in as many clusters in the site, are apparently attributable to two vessels (Feature 30 and Features 28 and 44, respectively), although they have proved capable of little restoration. These groups constitute the only information Riv. Bas. Sur. eb Pap. No. 9] HEART BUTTE RESERVOIR AREA—COOPER 21 available as to certain attributes of the pottery, such as size and overall form, and the combination of attributes which go to make up a pottery container. Feature 30 consists of 11 rim sherds and 325 body sherds, many of which are exceedingly small. The rim sherds and several body sherds have been combined into 3 larger rim and neck fragments, and 16 body sherds permitted restoration of a part of the lower body of the vessel. Features 28 and 44 comprise 10 rim sherds, which partial restoration has reduced to three larger pieces, and 101 body sherds. The tempering material included in the pottery, judging from ex- amination of the edges of the fragments in the collection and of several crushed sherds, is invariably composed of angular rock fragments, apparently crushed granite. The particles are relatively fine, averag- ing considerably less than 1 mm. in diameter; the finest are less than 0.25 mm. in diameter, and only a rare example exceeds 1 mm. Only moderately abundant, the tempering fragments are rather uniformly distributed through the paste in most sherds, although occasionally some clumping is observable. The paste, as revealed by the broken edges of the sherds, is most often fine-grained and compact, but occa- sionally has a rather gritty appearance as though fine silty material were included with the clay. The gritty pottery tends to break rather cleanly, while the broken edges of the sherds with finer paste are most often irregular and sometimes fairly contorted. Some of the latter have a slightly laminated appearance and there is an occasional split sherd. Surfaces vary from very smooth to gritty, but the preponderant surface, both interior and exterior, can be described as moderately smooth. Occasional sherds, usually but not exclusively from the neck area, are glassy to the touch, and havea high light reflectance. This ef- fect has been achieved by a usually horizontal smoothing, apparently with a very smooth, hard-surfaced object. There appears to be rela- tively little variation in hardness; of the samples tested, over 95 per- cent fall between celestite (3.5) and fluorite (4), the remainder are between 3 (calcite) and 3.5. Surface color is usually gray, although there is a gradation to a dull buff, at least sometimes on the same vessel. The two surfaces of a sherd are generally similar in color except for alterations, such as smoke blackening and the accumulation of carbonized material, inci- dental to use. The core is almost always entirely or predominantly gray, usually of a very dark tone approaching black but occasionally fairly light. The cross sections of sherds with buff surfaces, however, sometimes reveal thin strata of a similar color adjacent to the surfaces. Of the body sherds, 68 percent have alternating grooves and ridges on their exterior surfaces, presumably created by the application of a grooved or thong-wrapped paddle (pl. 12, c, 2, 6, 9), a treatment which has frequently been called simple stamping by archeologists in the Plains area. Seven percent bear incised decoration, and the BP BUREAU OF AMERICAN ETHNOLOGY [BuLL. 169 remainder are plain. A large proportion of the plain sherds are probably from the neck area or from the rim, although a very smali number are quite obviously from the lower portion of the vessel. The lack of grooves and ridges on a few sherds does not necessarily indi- cate that the paddle was not always used, since these features were barely discernible on many sherds from vessels which had been un- usually well smoothed subsequent to the paddling. A few sherds from the lower edge of the decorated area indicate that sometimes, at least, paddle marks extend this high on the vessel, but they are never dis- cernible within the decorated zone. They are not present on the neck and rim of Feature 30, but traces of vertical grooves and ridges can be seen on the outer rim of Features 28 and 44, despite considerable smoothing. Such vestiges are not visible on any other identifiable neck or rim fragments. Although on a large proportion of the sherds there is evidence of various degrees of smoothing subsequent to pad- dling, such evidence is lacking on some. The grooves, which appear to lie vertically on the vessel, with some crossing in the basal area, vary considerably in width, but most are from 3 to 4 mm. wide and are separated by ridges 1.5 to2 mm. wide. The extremes in width are ap- proximately 1.5 mm. and 7 mm., but only rare examples lie near these extremes. The remnants of vertical striations, usually nearly oblit- erated by subsequent horizontal smoothing, are observable on some sherds from the neck area. Interior surfaces are smooth and relatively even, lacking tool marks, except that in the area from the lip to the point of maximum constriction there are sometimes polished elongated horizontal facets resulting from the final smoothing of this area. The specimens in the collection cast little light on the method of manufacturing the pottery. There is, however, no evidence to sug- gest coiling, and it is probable that the paddle marks on the exterior surfaces were incidental to at least some of the later phases of the fabricating process. The collection provides little evidence relative to the general form or the size of the vessels represented, but some clues are available. There are, for example, no sherds indicating bow] forms, angular necks or shoulders, except for a single very small sherd which may be from a miniature vessel (pl. 7, 9), or bases which are pointed or flattened. A partially restorable vessel, Feature 30 (pl. 7, 3), had originally, judging from the portions present, a mouth diameter of approximately 205 mm. and a neck diameter of approximately 180 mm.; its maximum diameter and height are unknown. A partially restorable segment which appears to belong to the lower part of the body suggests a rounded base. The rim rises in a wide curve from an apparently rounded shoulder area and flares outward from the Eke gd) HEART BUTTE RESERVOIR AREA—COOPER 23 point of maximum constriction. The exterior surface of the rim has been thickened, by the addition of a fillet, for a distance of 9 mm. below the lip. The thickness varies from place to place on the vessel. The upper rim averages 8 mm., sherds from the neck average 5 mm., and those from the lower part of the body range from 3 to6mm. The thinnest sherds, 2.5 mm. thick, are from the decorated shoulder area. An even more scantily represented vessel (Features 28 and 44) ap- pears to have been similar in form and not greatly different in size. A mouth diameter of approximately 240 mm. and a neck diameter in the neighborhood of 215 mm. are indicated. The rim is thickened as on Feature 30 (pl. 7, 7). Other, smaller, rim sherds suggest that most of the vessels represented may be close to these in size, although one unthickened flaring rim indicates a mouth diameter probably not exceeding 150 mm. (pl. 7, 6). The commonest rim form (81 sherds representing 17 vessels) is characterized basically by a slight to pronounced convexity of the outer surface of the upper rim as viewed in vertical section, and a corresponding concavity of the inner surface (pl. 6). The exterior and interior walls are usually parallel and describe a smooth curve, but in one instance there is a sharp break on the outer rim 43 mm. below the lip and a somewhat less pronounced angle on the inner surface (pl. 6, 13). This latter specimen is also unusual in the series in its thickness and in the height of the decorated rim area. On 8 sherds (4 vessels) the concave-convex part of the rim is surmounted by a flare to a lip which is rounded, thinned, or thickened (pl. 6, 8, 9, 14), but otherwise the surfaces curve directly to a lip which may be flattened, rounded, thickened, or thinned (pl. 6, 7-7). Thinning of the lips of the last category was usually accomplished by beveling the inner rim margin (pl. 6, 1, 4-7). None of the sherds in our collection includes the total height of the concave-convex portion of the rim, but a single specimen collected by the Hewes party extends to the point at which the curve to the shoulder area begins. This sherd, which has been thickened on the interior adjacent to the lip, measures 32 mm. from lip to point of maximum constriction (pl. 6, 77). Thirty-one sherds (9 vessels) flare to a rounded, thickened, or thinned lip except that 25 (5 vessels) have been thickened on the exterior surface adjacent to the lip. On all but one of the latter the thickening is elliptical in cross section and varies from 5 to 10 mm. in width (pl. 7, 7-4, 7). Note variation on single vessel fragment in plate 7, 7; the exception, apparently created by folding the plastic clay outward, breaks sharply at the base of the thickening, 15 mm. below the lip, to give the upper rim a roughly triangular cross section (pl. 7,10). 42214158 8 24 BUREAU OF AMERICAN ETHNOLOGY [Buny. 169 Decoration occurs on the lip, inner rim, outer rim, and shoulder area and is, with rare exceptions, in two techniques, incising and impressing with single twisted cords. Except for one sherd in the collection made by the Hewes party and described below, incising is restricted to the body—probably to the shoulder area only—and cord impressing is apparently found only on therim. Incising is invariably fine, 1 to 2 mm. wide and usually so shallow as to defy measure- ment except at the beginning of a stroke, where it may approach 1 mm. in depth. A single body sherd in our collection combines punctations with incisions (pl. 7, 9), and there are two punctations on a cord-impressed rim sherd in the Hewes collection (pl. 6, 15). The sherds in our collection are too small to permit any comprehensive determination of the designs. It is evident, however, that they con- sist of groups of parallel lines and that these groups are combined in various ways. A few sherds reveal a distinctive interrupted-line ar- rangement which appears to be similar to an incised design at the Double Ditch site (Will and Spinden, 1906, pl. 38, A). There is no way of determining what proportion of the vessels from the site bore body decoration or how such decoration is correlated with other characteristics of the pottery. It should be stated, however, for what it may be worth, that approximately 30 percent of the body sherds from Feature 30 bear incised lines in contrast to a figure of approxi- mately 7 percent for the remainder of the body sherds from the site. It seems certain, judging from the sherds comprising Features 28 and 44, that body decoration was not always present. If it could be assumed—which it cannot—that the various zones of the complete original vessel were proportionately represented by the sherds of Feature 30 and that other conditions were such as to produce a fragmentation representative of that for pottery in the site in gen- eral, a tentative conclusion might be drawn that approximately 1 of 4 vessels bore body incising. While the assumptions necessary to reach this conclusion are highly questionable, it may not be unrea- sonable to guess that a minority of the vessels which contributed to the sherd collection had been incised. The cords used to impress designs on the rims invariably had a left-hand twist, but they varied in diameter and in the tightness of the twist. Plate 6, 17, is illustrative of the finest, most tightly twisted cord used on the pottery from the site; plate 6, 4, of the largest cord. The range in width of the cord impressions is approximately 1.5 mm. to3mm. Rims with exterior thickening invariably have the resulting band decorated with closely spaced oblique lines, extending from the upper left to lower right or vice versa (pl. 7, 7-4, 7,10). In addition, the lip of rim sherds of Feature 30 bears a series of notches, perhaps made by impressing with the fingertip, averaging approximately 10 mm. long and 4 mm. apart (pl. 7, 3), and the lower border of the RIv. Ss. . AGES al HEART BUTTE RESERVOIR AREA—COOPER 25 collar of another specimen is notched at intervals by the removal of clay with the fingernail (pl. 7, 70). A rim sherd in the Hewes collection has a series of deeply incised lines lying obliquely from upper right to lower left across the thickened portion (pl. 7,4). The exterior surfaces of concave-convex upper rims are decorated, except for 2 entirely plain specimens, with cord impressions often (10 sherds, 5 vessels) as a series of horizontal lines extending to the lip (pl. 6, 1-4). In one instance the horizontal series is interrupted by a series of diagonal, or perhaps curvilinear, lines (pl. 6,7). The three sherds representing this vessel also have a band of decoration 9 mm. wide adjacent to the lip on the inside of the rim. This band is composed of closely spaced cord-impressed lines lying obliquely from upper right to lower left. Three other fragments, all small and not extend- ing to the vessel mouth, reveal a similar interruption of horizontal lines by diagonal, or curvilinear, lines, and there is a similar specimen in the Hewes collection (pl. 6, 77). When the upper part of the rim flares outward above the concave-convex section, the cord-impressed lines do not extend onto the flaring portion (pl. 6, 8, 9, 14). Other designs on rims of this form are oblique lines, probably between a pair of horizontal lines (pl. 6, 5, 7, 16, 18) and groups of opposed diagonal lines (pl. 6, 6). On the single rim sherd which has a sharp break in the rim curve (pl. 6, 72), three diagonal lines break a series of horizontal lines. This sherd, although similar in paste, color, and tempering, differs from other rim sherds in the site and resembles pottery reported by Hewes from a site (32SI4) on the Missouri River near Fort Yates, N. Dak. (Hewes, 1949 b, pl. 6,7). Of the 5 sherds (3 vessels) from unmodified flaring rims, 1 certainly and 2 others (1 vessel) probably are decorated only on the lip, which in each instance is thickened to the exterior, is flattened, and slopes outwardly. Both bear closely spaced cord-impressed lines lying obliquely from upper left to lower right (pl. 7, 5). The remaining 2 sherds (1 vessel) are decorated on the lip and both surfaces of the rim with closely spaced, exceptionally fine cord-impressed lines. A band of horizontal lines occupies the space on the outer rim surface below a point 8 mm. from the lip, and above this is a series of oblique lines slanting from upper right to lower left (pl. 7,6). A series of lines arranged as horizontal chevrons lies across the lip and on the upper margin of the inner rim surface. The group of sherds which appear, on the basis of the decorative technique present, to be from the rim area, add nothing to the fore- going descriptions except that on a single sherd one and part of a second elliptical punctate lie adjacent to, and probably below, the vestiges of a cord-impressed line. The ceramics from the Koehler site, quantitatively limited as they are, appear to fall entirely within the range described for sites which 422141—58——4 26 BUREAU OF AMERICAN ETHNOLOGY [BuLu. 169 have been attributed to the Mandan of the 18th century (Will and Spinden, 1906; Will and Hecker, 1944; Strong, 1940). Not only do the descriptions of the general characteristics of Mandan ceramics of this period apply to our pottery, but many of our sherds appear to be almost identical to specimens illustrated in the cited reports (Strong, 1940, pl. 5; Will and Hecker, 1944, pl. 15; Will and Spinden, 1906, pls. 87-41, figs. 14-16) and most of them can be virtually dupli- cated in the limited surface collections from the Slant and Double Ditch sites in the possession of the River Basin Surveys, especially from the latter. The lack in the scanty Koehler site collection of the more elaborate forms and of such features as check-stamped surfaces, handles, and cord-wrapped stick decoration has little significance, since these are reported to be rare in the large collections from the Missouri River sites. The major difference between our sample from the Double Ditch site and the Koehler site materials is in the thickness of the sherds; the average for the latter is appreciably smaller than that for the specimens collected from the surface of the former. There seems to be no reason for comparing our pottery with that reported by Mulloy (1942) from the Hagen site, since the resemblances are of a much lesser order. WORK IN STONE Fragments of stone of various kinds—quartzite, cherts, chalced- onies, and various igneous forms—were abundant in the site, but artifacts were not numerous and the range of forms represented is relatively limited. Ground stone artifacts are rare and there are only 2 or 3 unshaped cobbles which appear to have been used as ham- merstones. There are a few small fragments of pumice, none of which shows any evidence of use. Chipped artifacts were made from what appear to be four varieties of chalcedony (including a considerable quantity of so-called Knife River flint), a gray chert, a gray silty chert, a fine-grained quartzite, and—rarely—jasper. Chert and chal- cedony were most often utilized for the smaller artifacts, while the larger ones were predominantly made from quartzite and the silty chert. There are 10 complete or fragmentary projectile points in the col- lection, 9 of which are small, light, and triangular and have straight or very slightly concave bases. All are illustrated in plate 8, 7-0. Of the 9, all but 1 have a single pair of side notches. They are deli- cately chipped on both surfaces te produce a thin biconvex cross section, invariably falling between 3 and 4 mm. in thickness. The sides are sometimes slightly convex. The 4 complete or nearly complete specimens have actual or estimated lengths and widths as follows: 28.5 mm. and 14 mm. (pl. 8, 7), 24 mm. and 14 mm. (pl. 8, 2), 25 mm. and 14.5 mm. (pl. 8, 2), and 28 mm. and 13 mm. (pl. 8, 4). Riv. BAS. SuR. 7 eee! Bae eo HEART BUTTE RESERVOIR AREA—COOPER 27 Plate 8, 70, shows the single small unnotched point, which has a maxi- mum width of 14mm. The points represented by the smaller frag- ments probably did not greatly exceed the size range of the complete specimens; the largest (pl. 8,5) is 18.5 mm. wide and is, in its incom- plete condition, 29 mm. long. The specimen shown in plate 8, 2, has been reworked to its present form by chipping from one face; while the specimen has probably been broken, all edges including the con- cave one, have been chipped. The points are made from a homo- geneous chert (pl. 8, 70), silty chert (pl. 8, 2, 4), “Knife River flint” chalcedony (pl. 8, 3, 6-8), and a light yellowish chalcedony (pl. 8, 1,5). A-single basal fragment, 20 mm. wide and 23 mm. long, with straight base and parallel sides, apparently is part of a larger projec- tile point (pl. 8,9). Of chalcedony, it has a thin (3.5 mm.) biconvex cross section. Three fragments in the Hewes collection are apparently from both notched and unnotched triangular points. Twelve complete and fragmentary end scrapers were found in the excavations. All complete or nearly complete specimens and a speci- men from the Hewes collection are illustrated in plate 8, 72-23. All are basically triangular in form and are relatively small; the complete specimens vary in width from 20 mm. to 26.5 mm, and in length from 23 mm. to 33mm. The ventral surfaces are unworked and represent the original unmodified flake surface, except in two instances where a small amount of chipping was probably necessary to remove some irregularity. The dorsal surface has usually been shaped by overall chipping, although this is not true of the specimen shown in plate 8, 23. The working edge is steeply inclined except in one instance (pl. 8, 20). One specimen has a graverlike point at one end of the steep scraping edge (pl. 8, 74). All but two of the end scrapers are made of “Knife River flint” chalcedony. Of the two exceptions, one is of a compact gray chert, the other, of dark red jasper. The only end scraper in the Hewes collection is like those described above, but is larger; it is 45 mm. long and 30 mm. in greatest width (pl. 8,12). A single specimen, somewhat similar to the end scrapers in general form, although larger, has a concave, gougelike working edge (pl. 8, 24). Of “Knife River flint,” it is 40 mm. long and 33.5 mm. wide. One or two coarse flakes have been removed from the ventral surface, but otherwise it resembles the end scrapers in being worked only on the dorsal surface. The site produced a rather wide variety of blades, varying con- siderably in size and form. One group of bifacially flaked specimens, while varying in form and size, can be set apart on the basis of rela- tive length and width and the invariable presence of retouching on the edges. Ends may be pointed, rounded, or straight, and sides are straight, or moderately convex, but all are alike in being relativély long and narrow. The complete range is illustrated in plate 9, 7-1/7. 28 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 Four of the 18 complete and fragmentary specimens are of “Knife River flint,” 4 of other chalcedony, 4 of silty chert, and 1 is of fine- grained quartzite. The smallest complete specimen (pl. 9, 17) is 46 mm. long and 15 mm. in maximum width, the largest (pl. 9, 2) is 109 mm. long and 28 mm. wide. The specimens are 3 to 4 mm. thick. Some of them resemble rather closely specimens from the Double Ditch site, described and illustrated by Will and Spinden (1906, pl. 34). A fine-grained quartzite fragment is long, narrow, and relatively thick (pl. 8, 29). It may have been intended to serve as a drill, although its point seems rather blunt to have so functioned. It is carefully worked on both faces and is 7 mm. thick and 15 mm. wide. All but 2 of 11 rather large blades are incomplete; they are made of fine-grained quartzite or silty chert with one exception, which is of “Knife River flint.” They have convex sides and pointed (pl. 10, 1, 3,6), straight (pl. 10, 5, 8), or rounded (pl. 10, 2, 4,7,9) ends. They are shaped by percussion flaking on both surfaces, and secondary chip- ping is sometimes present, sometimes absent on the edges. On some specimens, only parts of the edges have been retouched. Two complete specimens are 59 mm. long and 40 mm. wide and 127 mm. long and 50 mm. wide (pl. 10, 7), respectively. Specimens of this class vary in thickness from 8 mm. to 13 mm. Like the long, narrow blades pre- viously described, these objects seem to bear a close resemblance to specimens reported and illustrated from the Double Ditch site (Will and Spinden, 1906, p. 164 and pl. 33). A few larger and heavier blades made from the same materials may have been used either as scrapers or as choppers. The only complete specimen (pl. 10, 73), 112 mm. long, 68 mm. wide, and 23 mm. thick, is the most carefully made of this group of artifacts. On the others, only enough percus- sion flaking has been done to create an ordinarily rather blunt edge and sometimes to reduce a fragment to convenient size (pl. 10, 11, 12, 14-16). . Other chipped-stone objects consist of irregular flakes, large and small, with one or more retouched edges. The end of one very small, thin flake of chalcedony has been chipped on one surface to create a very fine point (pl. 8, 77). It is difficult to visualize a function for this specimen, since it is much too thin and fragile for use as a graver. The smaller retouched flakes are usually chalcedony, while the larger, heavier retouched fragments are most often quartzite and silty chert (pl. 9, 12-14). Objects of stone other than those which have been altered by chip- ping are very rare. A thin subtriangular object smoothly ground into shape from Amazon stone, a green variety of microline, has been perforated, presumably for suspension (pl. 8, 28). Above the bi- conical perforation, which is approximately 1 mm. in diameter, and at the apex of the triangle, can be seen the remains of a similar hole Riv. Bas. Sur. rA— Pap. NO. 9] HEART BUTTE RESERVOIR AREA—COOPER 29 which had apparently been drilled too close to the edge of the stone. This object, supposedly a pendant, is 16 mm. long, 12 mm. wide, and 3 mm. thick. Three thin fragments of a brick-red stone with lamin- ated structure, tentatively identified as volcanic ash, are probably split from a single tabular piece which was rectangular in outline. Most of the original surfaces have scaled away, but a small remnant near a rounded corner of one of the pieces bears a short segment of an incised line (pl. 8, 26). Another thin tabular fragment of similar material with a rounded edge bears the remnants of 3 or 4 lines which seem to radiate to the edge of the object from a common point (pl. 8, 25). This piece also appears to have been split from a thicker stone. Although they are so fragmentary that no definite statement is pos- sible, these objects seem to me to be reminiscent of incised rectangular and tabular stones reported by Will and Spinden from the Double Ditch site (1906, pp. 165-166 and fig. 4). Of similar material is an asymmetrical but basically disk-shaped bead with a conical central perforation (pl. 8,30). It has an average diameter of approximately 17 mm., an average thickness of 7 mm., and a perforation 4-7 mm. in diameter. A fragment of fine-grained red stone is probably part of a circular bead. It is crudely shaped and is broken through a conical perforation (pl. 8, 27). The diameter of the bead was 13 mm., the thickness 8.5 mm. Finally, as far as tools or other objects of stone are concerned, there are two unshaped cobbles of igneous rock each with one end somewhat battered, suggesting their use as hammer- stones. As in the case of the pottery, it appears that the artifacts of stone from the Koehler site fall within the range of the materials from the Double Ditch site. Although the descriptions of projectile points, for example, from the latter site are too general to permit a satis- factory comparison, there seems to be no doubt that there are at least some nearly identical artifacts. The similarities in some of the larger chipped blades have already been noted. I suspect, also, that there is a close correspondence in the materials utilized. For example, Will and Spinden mention the use of a “gray chert or fine grained argillite” for the larger blades (1906, p. 164), and I am of the opinion they are referring to material identical to that which is described here as a silty chert; stone of this kind is in our surface collection from the Double Ditch site. I do not remember having seen material of this sort, at least in quantity, in sites of other cultural affiliations, and its use may prove to be culturally diagnostic. WORK IN BONE AND ANTLER Objects of worked bone and antler recovered in the excavations are few in number and represent a very limited range of artifacts. A single well-made awl, 108 mm. long, is made from a section of long 30 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 bone, probably the metapodial of a deer or antelope (pl. 11,8). The butt has been modified by grinding, but traces of those features char- acterizing the articular surfaces of the end of the bone can still be recognized. The implement is relatively uniform in width to within 24 mm. of the sharp point, which is formed by a sharp taper of the two edges and the interior surface. Four spatulate objects, all but one fragmentary, may be classed together. Three are made from the lateral surface of a large rib, one is probably from the thickened border of a large scapula. The three fragments retain on the inner surface a layer of rough cancellous bone which shows no evidence of smoothing. The edges are roughly shaped, and only the ends appear to have been used. On 1 specimen, the end is rounded, and on 2 it is bluntly pointed (pl. 11, 7, 3,6). These ends bear 1 or 2 wear facets. The fourth specimen has been worked overall and has one end thinned and rounded. The other end, also rounded but considerably more tapered, bears a wear facet but has also been somewhat nicked as though the tool might have been used in chipping. This specimen is 120 mm. long and 19 mm. wide (pl. 11, 2). These specimens appear to be similar to bone objects from the Double Ditch site illustrated by Will and Spinden (1906, pl. 35, , 0), and as far as general form and the nature of the working ends are concerned, to objects from the vicinity of Mobridge, S. Dak., described by Wedel (1955, pp. 123-126). Wedel points out the resemblance of his specimens to an object de- scribed and figured by Orchard (1916, p. 9, pl. 5), Hels 12 as a quill- flattener and attributed to the Sioux. There are two objects which may have been used primarily for chip- ping stone, judging by the nicked condition of their blunt points. One is an otherwise unmodified tip which has apparently been broken rather than cut from an antler (pl. 11, 77), the other appears to be made from the thickened border of a large scapula. The distal half of this specimen has been worked overall to remove all traces of can- cellous bone and to create a symmetrical taper to a rounded point (pl. 11, 12). The antler specimen is 95 mm. long and about 18 mm. in diameter at the butt, the bone specimen is 223 mm. long and varies from 10 mm. to 35 mm. wide, near the tip and at the base, respectively. A small highly polished section of bird bone is presumably a bead. Marks of cutting are visible at both ends, and there are several short transverse scored lines irregularly spaced at various points on the specimen, which is 14 mm. long and 6 mm. in diameter (pl. 11,4). A fragment of a small mammal rib is unmodified except for a slight polish and half of a biconical perforation at one of the broken ends (pl. 11,4). The specimen is 43 mm. long and 2.5mm. wide. Its func- tion is uncertain, but it is suggestive of specimens from the Double Ditch site (Will and Spinden, 1906, p. 172 and pl. 36,z). A split sec- tion of bird bone, rounded at one end and broken at the other, is also Riv. Bas. Sur. Tee Pee HEART BUTTE RESERVOIR AREA—COOPER 31 of uncertain function (pl. 11,9). It is 5 mm. wide and, in its present condition, 68 mm. long. All edges but the broken one are well pol- ished. The remaining specimen in the River Basin Surveys collec- tion is the distal end of a bison metapodial, which has been cut from the shaft and is presumably a byproduct of preparation of the shaft for manufacture of an artifact. The collection made by the Hewes party includes three objects of bone showing more or less modification. There are 2 fleshers, 1 broken near the working end, made of bison metatarsals by cutting diagonally through the shaft from the anterior to the posterior surface and pro- ducing a rounding chisellike edge where the cut meets the posterior surface. In each instance, a few narrow notches have been cut into the edge and the wear produced by use is on the cut side of the edge. The butt consists of the unmodified proximal end of the bone. The illustrated specimen (pl. 11, /0) is 158 mm. long, and the other is al- most identical in size. A fragment of rib of a large mammal, roughly broken at both ends, has a series of 10 short transverse scored lines near one end, but is otherwise unmodified (pl. 11, 7). There is no evidence of wear or polish anywhere on the specimen. TaBLe 1.—Artifacts from the Koehler site North Da- River Basin} kota Uni- Artifact Surveys versity- Illustrations collection | Historical Society collection Pottery: Ring sherds) (uchiding Jip) =. 28 2 62 10 | Pls. 6, 1-15; 7, 1-8, 10. Rimisherds (notincluding lip)ies. 2 eee 23 2| Pl. 6, 16-19. IB Od VISHCTOS Ste saa es tet lg 1, 288 215 | Pls. 7, 9; 12, c, 2-6, 9. Stone artifacts: Preipctilemoints 226 2. fon OPA oe ee 10 3 | Pl. 8, 1-10 IE TICUSCTADONS oe es ee eo ee eee ea nn See iby 1 | Pl. 8, 12-23 “Gouge 2. = £2 ce 1 OME See. ID Fol Go pgp See Re 1 0} Pl. 8, 29 Long, narrow blades-_- 13 0 | Pl. 9, 1-11 APE OI DIAC CG eee eee ee ye te ee een, BN il 2 | Pl. 10, 1-10 mCUOD DENS ne ao see lee eR Sa 10 1 | Pl. 10, 11-16 Graverlikeropiects: 3c. eee ie Ae eet 1 On Pel Sart Cincilaribead ase aes ie es ee ee 1 O08 pel Sirey: DiskayGa deme eee sae 5 ee ee a SS SEES 1 0 | P18, Ingised tabletss20 <5. . OY oe I ee Se 22 0 | Pl. 8, 26, 26 AINAZON Stone Pendantee. =. 2. .e a ee ee eS 1 0} Pl. 8, eammierstones®=-e- ose a oe ee 2 1 Bone artifacts: 1 OQ aebabl 2 OF Pei 0 2 SPV 0 4 0 | Pl. 11, 1-3, 6 0 G2 ahle 1 0| Pl. 11, 4 1 OF eats 1 ON Pela 1 Quantities include complete and fragmentary specimens. 3 4 fragments. FAUNAL REMAINS Bones were the most abundant remains in the site. A large propor- tion were exceedingly fragmentary and, not being identifiable, were 32 BUREAU OF AMERICAN ETHNOLOGY [BuLy. 169 not all collected. The bones in general were in such small fragments (pl. 5, 6) that it seems certain they were deliberately crushed, prob- ably to be boiled for the manufacture of bone grease (Leechman, 1951). An occasional bone was scorched. All bones which seemed to present any opportunity for identification (virtually all of these except vertebrae, tarsals, carpals, and phalanges were also fragmen- tary) were collected, but they probably constitute less than half the bulk of the total uncovered in the excavations. Of the 959 bones iden- tified by Dr. Theodore E. White, 731 are of bison (Bison bison), and 217 are of the domestic dog (Canis familiaris). Other forms repre- sented are beaver (Castor canadensis), by 9 bones; prairie dog (Cyn- omys ludovicianus), 1 bone; and a microtine rodent, 1 bone. It will be noted that a number of mammals ordinarily utilized by the Indians— deer, antelope, and elk, for example—are missing from this list; fur- thermore, bird and fish remains are entirely absent, except for the two artifacts made from bird bone previously noted. Horse bones are like- wise absent. Approximately one-third of the dog bones and 22 per- cent of the bison bones are from immature individuals. It is not pos- sible to arrive at a sound estimate of the numbers of individuals repre- sented by the bison and dog bones, but, on the basis of the number of the most abundant element present (divided by 2), it can be said that there are at least 11 bison and at least 3 dogs. Probably considerably more individuals than these are represented. Molluscan remains were scattered throughout the site, but by far the largest number were found in a single deposit, Feature 29. As identi- fied by Dr. J. P. E. Morrison, they include six species of fresh-water mussels, most of which are represented by only a few specimens. Of the 750 identified shells, 698 belong to the species Lampsilis siliquoidea (Barnes) ; others are Lasmigona complanata (Barnes), 43 specimens; Lampsilis ventricosa occidens (Lea), 5 specimens; Amblema costata (Rafinesque) ; and Quadrula quadrula and Anodonta grandis plana (Lea), 1 specimen each. Seventeen fossil snail shells whose origin is the Fort Union formation are probably accidental inclusions in the cultural deposits; the species are Ceriphasia nebrascensis (Meek and Hayden) and Compeloma multistriata (Meek and Hayden). CONCLUSION The investigations in the Heart Butte Reservoir area, limited as they have been, provide some information relative to the aboriginal utilization of an area which until now has been entirely unknown ar- cheologically. The number of sites revealed by the fairly intensive reconnaissance of a restricted segment of the Heart River valley indi- cates that the area west of the Missouri River in North Dakota was not avoided by aboriginal peoples, but much additional investigation will be necessary before even a sketchy knowledge of the prehistory of the Bay Regi” =«© HEART BUTTE RESERVOIR AREA—COOPER 33 area is acquired. The excavations in the Koehler site, although on a regrettably small scale, constitute a first step in the acquisition of such knowledge. The evidence revealed by the excavations in the Koehler site suggests occupation by small groups, probably over a period of some years. There is no evidence, however, of any cultural change from bottom to top of the cultural deposit, so no considerable time span is indicated. No evidence of structures of any kind was uncovered, nor was there any evidence of storage pits, which facts lead to the conclusion that the mode of life was unlike that represented by the more or less perma- nent earth-lodge villages on the Missouri River to the east. The absence of remains of corn or other cultivated plants and of such agri- cultural tools as scapula hoes, while not conclusive, in view of the gen- erally scanty specimen yield, suggests the probability that the inhabi- tants were not engaged in agriculture during their occupancy of the site. On the other hand, the relative abundance of bones and of mussel shells indicates a heavy reliance upon hunting and gathering for food. If the bones found in the site can be considered conclusive, the bison herds were exploited almost to the exclusion of other game ani- mals; no deer or elk bones have been identified in the collections. The extremely fragmentary condition of the bones, probably indicative of the extraction of bone grease, would seem to imply that the site was not simply an overnight stopping place for groups on the move, but that it served as a headquarters during the hunt. The artifact inventory, while leaving much to be desired from the standpoint of quantity and range of forms, seems to point unmistak- ably to a close relationship with certain sites which have been investi- gated on the Missouri River near the mouth of the Heart River. Spe- cifically, resemblances to artifacts from the Double Ditch site, for which comparative data are available, seem very close. Many arti- facts reported for the Double Ditch site are lacking in the Koehler site, as is to be expected in view of our small sample, but, on the other hand, it appears that all of the pottery and other artifacts from the Koehler site can be duplicated in the large collection from the Double Ditch site. The lack of metal and other White contact materials in our excavations is compatible with such a relationship, for the only items of trade goods reported from the excavations in the Missouri River site are two pieces of copper (Will and Spinden, 1906, p. 168). It is suggested, in view of its location, its general character, and its apparent cultural relationships, that the Koehler site was a recurrently occupied hunting camp of a people closely related culturally to the occupants of the Double Ditch site. These occupants were presumably Mandan, since the Slant Village, virtually identical culturally (Strong, 1940, p. 363), was identified with this tribe by one of Lewis and Clarke’s informants. While the absence of materials of Caucasian 34 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 origin cannot be considered proof of a precontact date for the site, there is no doubt that a date prior to the acquisition of large quantities of such materials is indicated. This negative evidence, together with the positive evidence of a close cultural relationship to the Double Ditch site, points to occupancy in the 18th century, and perhaps in the first half of that century. It appears probable that most of the other sites recorded in the reservoir area (see Appendix) can be similarly identified, although the collections are too scanty to permit more than tentative identifications. In each instance where pottery was recovered, it resembled that from the Koehler site. Whether the sites which yielded only flint materials represent different complexes or are simply workshop areas cannot be stated on the basis of present information. At the moment, evidence of any appreciable time depth in the area is lacking, but further investi- gation may alter this situation. At any rate, there seems no doubt that the western tributaries of the Missouri River constitute a fruit- ful field for future archeological research. LITERATURE CITED Dick, LEE R. 1948. The biotic provinces of North America. University of Michigan. FENNEMAN, NEVIN M. 1931. Physiography of Western United States. New York and London. HEWES, GORDON W. 1949 a. The 1947 summer field session in archeology, University of North Dakota. Notebook No. 1, Lab. Anthrop., Univ. Nebraska, pp. 21-24. 1949 b. Pottery from the sites excavated by the 1947 North Dakota field session. Notebook No. 1, Lab. Anthrop., Univ. Nebraska, pp. 58-67. LEECHMAN, DOUGLAS. 1951. Bone grease. Amer. Antiq., vol. 16, No. 4, pp. 355-356. LEONARD, A. G. 1912. Geology of south-central North Dakota. State Geol. Surv. North Dakota, Sixth Biennial Rep., pp. 21-99. MULLOoy, WILLIAM. 1942. The Hagen site. Univ. Montana Publ. Soc. Sci., No. 1. ORCHARD, WILLIAM C. 1916. The technique of porcupine-quill decoration among the North Ameri- can Indians. Contr. Mus. Amer. Ind., Heye Foundation, vol. 4, No. 1. Over, W. H. 1936. The archaeology of Ludlow Cave and its significance. Amer. Antiq. vol. 2, No. 2, pp. 126-129. STRoNG, WM. DuNCAN. 1940. From history to prehistory in the northern Great Plains. Smith- sonian Mise. Coll., vol. 100, pp. 353-394. TISDALE, ERNEST BH. 1941. The geology of the Heart Butte quadrangle. North Dakota Geol. Surv., Bull. 13. Ry. BAS. Sur. Par. No)” HEART BUTTE RESERVOIR AREA—COOPER 35 WEDEL, WALDO R. 1955. Archeological materials from the vicinity of Mobridge, South Dakota. Bur. Amer. Ethnol. Bull. 157, Anthrop. Pap. No. 45, pp. 69-188. WILL, GEORGE F. 1924. Archaeology of the Missouri Valley. Anthrop. Pap. Amer. Mus. Nat. Hist., vol. 22, part 6. WILL, GrorcE F., and HEcKrER, 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, archaeology, and language. Pap. Peabody Mus. Amer. Archaeol. and Ethnol., Harvard Univ., vol. 3, No. 4. APPENDIX SITES IN THE HEART BUTTE RESERVOIR AREA Following is a list and brief description of sites found in the Heart Butte Reservoir area by the River Basin Surveys in 1946 and 1948 and by the University of North Dakota-State Historical Society party in 1947. Except for site 32GT1, where fairly large-scale excavation was accomplished, and site 32GT5, where minor testing was done, none of these sites was excavated, even on a small scale. Site 32GT1 (the Koehler site), a camp site of a pottery-using group in the NE% sec. 9 TI36N R89W, was probably the most intensively occupied site in the reservoir area. Both the River Basin Surveys and the North Dakota University-Historical Society party undertook excavations here. The description of this site comprises the body of the present report. The site is flooded by the reservoir. Site 82GT2, on a low terrace on the right side of the Heart River in the NE%4 NW), sec. 9 T136N R89W, yielded bone fragments at the time of the reconnaissance. The landowner reported pottery and flint artifacts had been found here in the past. Situated just across the river from the Koehler site, it may relate to the same occupation. It lies within the pool area and has been flooded. Site 82GT3, on the left side of the river in the NE14 NEY, sec. 9 T136N R89W, was on the same terrace as the Koehler site, but across ashallow ravine. Fragments of bone only were observed by the recon- naissance party and Hewes reports that the same material was found in garbage pits dug by his party. It may be an extension of the Koehler site and has also been flooded. Site 32GT4 consists of 5 groups of stones, 1 of which constitutes ap- proximately half a circle and 3 of which are groups of stones some- what resembling it but being rather irregularly arranged. The fifth has been disturbed by excavations of the landowner, who assumed it marked a grave but who reports having found no bones or other evi- dence. The site is in the SE% sec. 4 T136N R89W and is within the area flooded by the reservoir. Site 32GT5, in the NW], sec. 13 T136N R89W, is a small, shallow rock shelter formed by the erosion of very soft sandstone underlying more firmly cemented materials. Numerous fallen slabs lying on the slope below the shelter and elsewhere in the vicinity (pl. 12, a, 6) sug- gest a lack of stability in the situation which would preclude the pos- sibility of occupation over a long period of time. The present floor 37 38 BUREAU OF AMERICAN ETHNOLOGY [BULL 169 is composed of an exceedingly fine sand. The shelter faces to the northwest, toward the prevailing winds, and on windy days would be untenantable; sand whipped up by such a wind made work almost impossible during a half day spent here by the River Basin Surveys party. Evidence of casual occupation was found by the Hewes party in 1947, when a few sherds from a single pottery vessel, points, and many spalls and chips are reported to have been found (Hewes, 1949 a, p. 22). The pottery, which is apparently in the same tradition as that at the Koehler site, has been described by Hewes (1949 b, p. 61). It has a grooved and ridged body and a concave-convex upper rim decorated with a band of diagonal cord impressions (pl. 12, ¢, I, 7). A point in the collection on loan from the State Historical Society of North Dakota (pl. 12, ¢c, 8) is small and triangular and has a slightly concave base and a pair of side notches. The sides are serrated. The spalls and flakes are of “Knife River flint.” During a very brief examination of the site in 1948, the basal frag- ment of a small, probably triangular, point with a concave base and a number of flakes and spalls of “Knife River flint” were found on the surface. In addition, a small sherd from the lip of a vessel and a few additional fragments of “Knife River flint” were collected just beneath the surface in a 5- by 10-foot test trench at the outer edge of the shelter. Otherwise the trench yielded no certain evidence of occupation, al- though there were occasional flecks of charcoal mixed with the other- wise sterile fine sand. The sherd is too small for satisfactory identi- fication, but could very well be from a vessel like a number represented by the sherds from the Koehler site. Immediately above the dam, the site is believed to have been destroyed by flooding. Site 82GT6 is a deposit of bison bones, exposed in the eroding bank of the Heart River in the NE% sec. 8 T136N R89W. At the time it was observed in 1948, the deposit was 0.1 to 0.75 feet thick, but it was reported that, previous to recent cutting by the river, it was as much as 3 feet thick. The top of the layer of bones was at a depth of about 10 feet beneath the surface. A fragment of a projectile point, perhaps corner notched, was found at the edge of the river below the deposit, but it may not have originated in the bone deposit. Presumably a bison kill, this site would undoubtedly have repaid investigation had resources been available. Lying on the immediate bank of the river, it was flooded soon after water storage began. Site 32GT8, in the SW sec. 10 T136N R9IOW, is apparently a small camp site which occupies a terrace finger on the south side of the Heart River. Four small pottery sherds, collected from the terrace slope below the site, are decorated with cord-impressed lines and re- semble pottery from the Koehler site. The site is within the area of flooding. Riv. Bas, SUR. HEART BUTTE RESERVOIR AREA—COOPER 39 Site 32GT9 is a single boulder circle reported to lie in the SEY SW), sec. 1 T136N R90W, on the south side of the Heart River. No artifacts were found on the site, which has been covered by the reser- voir waters. Site 32GT10 is an occupational area, apparently a small camp site, on a terrace on the north side of the Heart River in the NW, SEY sec. 2'T136N R90W and within the reservoir pool area. The cultural evidence on the surface consisted of a single cord-impressed rim sherd, similar to those found in the excavations at the Koehler site, two re- touched flakes, and nine unworked fragments of stone. ‘The stone is “Knife River flint” and gray chert. Site 32GT11, in the NW14 SW sec. 3 T136N R9OW, is an occupa- tional site buried beneath slope wash on a terrace which is being cut by the Heart River. The occupational layer, in which basin-shaped fireplaces and deposits of bison bones and mussel shells were observed, lies an average of 3.5 feet beneath the present surface. Two pottery sherds, one plain and one simple stamped (bearing parallel grooves and ridges), a triangular chert end scraper, a broken blade of “Knife River flint,” and cores, spalls, and flakes, mostly of “Knife River flint,” and an unfinished bone awl were collected. The site has been flooded. Site 32GT12 is on high land on the north side of the Heart River in the SEY SW sec. 85 T187N R9OW. Numerous flakes and a few cores, all of “Knife River flint,” were collected from the surface of the site, which may be a quarry, since the materials are reported to be exposed in shallow depressions. The site is apparently not subject to flooding. Site 32GT13, in the SEY, SEY, sec. 3 T136N R90W, is a camp site lying on a terrace about 15 feet above the Heart River and on the north side of that stream. Four small simple-stamped sherds and fragments of “Knife River flint” were found on the surface of the terrace, which lies well below the reservoir level. Site 32GT14 is a camp site on a fairly low terrace on the north side of the Heart River, in the NE14 SW sec. 34 T137N RO9OW. Ten small sherds, one bearing decoration with single cord impressions and others with simple-stamped surfaces, and a number of chalcedony fragments, mostly “Knife River flint,” were collected from the sur- face, which is now covered by the reservoir. Site 82GT15 occupies the crest of a knoll on the north side of the Heart River in the NW14 NW1){ sec. 34 T137N R9OW and within the flooded area. Fragments of bone and of flint and other stone were observed on the surface and a single end scraper of “Knife River flint” was collected. 40 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 Site 32GT16 is a small area on the north side of the Heart River which yielded a number of flakes and cores of “Knife River flint.” Three depressions, which may represent quarry pits, are reported. The site is in the W14 W134 sec. 35 T137N R9O0W within the flooded area. Site 32GT22, not observed by the River Basin Surveys party, is reported by Hewes (1949 a, p. 22) to be a quarry yielding chalcedony spalls and a few tools in the SW NEY sec. 1 T136N ROOW. It occu- pies a knoll on the north side of the Heart River and will apparently be flooded. Site 32SK4 is a pottery-bearing camp site on a low terrace on the north side of the Heart River in the NW14 SWY4 and SWY4 NW, sec. 12 T137N R91W. Approximately 60 small sherds, 2 end scrapers, and a quantity of chips were collected from the surface. The sherds, most of which are simple stamped, resemble the pottery from the Koehler site. This site probably lies slightly above the reservoir pool and, accordingly, will be available for future investigation. ue ata nee sia seri } ¥ rl pa an iy a a a an BULLETIN 169 PLATE 1 BUREAU OF AMERICAN ETHNOLOGY “IOATY Jieof{ SsOl9v WIOJ; YOU OT, “q Ss el eee I [ty quaselpe WO] (wvaij}sumop) ysvo OT, “vp "(L.LOZE) es Ja[Yaoy Jo smorA PR fs nt BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 2 Koehler site (32GT1) during excavation. a, b, Excavation trenches. Heart River in background. c, Excavation unit 1 in early stage, toward site west. Excavation floor nearest camera is at top of cultural deposit. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 3 Koehler site (32GT1). a, Profile on W15 line, N50 to approximately N62, with zones marked. ‘Top to bottom, zones B-F. Fireplace, Feature 32 at left. 6, Typical profile of N45 line, W70.3 to W75.2, with zones marked. Top to bottom, zones A-F. c, Profile on W15 line, N85 to N92, showing deposit of mussel shells, Feature 29, and beneath it, shallow pit, Feature 43. Recent pit at right. d, Pottery fragments, Feature 28, in situ. e, Pottery fragments, Feature 30, in situ. BULLETIN 169 PLATE 4 BUREAU OF AMERICAN ETHNOLOGY Jo 941g oa) SUA ONAN ES) "OUI] S/T N uO ‘7 oinqeay ‘sov[daig ¢¢ ainjeag ‘aov[dary fo a[yog ‘ q -? “Cas te Sale: SE, A V souoz ‘u10}}0q 0} doy, “oury] > “Nae c P| ee Za a LV pevg Bi SAG of i eS) be see N uo eyoud yeoidA 7, Ne ‘ D “(LLOZE) aus 1914903 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 5 b, Discarded unidentifiable 0, showing nature of cultural debris. 5 c, Top of deposit of mussel shells, Feature 29. 1, square N170W in excavation unit bones recovered from Feature 19. a, Top of cultural deposit Koehler site (32GT1). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE Pottery rim sherds from the Koehler site (32GT1). Interiors of profiles to left. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 7 CM Pottery sherds from the Koehler site (32GT1). All are rims except 9, which is from the shoulder area. Interiors of profiles to left. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 8 Stone artifacts from the Koehler site (32GT1). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 9 Chipped stone artifacts from the Koehler site (32GT1). J-I1, blades; /2-/4 retouched flakes. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 10 CM Large blades and ‘“‘choppers” from the Koehler site (32GT1). BULLETIN 169 PLATE 11 BUREAU OF AMERICAN ETHNOLOGY F2G7T1- 2 ut =. iS eo £7) Koehler site (32GT1). Bone and antler artitacts from the BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 12 a, Site 32GT5, rock shelter (arrow), to southeast. 6, Site 32GT5, rock shelter, to northeast. c, Artifacts from Koehler site (32GT1) and rock shelter, site 32GT5. from Koehler site; /, 7, 8, from site 32GTS. 2-6, 9, body sherds SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 169 River Basin Surveys Papers, No. 10 Archeological Investigations at the Tuttle Creek Dam, Kansas By ROBERT B. CUMMING, Jr. 41 422141—58——_5 had Visas) POY Uehe Py) es, .~ ‘+ ‘ , ) cry rei WAST hei *\) UMreagagie bs . Ne Ce heey Nap . s s, . er , a ftiath 4o9T) ohiw'i ot le anotanu a ‘Saeed ee VATE CED PARAS ED Oe CONTENTS PAGE LES DCT STG get Sm vad RA A ah es lM Seti A ee ed i 45 Prciatioitind eh vironment. eet 9. eit Tao) allt oP nits Poe ee 46 Enthnological and archeological background --_-_--____-____---___-___- 47 Dwenecsee: Mound: (f4PO%4) 22526 2S SS ke 48 LCST ae lee ee Pa Ey eee ee ee ey ee Sep ee 48 PRD AEAC UN ae Mee at RATS Secs ee, an Deg AA tre 5 chee a eee: 52 Cais elusiOnas 22 ot aa see Sn Te eT Pin eae se aoe ae 54 Wend pulwaynciie (PaP OU) 2 ek ce oN eee oe ee eek el ee Sk 56 LADEN OR (Ovi fr thoi MMA, 5 SEs SAS? Sec CRS cM ss BAY yal See 8 emer 6 bt See oS 56 "NLT CG eral Sg pee ce Ree te aaa ET SRE? SEE wig pe San SOE ee OO 57 AP Rete GON ta Nie ete ey Ee a og Le Se te he 57 snrre Rly ibe ke Ol) eae Se Le ee ee eg 58 | SMTVEYYG ASSET [Di aan ene Rae a i, 5 Aa ee Re ae SN a ae a ae 58 PACUTAGUS Se eas Bees Se ee Sie Re 2 SNe CO ke ene A A er ede 58 POLO EDs S010) 1s] ee nn EN aes Me ek Se OP, RR ae eS 60 Pie ed ee 9 Depa BT i base tg hi Es wie ge eT tin 8 Bile os ws 61 SERA Te Sete eee SON Up ee. ee Maa eer aE a 62 LTT CEESEY In ELA CYr aN EI i eel eee Se eS a 62 Appendix. Skeletal remains of the Sweat Bee Mound site (14P014)_____ 65 natn ereneme enter ss keke Mac Nie Ow BE EL AEG ea 2 8 65 RUC Alivcys [1h SNe Ae Be Oa a) Re ee BY Ee eae ON eLe BENE ey.) SM—y Pe ee eee 65 OURDIT ONL (6 (1 ts NM al le Ot Og OO aa ee A Ah om ee 65 CP LTIRPIS Vere TS Eis gfe allt ROE MY NSS ESLOL UI Ps Se oe a CP yee ee 68 PPS ELAM CA Gham Ute Je teat melo ey ae se ce kg nana hay ae Ret oy eet 69 RuUmiiary nd: COWCMSIONS 2. ote ao! oe wee en eters oo ee 70 ied bel ease semen Na ae: Cpe OP URS eG oe a US lo ial ILLUSTRATIONS PLATES FOLLOWING PAGE 13. a, A view of Sweat Bee Mound before excavation. 6, Sweat Bee Mound after clearing off the overgrowth__._..._.-_.__-_---- 78 14, a, Sweat Bee Mound after removal of the soil fill. 6, Excavating Sweat Bee Moundis ats 2 e251 Aes PAD TM ene oe) se ay Be eae A 78 15. a, Excavating and mapping Sweat Bee Mound. 5, View of the upper burials of Feature 1 of the Sweat Bee Mound_________________- 78 16. a, View of the lower burials of Feature 1 of the Sweat Bee Mound. 5}, Views of the male skull of burial 2, No. 50, of the Sweat Bee iglini sitencss LL See BPS Ae eee eeynne td AAW eed ee tee 78 17. Views of the female skull of burial 1, No. 49, of the Sweat Bee Mound Thiet. 8, 2 Lee Cee OE Ore ee Se ee Ow eee ee Pere eee 78 18. Artifacts from the Sweat Bee Mound site_____._._.______________- 78 19. Artifacts from the Sweat Bee Mound site_.._____._._...--_--_-_-- 78 20. a, A view of the remaining portion of the Spillway site. b, Testing Vie Bie arin Eten ees ee oe ME ee RS ONE ihe Le Pe. 78 2S Aritacts (rom: the omillwan Site. . 22220-20630 oe oe tcc cen 78 44 CONTENTS PAGE 22. a, Testing at the Reany site. Spillway site above cut in background. b, The Spillway and Reany sites being destroyed by construction activities. Sweat Bee Mound in the foreground _-_-_------------ 78 23. Artifacts from the Reany site__.-.-_----------------------------- 78 24. Artifacts from the Reany site___--.------------------------------ 78 MAPS 3. Site location map of the Tuttle Creek Dam area__------------ facing 46 4. Site map of Sweat Bee Mound (14P014)-------------------------- 50 ARCHEOLOGICAL INVESTIGATIONS AT THE TUTTLE CREEK DAM, KANSAS’ By Rosert B. Cummrne, JR. INTRODUCTION The Missouri Basin Project of the Smithsonian Institution River Basin Surveys, conducted archeological excavations in the construc- tion area of the Tuttle Creek Dam site from June 10 to June 30, 1953. These field activities and the present report resulting from them were a part of the Inter-Agency Archeological Salvage Program. This program combines the cooperative efforts of the Smithsonian Institution, the National Park Service, the Corps of Engineers, the Bureau of Reclamation, and various State and local agencies. De- tails of the organization, background, and accomplishments of this program are published elsewhere (e. g., Brew and others, 1947; Wedel, 1947; Roberts, 1952) and need not be detailed here. Lack of funds prevented extensive fieldwork in the season of 1953 in any of the lo- calities of the Missouri River Basin. However, some money was avail- able and this was diverted to the areas of most critical salvage needs. Tuttle Creek was one of these. Previous archeological investigations in the area to be flooded by the Tuttle Creek Reservoir had located upward of 119 sites, 9 of which were situated in the immediate construction area of the dam (Solecki, 1953 a, p. 6). Of these, 5 would not be destroyed until the later phases of dam construction but 3 were already partially destroyed and the fourth was in imminent danger of destruction from dam-building activities. The field party, limited by funds to only 3 weeks, selected 1 site (14PO14) for complete excavation and conducted sampling tests in the other 8. The field party was under the direction of the writer and James M. Shippee, of the Smithsonian Institution, River Basin Surveys staff. Linwood L. Hodgdon, assistant professor of Anthro- pology at Kansas State College, acted as field assistant. Marvin Carl- son, Randall Weeks, Alfred Johnson, and John Hennes made up the remainder of the crew. The writer wishes to express sincere apprecia- tion for the splendid work of the entire crew during the field season. Long range direction of the project was given by Dr. Frank H. H. 1 Manuscript submitted April 1954 ; some revision July 1956. 45 46 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 Roberts, Jr., Director of the River Basin Surveys, and Robert L. Stephenson, acting chief of the Missouri Basin Project. The assist- ance given the field party by Burney V. Reany, project engineer, Corps of Engineers, and by the many others who so generously volun- teered their help is gratefully acknowledged. The writer also wishes to express his apprecation to Dr. John L. Champe, of the University of Nebraska, for his assistance and advice; to Dr. G. Neumann, of the University of Indiana; and to Dr. T. D. Stewart, of the National Museum, for their criticisms and suggestions; and to Roy W. Drier, of the Michigan College of Mining and Technology, for the metallo- graphic examination of the copper specimens. At the termination of work on site 14RY10, evidence of a dwelling structure had been found, but time limitations did not permit excava- tion of the structure. Subsequent investigations at this site were made possible by a volunteer party organized by Dr. Linwood L. Hodgdon and consisting of himself, John Hennes, Warren Shaw, and James Tuback of Kansas State College; Dr. John L. Champe, Mary Kiehl, and Raymond Wood of the Laboratory of Anthropology, University of Nebraska; Harold A. Huscher of Columbia University; and Dr. Theodore E. White, Lee Madison, and Francis Brown of the Missouri Basin Project. This volunteer party worked at the site for about a week in July 1953. It is contemplated that Dr. Hodgdon and others will cooperate on a separate report of the overall activities at this site in the near future. Consequently only a brief account of the testing accomplished here in June 1953 by the Smithsonian Institution party will be included in the present report. SITUATION AND ENVIRONMENT The proposed Tuttle Creek Reservoir area is located in the Big Blue River basin in northeastern Kansas (map 3). The dam site is located approximately 12 river miles upstream from the junction of the Big Blue River and the Kansas River, or about 6 miles north of the town of Manhattan, Kans. The Tuttle Creek Dam, a Corps of Engineers project, is to be an earthfill structure 7,350 feet long and 136 feet high. Its purpose is the storage of flood waters from the 9,550 square miles of drainage area of the Big Blue River, above the dam. The Tuttle Creek area lies within the Interior Plains physiographic division, Central Lowland province, dissected till Plains section (Fen- neman, 1931), and within the Tall Grass, Prairie Grassland vegetation area of Shantz and Zon (Kroeber, 1947). Locally the Big Blue River flows through a flat alluvial valley, a little over a mile in width, with the dissected tableland bluffs rising several hundred feet above the flood plain. In the area as a whole, dark brown silty soils underlain by a yellowish brown subsoil are characteristic. The rich, alluvial soil A3AA jac waaAO 3 ~ AM HOFFAOOS aT 12 to on mart bslanba 27.10 ae . 4 . Fe rp Gan ‘ ; ; O—58 (Face p. 46) 422141 ‘| = . 14P014 TUTTLE CREEK DAM AREA SITE LOCATION MAP Adapted from Corps of Engineers Map No. A-9-7-1I4 T331 9-22-1953 Map 3.—Site location map of the Tuttle Creek Dam area. 4° °° of } eae ee Me _ « ef ES tye ee mmo 8 eer , i + + ¥w 2 it 3 i ed : : iy 2% TaSt.. : 1.2.4 i ‘ é Dar eRe | Ieekr to Cami ndigagal) ajie-—.t aa q ; vy * i : 4 . / re r ‘ i} it)3 obs, Om y ‘ oy ; i . : ik i ie! : y A, 4 ‘ ‘ f hd ae v , a | , : TP lt Att tah Pr iw ‘i i. 5 i “Ger i ae, Pa : ; y ae ne eh , > - 4 k t % ae ; J ; : , i eo afi gn wens We : 7 ' i aT Oy Fy ee | | | | pear | Riv. Bas. Sur. Pap. No. 10] TUTTLE CREEK DAM, KANS.—CUMMING 47 of the flood plain is excellent for farming, and the bottom lands are and have been under cultivation for many years. The upland slopes are, for the most part, in pasture. On the top of the bluffs a shallow soil cover is underlain by the limestone and shales of the Wolf Cap group of the Permian. The climate throughout the general region is characterized by warm summers, with an average temperature of 70° to 80°, and cold winters. Temperatures are very changeable and show a wide range of variation. Maximum temperatures as high as 121° and minimum temperatures as low as 40° below zero have been recorded (Kansas Agr. Exp. Sta., 1937, p. 5). The prevailing winds are from the south in summer and from the north in winter. The growing season averages 170 to 180 days. Average annual precipitation is from 32 to 34 inches, with about 70 percent falling within the warm season (U.S. Dept. Agr., 1941), thus making the region quite suitable for primitive horticulture. There is an abundant growth of trees and shrubs along the river valley. The following have been observed within the reservoir area: cottonwood, elm, burr oak, walnut, hard maple, sycamore, willow, lin- den, ironwood, box elder, buckeye, hackberry, juniper, honey locust, pignut, redbud, dogwood, buck brush, and sumac. Food-bearing plants native to this area include mulberry, plum, hazelnut, elderberry, gooseberry, chokecherry, and grape. Game formerly abundant throughout this area included elk, deer, antelope, bison, bear, wolf, cougar, wildcat, otter, and turkey. At the present time mink, muskrat, red squirrel, gray squirrel, woodchuck, beaver, opossum, raccoon, coyote, and rabbit may all be found. Quail and prairie chicken inhabit the uplands and ducks and geese are and were plentiful during the migratory season. The valley of the Big Blue River, in northeastern Kansas, provided excellent resources for hunting and gathering cultures, and an exten- sive utilization of long duration of these resources is indicated by the Survey’s location of approximately one hundred nonceramic sites in the Tuttle Creek Reservoir area alone. ETHNOLOGICAL AND ARCHEOLOGICAL BACKGROUND During the historic period the Tuttle Creek Dam area was well within the habitat of the Kansa Indians. To the north of their range, near the Nebraska border, were the Pawnee, to the south and southeast the Osage, to the east across the Missouri the hunting range of various tribes such as the Iowa and Sac, and to the west, in the high plains, the hunting ground of the Apache, Comanche, Sioux, Cheyenne, and others. During the beginning of the historic period the Wichita in- habited the area about the great bend of the Arkansas River. According to Kansa tradition, the Kansa moved upstream along the Missouri River as far as the Nebraska border and were then forced AS BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 southward (Wedel, 1946, p. 6) ; however, this has not as yet been veri- fied by archeological investigations. The distribution of identified Kansa village sites, recorded by Wedel (ibid., p. 2), shows the Kansa to have been along the Missouri River, at the Doniphan site, near the town of Doniphan in 1724; along the Kansas River, at the Salt Creek site, about 6 miles from the town of Leavenworth in 1757; at the “Old Kansas village” reported by Lewis and Clark, between the junction of Soldier Creek and Cross Creek with the Kansas River in 1804; and at the Blue River site, about 2 miles east of the town of Manhattan, and about 6 miles downstream from the Tuttle Creek Dam from about 1800 to 1830. Three known village sites, between the junctions of Cross Creek and Soldier Creek with the Kansas River, were inhabited dur- ing the period of 1830 to 1846. From 1847 to 1873 the Kansa lived in three villages below the town of Council Grove on the Neosho River. From there they were removed to Indian Territory. The section of the Big Blue River valley to be inundated by the Tuttle Creek Reservoir was virtually unknown archeologically until the summer of 1952. At that time a Smithsonian Institution recon- naissance party consisting of Ralph S. Solecki and James M. Shippee examined most of the area. The reconnaissance of the area, even yet incomplete, yielded evidence of 119 archeological sites. The only ex- cavated site near this area was the Blue River site, the historic Kansa village, 2 miles east of the town of Manhattan (Wedel, 1946, p. 2). The abundance of archeological remains that have been and are being found in this area clearly indicate that northeastern Kansas has been inhabited from very early prehistoric to historic times. Suggestions of Paleo-Indian occupation are found here as evidenced by occasional artifacts identified with the Folsom and Plainview groups (Solecki, 1958 b, p. 52-53, and Shippee, 1953, p. 54). Sites of other preceramic or nonceramic groups are abundant. Cultural manifestations of Woodland, Upper Republican, and Nebraska cultures have been found, and it is quite probable that further investigation will also dis- close elements of the Oneota and Hopewellian cultures. SWEAT BEE MOUND (14P014) FIELDWORK This site is located in Pottawatomi County, Kans., in the NW14 NE}, sec. 19, RSE T9S. It consists of a rock mound situated on a prominent elevation along the top of the ridge bordering the east side of the Big Blue River flood plain (pl. 13, a). Site 14PO13 is about 500 yards farther north along the same ridge. From the mound, which is about 180 feet above the Big Blue River, a fine view may be had of the valley below. The slopes of the ridge are wooded, and the top sparsely covered with grass and brush, and studded with limestone Riv. BAS. Sur. BRO toy”)«XTUTTLE CREEK DAM, KANS.—CUMMING 49 rocks. Beneath a shallow, dark, soil zone at the surface is a gravel stratum underlain by bedrock limestone and shale. The method of excavation consisted of first clearing the mound of grass and brush (pl. 18, 6), laying out a grid of 10-foot squares, and establishing a site datum. The partial dirt cover over and between the rocks was then removed, outlining the entire extent of the mound. Surface elevations were taken and the mound was mapped (pl. 15, a). The rock-slab covering was then removed, leaving a small profile section in place until excavation of all else was completed. The en- tire mound fill was screened (pl. 14, d). Before excavation the mound appeared as a small, circular, earth and stone hummock, grass-covered between the rocks with a clump of brush near the south side. After clearing off the brush, grass, and dirt, above and between the stones, the mound was found to be irregularly circular in shape, 26 feet in diameter and 114 feet high (map 4). The mound structure consisted of a pile of irregularly shaped limestone slabs, placed without any semblance of order, but in sufficient quantity to completely cover and protect that which lay beneath. The covering rocks ranged from small fragments to slabs 3 feet long, 114 feet wide, and nearly a foot thick (pl. 14, a). It seems most probable that the dirt over and about the stones had blown in during later years, and that the mound, as originally built, was a stone-slab mound. The remains of 7 individuals were found within the mound, 6 of whom were in 1 pit (Feature 1) and 1 (Feature 3) on top of the gravel underlying the mound. After the mound had been cleared off and some of the upper stones removed, a basinlike pocket was found in the rock covering, 7 feet in from the east edge and 3 feet south of the center of the mound. This proved to be the location of Feature 1. Feature 1 was an oval- shaped pit 3 feet in north-south diameter, 34 feet in east-west di- ameter, and 2 feet in depth from the surface of the gravel underlying the mound. Limestone slabs had been placed around the perimeter of the pit and rocks thrown within it, covering a group of three burials that lay within the pit at the level of the base of the surrounding slabs. Two distinct groups of burials were found within Feature 1. Burials 1, 2, and 3 were at the level of the gravel base of the mound, and were surrounded by the rock slabs (pl. 15, 6). Burials 4, 5, and 6 were within the pit below the base of the mound (pl. 16, a). The upper-level burials were mixed with soil and stones and covered over by a foot of stone rubble. The bones were crushed, badly weathered, and, to a considerable extent, jumbled together. Burial No. 1, in the center of the pit, was a semiflexed, primary burial. A large rock slab had settled through the midsection of the skeleton, that of an old adult female, crushing the bones beneath. ‘The 50 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 250NE100 90 NEIOO s +!90NEI60 SWEAT BEE MOUND SITE 14P014 LIMITS OF MOUND AND EDGE OF EXCAVATION CONTOUR INTERVAL = 0.2' T325 : 9-14-1953 Map 4.—Site map of Sweat Bee Mound (14P014). burial was oriented along a north-south axis, with the head to the north, facing east, and rested on its left side. The manner of displace- ment of the bones of the other upper-level burials in the pit indicated that burial No. 1 was the last to be placed in the pit, and that the other burials had been pushed aside to make room for it. Riv. BAS. Sur. — Pay, No 10] TUTTLE CREEK DAM, KANS.—CUMMING 51 Burial No. 2 was the remains of an old adult male. The right side of the skull was completely crushed and the skull was no longer in articulation with the spinal column. The state of preservation of the bones, combined with the fact that they had been somewhat displaced, made it difficult to determine the original manner of deposi- tion. It appears to have been a primary semiflexed burial, oriented north-south, with the head to the north. Burial No. 3, the bones of which are crushed and disintegrated, was that of an old adult male. It appears to have been a primary fully flexed burial, oriented southeast-northwest, with the head to the southeast. Many of the bones of the lower burials were badly disintegrated and jumbled. Burial No. 4, the central one, was that of a middle- aged adult male. It was a primary, semiflexed burial, oriented ap- proximately southwest-northeast, with the head to the southwest, facing north. The skeleton was on its left side. Burial No. 4 had evidently been placed in the pit at a later date than burials 5 and 6, for the latter appear to have been pushed to either side of the pit in order to make room for it. A group of 4 circular copper beads (14P014-116), 1 large and 3 small, were found resting on the right mandibular condyle, and 2 more (14P014-118) were located beneath the left temporal. An additional pair of the same type of bead was found while screening fill from the area immediately about the skull. Both the right and left mastoid process of the skull were stained green from the copper. The beads had apparently been attached to the ears at the time of burial. A small disk shell bead (14PO014-117) was also found close to the skull. Fragmentary bits of copper and a tubular conch shell bead (14PO014-119) were found close to one wrist of the skeleton. Burial No. 5, a semiflexed primary burial, was that of a middle- aged adult male. It was oriented approximately east-west, with the head to the west, facing to the north. Several of the long bones of this skeleton are unusually large. An ovate stone scraper (14PO014-131) was found directly over one of the scapulae. Burial No. 6 was the badly disintegrated remains of a middle-aged adult female. The outline in the soil indicated that it was probably semiflexed, with the head to the west. Two stone scrapers (14PO14-123 and 124) were found in the fill about the lower burials. A group of 4 chert nodules and 1 piece of hematite (Feature 2) were found in the gravel beneath the mound fill, 11 feet north of Feature 1. In addition to the burials of Feature 1, the fragmentary remains of an extended, supine, adult burial (Feature 3), burial No. 7, were found lying directly on the gravel deposit under the mound, and 52 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 about 5 feet north and 5 feet west of Feature 1. The lower leg and foot bones were present and in articulation but crushed by the over- lying rocks. A section of the right radius and ulna, and several phalanges indicated that the arm had been extended along side of the body, which was oriented northwest-southeast, with the head to the southeast. The remainder of the body bones were missing or frag- mentary. The outline of a pit was evident in the reference block of the mound profile at the spot where the skull should have been; hence the skull may have originally been present. No associated artifacts were found. Burial No. 7 appears to have been intrusive and to have no connec- tion with the other burials in the mound. Individual extended burials, placed beneath piles of rocks, were customary among the Kansa, and it is quite possible that the nearby site, 14P013, is a Kansa site; however, there was no certain evidence to indicate that the burial was intrusive. ARTIFACTS A total of 47 artifacts was recovered from the mound. No pottery was found. Thirty-five stone artifacts, mostly scrapers, cutters, and choppers, were recovered from the dirt between or below the rock-slab covering, and a few specimens were found in association with the burials. Four chert projectile points (pl. 18)? and the base of a point were found in the mound fill. Specimen No. 4 is a small, delicately chipped, straight-based and straight-sided triangular point with two small side notches and one base notch, all of the same size. The tip of the point is missing. Specimen No. 31 is a medium-sized point with sides slightly convex in outline, is flat on one side, and has an expanded stem containing a small, shallow notch. Point No. 29 is double con- vex in cross section, apparently straight based, and evenly chipped. Half of the stem and the lower portion of the blade is missing. No. 5 is a rather small, evenly chipped point, convex sided and concave based in outline. No. 37 is the base of a straight-based projectile point. Table 1 gives data on the projectile points. TasBLe 1.—Projectile point data Cat. No. Length] Width | Thick- | Descrip- Provenience ness tion 1 Mm Mim. | Mm J4PO14-4 -t ee tee es 11 2| NBa2 | Grass roots on mo LAE OTS 30s Bese eee 32 22 6 SCa2 Mound fill, 1.1 feet aban surface. TAR O14- 90 Mie tee Ee 239 23 5 SCa2 Base of mound fill. TP O14 pale. ee eee 25 17 3 NBb Mound fill, 0.7 foot below surface. 14'P O14 Oye Sep ee he Ol ee I) arma se ae Ue Mound fill. 1A descriptive codification of projectile points formulated by Strong, given as an aid for the comparison of jee ports with projectile point data recorded elsewhere from the Plains area (Strong, 1935, p. 88). stimate 2 Artifacts are identified in plates 18, 19, 21, 23, and 24 by their specimen catalog numbers. Riv. BAS. Sur. Par. No. 10] TUTTLE CREEK DAM, KANS.—CUMMING 53 The point end of a chert blade (No. 38) was found in the mound fill (pl. 18). The fragment, 33 mm. long and 26 mm. wide, is double convex in cross section and is neatly chipped. Two small end scrapers were found (pl. 18). Specimen No. 9, from the mound fill, is a snub-nosed scraper with an expanded bit, flat on both sides, and retouched on the end and part of one side. It is 25 mm. long, 16 mm. wide, and 6 mm. thick. Number 131, a small flat ovate scraper, is retouched on both sides and one end, and is 29 mm. long, 23 mm. wide, and 7 mm. thick. It was found directly over the scapula of a lower burial of Feature 1. Thirteen of the 35 chipped stone artifacts are scrapers (pl. 19). They are made from irregularly shaped chert spalls, and show a minimum of modification. The ventral surface is a single cleavage plane. One or more of the edges are retouched. No. 32 is a high- backed, subelliptical scraper with a concave ventral surface. The en- tire dorsal surface is flaked, and one edge is retouched. No. 1, found on the surface, is a nearly circular scraper. Its dorsal surface is con- vex, and the ventral surface has a prominent medial ridge. No. 8, also from the mound surface, is an ovate, stemmed scraper, and is flat ventrally and convex but unmodified dorsally. It is retouched around the entire edge. Two scrapers were found associated with the lower burials in Feature 1, Nos. 123 and 124. No. 123 has a broad blade that is more or less oval and tapers toward oneend. The ventral surface is moderately convex, the dorsal surface beveled, with the maximum thickness close to the left edge. The broad end and one side are re- touched. No. 124 is a rather small subelliptical scraper, concave ven- trally, beveled dorsally, with the maximum thickness close to the left edge. It is retouched on both ends and on one side. Table 2 follows. TABLE 2.— Measurements of scrapers} Cat. No. Length Width Thickness Mm Mm. Mm BPS Se SEE ot CR EE ee Le eee ne SEE eS SS Seen I eae eae 59 33 14 Da ae ace aac Ric: Laat loll 8 6 a a8 aia ME LIN apse Tie Lobes S| “RR a aya 44 40 20 ae Nie USE 1s AE ESS PR EP ere PE Re oe 47 44 110-5}. cet pasa ot BESS MER 0a OR BSN PD) RSs ee SS 71 54 12 192 Sake RS ER eee eee OE ee ee ee ee eee ee ee ee 52 27 9 UATICO ACE: S5re LOPS ER Bee oe ered peek Re A 34-76 24-55 7-29 PAV OREO U2 he oak = _ No emilee oie then Li OE esa ee A ot 52.5 36.6 12.6 1 Measurements are given for 5 of the scrapers illustrated (pl. 19), but the range and average are given for the total of 13 specimens. Five specimens are a type of cutting or scraping tool (pl. 19). They are of the spall variety, medium sized, generally ovate in shape, and bifacially flaked without secondary retouching. The flaking is coarse. The cutting edge extends all the way around or on two sides and one end. The implements range from 69 to 80 mm. in length, 40 to 54 mm. in width, and 17 to 23 mm. in thickness. No. 48 is of chert, oval in out- 54 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 line, double convex in cross section, with a depression in the center of one side. It is edged on both ends and one side, and measures 68 mm. in length, 47 mm. in width, and 19 mm. in thickness. This was found in association with the upper burials of Feature 1; the others all came from the mound fill. Five roughly modified cores show some evidence of use. Specimen No. 46 (pl. 19) is a large, coarsely flaked bladelike chopper, flaked on both ends and on one side. It is 102 mm. long, 76 mm. wide, and 25 mm. thick, and was found in association with the upper burials. A rectangular piece of reddish quartzite, 117 mm. long, 100 mm. wide, and 68 mm. thick, was found in the mound fill (pl. 19). The rock is smooth and the corners well rounded. It might have served as a hammerstone or as a small anvil; however, there is no clear-cut evi- dence of battering. A pendantlike object of catlinite (pl. 18) came from the mound fill. The specimen (No. 28) is subrectangular with one end having a small stem or projection 3 mm. long and 4 mm. wide. The object is 33 mm. long, 26 mm. wide, and 2 mm. thick. The two pieces of worked shell recovered were associated with the lower burials of Feature 1. One is a disk shell bead (No. 117) 6 mm. in diameter, and the other a cylindrical conch shell bead (No. 119) 27 mm. long and 11 mm. in diameter (pl. 18). A notch has been worn into the same side of each end of the bead. Feature 2 yielded a piece of worked hematite (No. 130). In outline it resembles half of a cross-sectioned elliptical object. The ventral and dorsal surfaces are flat, and the edges are rounded at the tapered end. The piece is 55 mm. long, 43 mm. wide, and 23 mm. thick. Eight copper disk beads (pl. 18) were associated with the lower burials of Feature 1. The beads, made by rolling copper strips 5 to 7 mm. wide and 2 to 3 mm. thick, range from 9 to 12 mm. in diameter, with a hole 4 mm. wide. Metallographic examination of the beads showed them to be of native copper. CONCLUSIONS The cultural identity of the mound is, as is so often the case with burial mounds, difficult or impossible to definitely determine. There is however, some reason to believe that the mound belongs to the Wood- land Complex. Two of the five points (Nos. 29 and 31), found well within the mound fill, resemble Woodland points attributable to the Valley Focus. The other three points do not resemble conventional Woodland types, however ; one of these occurred at grass-root level and hence may be unassociated with the mound. Burial mounds of earth and stone construction are numerous and widespread in Missouri, and are associated with the Woodland Pat- tern. The burials are sometimes in a pit dug into subsoil beneath the Riv. BAS. Sur. Par, No. 10] TUTTLE CREEK DAM, KANS.—CUMMING 55 mound, on a stone platform with or without a stone covering, in stone chambers or vaults, or in cists (Chapman, 1948, pp. 110-125). A group of stone vault mounds, ascribed to the Hopewellian Complex, is situ- ated at Kansas City, Mo., and stone vault mounds have also been found in Kansas, in the northern part of Doniphan County (Wedel, 1943, p- 159). The Younkin Mound, a rock and earth mound with a stone-slab floor, located about 6 miles northwest of Junction City, Kans., has been ex- cavated, and is ascribed to the Hopewellian Complex (Spaulding, 1949, p. 106). Its only resemblance to site 14PO14 consists of rock- covered burials and a highland location. Stone-slab mounds at site 14EW18, 10 to 30 feet in diameter and 2 to 3 feet in height, have been found on the high bluffs near the mouth of Bluff Creek, Kanopolis Reservoir, Ellsworth County, Kans. (Kivett, 1947, p. 9). These mounds are similar in external appearance to the Sweat Bee mound; however, they are unexcavated and their cultural identity is unknown. Burial cairns, excavated at site 14EW24, on the bluffs above the Smoky Hill River in Ellsworth County, are somewhat similar but con- siderably smaller than 14PO014. In one cairn, 12 feet in diameter, a flexed skeleton, crushed by the overlying rocks, was found in rec- tanguloid cist (Smith, 1949, p. 229). Subsurface burial pits, covered by an extensive layer of rocks, have been found in Nebraska. An example of this is the Lindsey site (25CC29) in Cass County, 114 miles west of the town of Weeping Water. Primary burials were found in five pits or cists, dug into the subsoil of the base of a rock-slab-covered area 30 feet in diameter. The site belongs to the Woodland Pattern (Kivet, Marvin F., per- sonal communication. Data on file at Nebraska State Historical Society). Turtle Mound, one of a series of mounds along the Missouri River bluff near the town of Rock Bluff, Nebr., proved upon excavation to be a 24-foot circular limestone slab-covered mound with one concen- tration of human bones close to the eastern border of the mound (Gil- more, 1932, pp. 167-169). A rock-covered burial pit containing four individuals in an oval- shaped pit was found at site 25NC201, 514 miles northwest of the town of Fullerton, Nance County, Nebr. This was on a high ridge over- looking Timber Creek. ‘The few sherds found in the pit indicated that the burial was Woodland (Cumming, 1953, pp. 8 and 9). A characteristic of the Woodland burial complex is the wide range of variation of the burial pattern. In Missouri both semiflexed and extended burials have been found within the same mound. Multiple, semiflexed, or flexed primary burials within mounds or pits in mounds, as at 14PO14, fall within the Woodland burial pattern. 56 BUREAU OF AMERICAN ETHNOLOGY [BuLL. 169 Although the Kansa, no doubt, frequented the region in which site 14P014 is located, it is doubtful that the mound burials are Kansa, with the possible exception of burial 7, which may be intrusive. While the Kansa often used bluff and hilltop locations for their burials, and usually covered the graves with rocks, they customarily buried their dead in individual graves and in an extended or semireclining posi- tion (Wedel, 1946, p. 27). The nearest known habitation site to 14PO014 is site 14RY10, a nearby village site adjacent to the west end of the dam axis. It is improbable that the mound burials are associated with this village, for preliminary investigations indicate that 14RY10 belongs to the Table Rock Focus of the Nebraska Aspect, and so far as is known, the Nebraska Aspect burial pattern is that of secondary burials in ossuary pits. It seems quite possible that the rock mounds and rock-covered pits containing multiple burials, and always placed on a highland loca- tion, found in northeastern Kansas and parts of Nebraska, may repre- sent an attenuated form of the Woodland burial mound building complex of Missouri. There seems to be a reasonable probability that site 14P014 is a manifestation of this complex. THE SPILLWAY SITE (14P012) FIELDWORK After completing work at the Sweat Bee Mound, 2 sites about 500 yards to the south, in the spillway area of the dam, were sampled. Site 14P012 was the least promising of the two, and little time was allotted to it. The Spillway site (14PO012) is located in Pottawatomie County, Kans., in the SW14 sec. 18, RSE T9S. The site is on the northeast side of the spillway of the Tuttle Creek Dam (pl. 20, a). Spillway excavation had, at the time of the investigation, destroyed over half of the site. The occupational area is situated on a relatively flat- topped ridge about 200 feet above the Big Blue River, and commands an excellent view of the river valley and of the hills bordering the tableland to the east. The soil about the area is rocky and sparsely covered with grass and sumac. The underlying bedrock of limestone and shale is covered with a gravel deposit containing many chert in- clusions. Over this is a thin covering of darker soil. Cultural ma- terials are spread over an area of about 1 acre and occur from the sur- face down to a depth of 0.6 to 0.8 foot. In the short time available, surface examination and the digging of one 3- by 4-foot test pit, 1 foot deep, were all that could be accomplished. From the surface, which was littered with chert chips, and from the test pit, a few sherds and chipped-stone artifacts were obtained. Riv. Bas. Sur. Be doy” TUTTLE CREEK DAM, KANS.—CUMMING 57 ARTIFACTS A total of 25 artifacts were obtained. The pottery sample consists of 12 small body sherds, all from the surface. These sherds are tem- pered with finely pulverized shell, their outer and inner surfaces are smooth but have a chalky feel, and the cores consist of a dark-gray paste. Colors range from light brown to buff, with the exception of one sherd which has an orange sliplike finish both inside and out. The ware is relatively thin, sherd thicknesses varying between 4 and 5 mm. Hardness ranges from 4 to 5. On all of the sherds the surface finish is plain, and three have narrow, sharply incised lines across them. The sample is too limited to permit making a cultural identification; how- ever, the sherds are suggestive of Glen Elder Focus material. The chipped-stone sample consists of 4 end scrapers, 7 side scrapers, 1 utilized core, and 1 blade or knife. All of the specimens are of gray or banded chert. Of the end scrapers (pl. 21), all found on the surface, 3 are medium sized and 1issmall. They are roughly rectangular, with the snub end rounded on two of the specimens. The under surface is unmodified and slightly concave; the upper surface fairly flat. Two of the scrapers are retouched on the front end and two retouched along the front and right side. The end scrapers range from 26 to 38 mm. in length, 21 to 31 mm. in width, and 5 to 11 mm. in thickness. The side scrapers (pl. 21) are of the spall variety. The chert chips from which they have been made are unmodified on one surface and retouched along one edge. A patinated portion, from the nodule from which the chips have been struck off, is present in three of the scrapers. The shapes are irregular, but. tend to be somewhat rectangular. One specimen (14PO012-5 (pl. 21)) is roughly semicircular and retouched along the working edge. A semicircular notch, 19 mm. wide, along one side, indicates that the scraper may also have been used as a bow-shave. Side scrapers range from 45 to 71 mm. in length, 29 to 47 mm. in width, and 10 to 15 mm. in thickness. One flat-bottomed, high keeled core shows evidence of use but no retouching on its work edges. The one blade or knife (14PO012-2) has a curved cutting edge and is coarsely flaked with the upper edge thickened and smoothed. It is 55 min. long, 30 mm. wide, and 9 mm. thick. CONCLUSIONS The artifact sample is too limited and nondiagnostic to make a cul- tural identification of the site. The sherds are suggestive of, but cer- tainly not identical to, material from the Glen Elder site (14ML1) in Mitchell County, and the White Rock site (14J W1) in Jewell County, Kans. It has been suggested that these latter two sites be assigned to the Glen Elder Focus of the White Rock Aspect (Kiehl, 1953, p. 4). 422141586 58 BUREAU OF AMERICAN ETHNOLOGY [BuLL. 16% Sites 14P012 and 14P013 are adjacent to each other and may be 1 site rather than 2. Scrapers are the most numerous of the stone arti- facts at both of the sites and are similar in style. On the basis of the pottery recovered, sites 14PO012 and 14PO13 are different sites; this may, however, be a reflection of unavoidable but admittedly inadequate sampling. THE REANY SITE (14P013) FIELDWORK The Reany site (14PO13) is located on the southwest side of the spillway, about 800 feet west of site 14PO012 (pl. 22, a). It isin the SW14SE% sec. 18 and NW1,NE\, sec. 19, RSE T9S. The site orig- inally covered an area of about 114 acres; however, at the time of investigation the north third of the site had been destroyed by con- struction operations. A day and a half was spent testing the remain- ing portion of the site. During this time the surface was intensively examined and nine test pits were dug in the area of the greatest sur- face artifact concentrations. Here, as at 14PO12, the surface was covered with chert chips, but sherds and chipped stone implements were more numerous than at the former site. ARTIFACTS A total of 244 artifacts was recovered. Potsherds were found on the surface and in the test pits, at depths of 0.1 to 0.6 foot. The pot- tery sample totals 177 body sherds (pl. 23) and 1 rim sherd. The ware is tempered with rather finely pulverized shell. ‘The exterior surface texture is smooth, the interior slick or imperfectly polished, and the core consists of a dark gray to black paste, compact, and often Jam- inated. The exterior surface, ranging from light brown to buff in color, is plain, with seven sherds having small punctate impressions made with a sharp pointed instrument held at an acute angle to the pot. Body sherd thickness ranges from 4 to 7 mm. and averages 5.8 mm. Hardness ranges from 4 to 5 in Mohs’ scale. The one rim sherd (14P013-86) (pl. 23), a surface find, is straight. Shallow diagonal notches, about 2 mm. in width, encircle the top of the lip. The rim is 7 mm. thick and 3.6 mm. high, and once had either a lug or strap handle riveted to it. The sight amount of curvature of the rim sug- gests that it is from a wide-mouthed pot. One small strap handle (pl. 23) was found in test pit No. 4. All of the 65 stone artifacts found were of chipped stone. These include projectile points, scrapers, blades, and one chopper. Of the 6 projectile points found (pl. 23), all from the test pits, 2 are com- plete, and 4 consist of the shoulder and about half of the blade. The points, made from chert, are small, triangular, straight sided, and del- Riv. Bas. Sur. Par. No 10] TUTTLE CREEK DAM, KANS.—CUMMING 59 icately chipped. Five of them have a straight base and one a slightly concave base. Specimen 14P013-61 (pl. 23) resembles a point but is concave near the tip, and may be a small scraper or knife. The projectile points range in length from 19 mm. to an estimated 30 mm., in width from 10 mm. to 15 mm., and in thickness from 2 mm. to 2.5mm. Table 3 gives projectile point measurements. TaBLe 3.— Measurements of projectile points Cat. No. Length Width Thickness |Description! Mm. Min. Mm. CM ee Wee ERE ee Oe eee ee ee 19 10 2n5 NBa (Meese swedes See Se tee ek RE ie. ob Dee 2 Peo seee es SaaS 20 13 2.5 NBa it) ees eee ree nC See Ne an eee ee ee ee eee gee ee 12 2.5 NBa 1a wt ps GP EA BD Se Bie EE EE OEE Sp iene nes bok alae dene jt ie ta PE 15 2.0 NBa FS tas pM RID FS See IN QO E08 E Raane he She as elect! |g aie sees eee eee ae 2.0 NBa noes ake Bee ber. ae epee SCAT, ces ae eee ee: apne a ee ae eee eee 10 2.5 NBb 1 See Strong, 1935. The most abundant artifact found at the site is a chipped scraper of gray or banded chert. Of the 58 recovered, 21 came from the test pits. The scrapers are of the spall variety and as a rule show a minimum of modification. Most of the scrapers are a variety of end scraper, and for convenience of description they have been lumped into three groups. Group 1 end scrapers (pl. 24), 27 in number, are subrectangular in outline and triangular in cross section. They are characterized by having a dorsal ridge off center, with a steep bevel on one side, a base that is either flat or slightly concave, and a bit that is abrupt and retouched. Either one or both of the sides may be retouched. Group 2 scrapers (pl. 24), 11 in number, are shorter and wider and more irregularly shaped than group 1 scrapers, and are characterized by a flat upper surface, unmodified or with the dorsal ridge removed ; otherwise, they are similar to those of group 1. Two of the scrapers are small, and resemble the familiar “thumb-nail” scraper. Group 3 scrapers (pl. 24) are small, ovate, flat to slightly concave ventrally, and have a low dorsal ridge that is approximately medial. Two specimens are retouched along the end and one side, and one on the end and both sides. Five specimens are classed as miscellaneous end scrapers (pl. 24). No. 37 is an elongate, coarsely flaked, keeled scraper. The bottom surface is slightly concave, and the upper surface has a low bevel at the bit end. The maximum width is at the bit end. No. 34 is a subrectangular end scraper, flat on the ventral surface, and concave on the dorsal surface. The working edge, which is abruptly beveled, is on the end and one side, with the corner rounded. Specimen 11 is a jagged piece of chert, roughly triangular, with the apex rounded and retouched. Maximum thickness is at the bit end, and the bevel is steep. No. 56 is a flat scraper with a right-angled corner and 60 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 the remainder rounded. The ventral surface is flat, and the working edge on the dorsal surface is retouched on all but a small portion of one side. The scraper is made of chert. Twelve specimens, four of which are incomplete, have been classed as side scrapers (pl. 24). They consist of irregularly shaped chert flakes, one side of which is flat and unmodified and one edge retouched. Table 4 gives measurements of scrapers. TABLE 4.—Measurements of scrapers Number Length Width Thickness Mm Mm. Mm End sereDetsy GrOUp Ise al cw oe eee QTc sae ces a eM es Pee eee Range. 2. 50-22: ._ Fae 38 BS 2 Fs ee Ea | ees 29-84 19-44 6-2 JW 01 enn | Cie Cae eee rae. Ui po | ce been eet 26. 11-5 End scrapers Broup 2..- dE PE we ey A EE ENE 1B eee pes Sel Re ges eae pees = Range wos Sth age Dee tel el ee 25-63 18-46 4-15 ee Ss BS SYS Le Oe ee Pe eee ae eee 44. 5 B2el 9.2 End oS group 1 ROE ic MOE HR) 2 FI oO Sil. sas. Ss a ee eee Ranges 28. cee. 0h Pots SPB e. . oe Oe Se | eee ae 26-33 20-27 6-7 Mae Ne er ee gee ee enone et CER: Sess Setar ot 29. 24.3 6.3 End scrapers, miscellaneous: Cavin INo Fazer CLT GE) Lot el Sas Ue aD eat Eee ek 64 35 18 Cat NORS4E oo co te. SPNALS ees pan S| eae ges a 85 42 19 Cat. No. Be SSE ERS ak. as SE A 2d, Slee 65 72 31 CatN 0:56.28 3: feo oh igns Spe eee ee |e ee 37 30 12 Side aes (Connpiets. yet 8 fede IC A a a eee ao ioe Soe 41-62 18-35 7-11 Man aA EAE OR SE ASS ee PLT TS sree Sa 49.3 25.3 9.1 Two blades or knives were found. Specimen 5 (pl. 24), found on the surface, is the end of a chert blade. It is convex on both sides, with a medial ridge on one surface. The tip is slightly rounded. The piece measures 49 mm. in length, 31 mm. in width, and 10 mm. in thickness. No. 74 (pl. 24) is an elliptical-shaped flat blade or knife, found in test pit 8. The tip of one end is broken off, the other tip is thickened and retouched like an end scraper bit. The edges of the blade are re- touched on both surfaces. The blade is 105 mm. long, 48 mm. wide, and 7mm. thick. One oval-shaped chopper (pl. 24) coarsely flaked, and made of quartzite, was found on the surface. It measures 90 mm. in length, 61 mm. in width, and 22 mm. in thickness. Four small pieces of metal were found. One small piece of thin sheet copper came from the surface and one from test 3. Test 2 yielded a small, thin fragment of iron, and test 1 a copper jingle or cone (pl. 23) which measures 18 mm. in height, 13 mm. in width at the base, and 1 mm. in thickness. CONCLUSIONS The artifact sample obtained at 14PO013 is too small to permit the making of a positive cultural identification of the site, nevertheless the range of probabilities can be limited. The copper jingle and frag- ments of copper and iron found in the test pits indicate a protohistoric or early historic date, and we know that the site is well within the terri- Par. No toy «TUTTLE CREEK DAM, KANS.—CUMMING 61 tory of the historic Kansa, and do not know of other tribes in this im- mediate area at that time. Unfortunately, little is known archeologi- cally of the Kansa. Ethnological data indicate that there were many similarities between the cultures of the Kansa and of the Osage, and it is interesting to note that the shell-tempered pottery and triangular points found at 14PO013 fall within the range of that described for the Osage (Chapman, 1946, pp. 22-23). It seems quite probable that the Reany site is a protohistoric or early historic Kansa site. The occurrence of chert all along the ridge, and the abundance of chert chips covering the surface of the site indicate that stone-chipping was a major occupation at this spot. SITE 14RY10 Site 14RY10 is a village site located 2,500 feet south of the south- west end of the Tuttle Creek Dam, in an area of construction activity (pl. 20, 6). The site is in the SW14 SE, sec. 24 TIS R7E. It is situated on a low terrace in the bottom lands, several hundred feet west of the right bank of the Big Blue River. To the south a small creek, the bed of which is 30 feet below the terrace surface, limits the occupa- tional area. The relatively flat surface of the terrace has an eleva- tion of 1,040 feet, 20 feet above stream level. The whole area, covered with a clayey loam topsoil, has been under cultivation for almost 100 years. Surface indications consisted of a sparse scattering of occupational refuse. Dr. Linwood Hodgdon had previously made an intensive surface survey of the site and had marked the areas of major refuse concentrations. A north-south base line was established immediately east of one of these areas and a grid system, based on 10-foot squares, was laid out. The first test, excavation unit 1, was dug 150 feet east of the base line. Here a 30-foot-long, 4-foot-wide trench was dug to subsoil, with a deeper pit at one end to check for possible stratification. The topsoil covering in this section of the site was 0.6 foot in depth and underlain by a sterile yellow clay subsoil. The cultural materials all occurred within the first 0.8 foot from the surface. No subsurface features were found. Work next proceeded on excavation unit 2, a 70-foot-long, 4-foot- wide trench, dug along the west side of the base line. The excavation was extended by digging a 100-foot-long and 3-foot-wide east-west lateral at line N310, a 35-foot lateral at N330, and 15 feet of trench along a third lateral at line N360._ All trenches were dug to sterile sub- soil with intermittent pits dug deeper. No subsurface features were encountered and artifacts were sparse. A large concentration of wattle was found just west and north of the N310 lateral trench indicating a possible house location. Lack of funds prevented further 62 BUREAU OF AMERICAN ETHNOLOGY [BuLL. 169 work by the Missouri Basin Project unit, but volunteer units continued working and uncovered the remains of one dwelling. Observation of the artifacts obtained indicates that the site may be assigned to the Table Rock Focus of the Nebraska Aspect. SUMMARY During the field season of 1953, a Smithsonian Institution party ex- cavated 1 site and tested 3 others in the construction area of the Tuttle Creek Dam, northeastern Kansas. None of the sites may be positively identified with any specific cultural group but indications point to some tentative identifications. The Sweat Bee Mound (14P014) was ex- cavated and it is suggested that this is a Woodland burial mound. Seven burials (one intrusive) were found here in a stone-slab mound. The Spillway site (14P012) was briefly tested and few artifacts were recovered. Ceramically it is suggestive of the Glen Elder Focus. The lithic materials indicate that it may be a continuation of site 14P013. The Reany site (14PO13) was briefly tested and it is sug- gested, on the basis of the pottery and the metal artifacts, that this might be a protohistoric or early historic Kansa campsite. Site 14RY10 was tested briefly and such indications as could be determined here suggest that this may be a village site of the Table Rock Focus of the Nebraska Aspect. It is felt that the potentialities of sites 14PO12, 13, and 14 have been exploited to the limits of practicability under the circumstances. These sites, at best, could yield but little additional information. Site 14RY10, which, for the most part, has been investigated by volunteer groups under Linwood Hodgdon and under John L. Champe, merits further work. The urgent need in this area now in order to begin to clarify the picture of the prehistory of this area is for excavation of a number of additional sites here. These additional sites have been suggested by Solecki (Solecki, 1953 a) and are here heartily endorsed. LITERATURE CITED Brew, J. O., CHAIRMAN ; and OTHERS. 1947. Symposium on river valley archeology. Amer. Antiq., vol. 12, No. 4, pp. 209-225. CHAPMAN, Cart H. 1946. A preliminary survey of Missouri archeology. Part 1, Historic Indian tribes. Missouri Archeol., vol. 10, October. 1948. A preliminary survey of Missouri archeology. Part III, Middle Mississippi and Hopewellian Cultures. Missouri Archeol., vol. 10, April. CUMMING, ROBERT B., Jr. 1953. Appraisal of the archeological and paleontological resources of the Lower Platte Basin, Nebraska: supplement. Missouri Basin Project, River Basin Surveys, Smithsonian Institution. Lincoln, Nebr. (Mimeographed.) By Ne toy TUTTLE CREEK DAM, KANS.—CUMMING 63 FENNEMAN, NEVIN M. 1981. Physiography of western United States. Map, physical divisions of the United States. New York. Gi~MorE, O. H. 1932. Turtle mound, in Cass county, Nebraska. Nebraska Hist. Mag., vol. 13, No. 3, July-September. Hooton, Earnest A. 1922. Theskeletal remains (from the Turner group of earthworks, Hamilton County, Ohio). Pap. Peabody Mus. Amer. Arch. and Ethnol., Har- vard Univ., vol. 8, No. 3. 1947. Up from the ape. New York. HrpiicKa, A. 1947. Hrdlitka’s practical anthropometry. Edited by T. D. Stewart. ‘The Wistar Institute of Anat. and Biol. Philadelphia. KKANSAS AGRICULTURAL WXPERIMENT STATION AND KANSAS STATE PLANNING BOARD. 1937. Agricultural resources of Kansas. Kansas State College Bull., vol. 21, No. 10. Manhattan. KIEHL, Mary. 1953. The Glen Elder and White Rock sites in north central Kansas. Proc. Nebraska Acad. Sci., Inc., Sixty-third Annual Meeting, May. Kivett, Marvin F. 1947. Preliminary appraisal of the archeological and paleontological re- sources of Kanopolis Reservoir, Ellsworth county, Kansas. Mis- souri Valley Project, River Basin Surveys, Smithsonian Institution, March. (Mimeographed. ) KRrRoeser, A. L. 1947. Cultural and natural areas of native North America. Berkeley, Calif. Martin, R. 1928. Lehrbuch der Anthropologie. Jena. Morant, G. M. 1923. A first study of the Tibetan skull. Biometrika, vol. 14, Nos. 3 and 4. NEUMANN, Grore K. 1950. Racial differentiation in the American Indian. Ph.d. dissertation. Univ. Chicago. Microfilm. Roserts, FRANK H. H., Jr. 1952. River basin surveys: the first five years of the Inter-Agency Archeo- logical and Paleontological Salvage program. Smithsonian Inst. Ann. Rep. for 1951, pp. 351-383. SHIPPER, J. M. 1953. A Plainview Folsom fluted point from Marshall County, Kansas. Plains Archeol. Conf. News Letter, vol. 5, No. 4, February. SmirH, CaRLyte S. 1949. Archeological investigations in Ellsworth and Rice counties, Kansas. Amer. Antiq., vol. 14, No. 4, April. SOLECKI, RaupuH §S. 1953 a. Appraisal of the archeological and paleontological resources of the Tuttle Creek reservoir, Marshali, Pottawatomie, and Riley Coun- ties, Kansas. Missouri Basin Project [River Basin Surveys] Smithsonian Institution, January. (Mimeographed.) 1953 b. A Folsom fluted point from Marshall County, Kansas. Plains Archeol. Conf. News Letter, vol. 5, No. 4, February. 64 BUREAU OF AMERICAN ETHNOLOGY [BuL. 169 SPAULDING, ALBERT C. 1949. The Middle Woodland period in the central plains. Proce. Fifth Plains Conf. Archeol., Note Book No. 1, Lab. Anthrop., Univ. Ne- braska. STRONG, WILLIM DUNCAN. 1935. An introduction to Nebraska archeology. Smithsonian Mise. Coll., vol. 93, No. 10. UNITED StaTES DEPARTMENT OF AGRICULTURE. 1941. Climate and man. Yearbook of Agriculture. Washington. WEBB, WILLIAM §S. 1946. Indian Knoll, site OH2, Ohio County, Kentucky. Dept. Anthrop. and Archaeol., Univ. Kentucky, vol. 4, No. 3, pt. 1. Lexington. WEDEL, WALDO R. 1943. Archeological investigations in Platte and Clay Counties, Missouri. U.S. Nat. Mus. Bull. 183. 1946. The Kansa Indians. Trans. Kansas Acad. Sci., vol., 49, No. 1. 1947. Prehistory and the Missouri Valley development program: Sum- mary report of the Missouri River Basin archeological survey in 1946. Smithsonian Mise. Coll., vol. 107, No. 6, pp. 1-17. WILFoRD, Lioyp A. 1941. A tentative classification of the prehistoric cultures of Minnesota. Amer. Antiq., vol. 6, No. 3. APPENDIX SKELETAL REMAINS OF THE SWEAT BEE MOUND SITE (14P014) INTRODUCTION The skeletal remains to be described consist of the remains of six adult burials, all from Feature 1, in a slab-rock mound. Feature 1 con- sists of an unlined pit, containing the lower burials, Nos. 4-6 (pl. 16, a), dug into the base of the mound, and a circle of rock slabs placed on edge just above and around the pit, containing the three upper burials, Nos. 1-3 (pl. 15, }). Apparently the pit had been reopened from time to time and burials added. In the lower group of burials the manner in which the bones were mixed indicated that the third burial had been added at a later date than the other two and that the bones of the first two had been pushed, each to one side, to make room for it. Many of the bones were poorly preserved. The same procedure seems to have been followed with the upper burials. Many of the bones of these were crushed by overlying rocks. Because of the fragmentary nature of many of the bones, and their displacement caused by forcing a third burial into each of the two groups, it was possible to definitely associate the skull and body bones of one individual only, burial No. 1, cat. No. 49. With this exception, the burial numbers apply only to the crania. METHODS The crania were sexed without further supporting evidence except for specimen No. 49. Aging is based on endocranial closure and tooth- wear. Observations of the long bones supported the conclusion that the total series consisted of 6 individuals, 4 male and 2 female. Measure- ments are taken in accordance with Martin (1928) unless otherwise indicated. CRANIAL DATA Condition of specimens—Specimen No. 47 (burial 3) consists of large portions of the occipital, right and left parietal, and frontal bones. The petromastoid and the tympanic portion of the left tem- poral is also present. The articulations between the parietal, frontal, and temporal bones are tenuous. Cranium No. 49 (burial 1) lacks the sphenoid, a portion of the base, and the processes of the maxillae. The ascending portion of the right mandibular ramus is missing. 65 66 BUREAU OF AMERICAN ETHNOLOGY [Buu 169 Specimen No. 50 (burial 2) consists of the cranial vault minus most of the right side and face, and all of the base. A portion of the left zygomatic and part of the maxillae are present. Specimen No. 65 (burial 5) consists of the left parietal and most of the right parietal bone, the squamus portion of the right temporal, the left half of the frontal and most of the left half of the maxilla and mandible. Cranium No. 66 (burial 4) consists of the cranial vault with the temporal bones, but minus the sphenoid, face, and base. The left zygo- matic, the glabellar portion of the frontal, together with the upper part of the nasal bones, and a portion of the body and alveolar border of the maxillae are also present. The mandible is complete except for the right ascending ramus. A severe warping of the cranial bones prevented restoration of the skull. Specimen No. 73 (burial 6) consists of fragments of the left half of the mandible. Measurements and indices——Archeological considerations suggest that site 14P014 belongs to the Woodland Complex. There is a possi- bility that it represents an attenuated form of the mound building complex to the east. In order to see what inferences may be gained from the physical types involved, a skeletal series from the Kansas City vault mounds, Hopewellian Woodland, located about 129 miles east of the Sweat Bee Mound site, has been chosen for comparative purposes. Stewart (in Wedel, 1943, pp. 245-265) has demonstrated the similarity of the Kansas City Mound series to that of an Ohio Hopewell series from the Turner site. The measurements and indices of the 14PO014 erania and a compari- son of these with the series from the Kansas City vault mounds (Stewart, in Wedel, 1943, pp. 250-254, 268) are given in table 5. Only 2 skulls from 14PO014, 1 male, No. 50 (pl. 16, b), and 1 female, No. 49 (pl. 17), are complete enough to yield comparative metrical data. Neither these nor any of the others, so far as could be observed, were artificially deformed. The skulls are dolichocranic, with the cranial indices falling close to those given for the Kansas City series. Head height is noticeably lower in the 14PO014 group than it is in the Kansas City series. The basion-bregma height in both crania No. 50 and No. 49 was estimated and hence it and the indices derived there- from are only approximate. The male skull is chamaekran or low ac- cording to the length-height index and the auricular height index, and low as judged by the mean height index. It is akrokran according to the breadth-height index. The female skull is in the medium height range of the various height indices. Head height of both the male and female crania of the Kansas City series is high. BY Reyo))«0«XTUTTLE CREEK DAM, KANS.—CUMMING 67 General head size, as indicated by the cranial module, is medium in the male and comparable to the Kansas City Hopewell. None of the 14PO014 crania is thick walled. The unusually small minimum frontal diameter of the Kansas City skulls is not present; 92 mm. as against 96 mm. for the 14PO14 skulls. Observations of facial structure had to be made, for the most part, from one female skull. The upper face height falls within the range of the Kansas City female series, but the face width is considerably less, resulting in a leptene face rather than the mesene one of the Kansas City series, Orbital shape is similar to that of the Kansas City series, hypsi- conch, but a marked difference is indicated by the nasal index which is hyperchamaerrhine in the 14PO014 skull and messorrhine in the Kanas City series. The upper alveolar arch is greater in length than width (maxillo- alveolar index 93.22) in female skull No. 49, while in the Kansas City series the reverse is true (maxillo-alveolar index 117.5). From this comparison of the two series it is apparent that although there are similarities, there are also distinct differences. Morphological observations—Observations of the individual crania are given in table 6. The following summary indicates the general characteristics of the 14PO14 series. Skull vault and base: Ovoid head form Divided type brow ridges of small to medium size Small to medium-sized glabella Low frontal height and small frontal breadth Medium frontal slope Small postorbital constriction Small frontal bosses Absent to small median crest Slight sagittal elevation Lack of postcoronal depression Medium parietal bosses Flat temporal region Medium to large mastoids Medium to large supramastoid crest Medium to pronounced occipital curve Slight to medium amount of lambdoid flattening Lack of occipital torus Small inion Lack of platybasia Shallow to deep glenoid fossa depth Small postglenoid processes Thin tympanic plate Elliptical-shaped auditory meatuses 68 BUREAU OF AMERICAN ETHNOLOGY (BuLu. 169 Facial skeleton : Square-shaped orbits of slight inclination Absent to shallow suborbital fossae Small to large-sized malars with slight lateral projection Slight nasion depression Medium nasal root height and breadth Medium to sharp nasal sills Slight amount of alveolar prognathism Medium palate vault height Medium to large mandible Bilateral chin form with slight chin projection Medium-sized genial spines Lack of mandibular torus Neutral to slightly everted gonians Dentition—The general condition of the teeth is good even though tooth wear varies from pronounced to very pronounced. ‘The crowns of the teeth are worn off markedly and in some cases the roots are exposed. Of a total of 74 teeth, some in place and others not, 2 cases of apical caries occur. Abscesses and other pathological conditions are not apparent, and the teeth were retained into late middle age and old age in spite of excessive wear. Pronounced tooth wear with a re- tention of teeth late in life, and a minimal number of abscesses and caries is characteristic of skeletal material from the Turner Mounds (Ohio Hopewellian) and from the Kansas City Mound series (Hope- wellian) (Stewart, in Wedel, 1943, p. 249). Pathology and anomalies.—No pathological conditions in the crania are apparent. No ear exostoses are present. The tympanic plate is perforated in specimens No. 66 and No. 50, and unperforated in two unarticulated left temporals. PHYSICAL TYPE The identification of the physical type of as small and fragmentary a series as that of the 14PO14 crania must necessarily be considered tentative and suggestive. With little supporting metrical data avail- able, reliance is basically on morphological observations. The 14PO14 crania appear to fit morphologically with the eastern Early Woodland long-headed group, characterized by the Otamid variety of Neumann (Neumann, 1950), a group found from Ne- braska to the east coast, and represented by such peoples as those of the Woodland Black Sand Focus in Ilinois, the Coastal Focus of the At- lantic coast, and the Karankawa of the Texas coast (Neumann 1950, p. 16). The 14PO14 crania are a long-headed group. The cranial index of skull No. 50 is 70.47. Cranium No. 66, although not complete enough for accurate measurement, is obviously long headed. The mean cranial index of the Otamid male series is 70.66 (Neumann, 1950, p. 20). Riv. Bas. Sur. 8. Pap. No. 10] TUTTLE CREEK DAM, KANS.—-CUMMING 69 It was possible to compare the morphological traits of the 14PO014 series with 39 of the 46 morphological traits listed by Neumann for the Karankawa series of the Otamid variety (Neumann, 1950, p. 21). Observations of the styloid processes, nasal height and breadth, nasal profile, anterior nasal spine, total prognathism, and palate shape could not be taken on the 14P014 series. The two series differed noticeably in only four traits. Glabella and brow ridge size are con- siderably smaller in the 14PO14 series, the iateral projection of the zygomatics is less, and nasion depression is less in the 14PO14 series than in the Otamid series. The tendency to a small to moderate de- velopment of glabella and of the brow ridges is characteristic of Ne- braska Woodland. In aseries of 16 male Nebraska Woodland crania, the modal distribution of glabella size is small (10), medium (3), and large (2), and of brow ridge size small (6), medium (5), large (3), and trace (1) (Neumann, unpublished data). The 14PO14 crania show a greater morphological resemblance to the Nebraska Woodland series than to the Karankawa Otamids. POSTCRANIAL DATA The poor state of preservation of the skeletal material prevented the recovery of many of the body bones. Most of the long bones recovered lack one or more of the articular ends. Burial 1, cat. No. 49, is the only one in which the paired bones and the crania were definitely associated. No pathological conditions were apparent in any of the skeletal bones. Bones measured or observed consist of the following: 9 femora (5 male, 4 female), 6 tibiae (3 male, 3 female), 7 humeri (3 male, 4 fe- male), 4 radii (2 male, 1 female, 1 sex ?),3 ulnae (2 male, 1 female), and 4 fibulae (1 male, 1 female, 2 sex ?). Measurements and observa- tions of the individual skeletal bones are given in table 7. Of the femora, two males show a pronounced development of the pilaster, as indicated by the middle shaft index. Pilaster development in the 14P014 female is greater than that given for the Kansas City females series, and close to that of the Turner Mound series. The sub- trochanteric region is less flat than in the Kansas City series. Of nine femora, the third trochanter was pronounced in one male, submedium in another, and medium in both the left and right femora of one fe- male. In the Turner Mound series the third trochanter occurred in some form in all of the males and 4 of 6 females (Hooton, 1922, p. 128). In the 14PO014 series the shape of the shafts of the tibiae is predomi- nantly the ordinary prismatic. Of 6 humeri, 2 female pairs had large sepatal apertures, 1 male a medium-sized aperture, and 1 male no sepatal apertures. 70 BUREAU OF AMERICAN ETHNOLOGY [BULL 169 Estimates of stature are dubious since they are based on individual long bones and very few of these. The estimates are based on Pear- son’s formulae (Hooton, 1947, pp. 728-729). Among the male burials there appeared to be one unusually large individual represented. Femur No. 97, with a maximum length of 515 mm., was considerably larger than the rest of the femora. The stature estimate based on this femur is 180.5 cm. (5 feet 11 inches). An unusually long left radius, 290 mm. maximum length, which probably belongs with femur No. 97, gives a stature estimate of 180.7 cm. (5 feet 11 inches). The other male long bones, if measurable, would yield shorter stature estimates. The average stature for the Kansas City male series is 160 cm. (Stewart, in Wedel, 1943, p. 259). The stature of female No. 49, based on the left humerus, is 153.5 em. Using Manourier’s tables (Hrdlicka, 1947, p. 182) the stature is 155.2 cm. (61.1 inches). The stature range for the Kansas City vault mound female series is 156 to 157 em. (Stewart, in Wedel, 1943, p. 259), and for the Turner female series 157 cm. (Hooton, 1922, p. 126). SUMMARY AND CONCLUSIONS The comparison of the Sweat Bee slab-rock mound crania with that of the Kansas City vault mounds has shown that although there are similarities, there are also distinct differences between the crania of the two groups. Excessive tooth wear accompanied by sound teeth relatively free of caries and retained late in life is common to the 14PO014, Kansas City vault mound, and Turner Mound groups. On the basis of the available data, it is suggested that the 14PO14 crania fit in morphologically with the early Woodland Otamids. The Otamid variety has been found, among other places, in the stone vault graves that occur along the Mississippi, Missouri, and Kansas Rivers, and is encountered in the Hopewell mounds of the Illinois valley (Neumann, 1950, p. 121). Riv. Bas. Sur. Pap. No 10] TUTTLE CREEK DAM, KANS.—CUMMING 71 TasBLe 5.—Cranial measurements and indices of 14PO14 series with comparative measurements of Kansas City Mound series (Hopewellian). Cranial measurements and indices of Comparative measure- indicated burial No., cat. No., and sex | ments of Kansas City Mound series ! B1, 49, | B2, 50, | Bs, 65, | B4, 66, female male male male Male Female Mm. Mm. Mm. Mm. Mm. Mm. Moaxmumsieneths +2 2225-2 522 173 AODy |e oS ake ER ee 189 (5) 178 ry Meximumybreadth.. 2-22 .=—- 222 128 136) [22 ey | ee 137 (5) 134 (3) Basion-bregma hgt_..-_--------------- *125 130-1» [Bae ae see BS ee 142 (2) 139 (3) Aaricularihetss cles secant oe CTS 111 1064-2220. 110+ 118 (6) 115 (6) Thiekness lt. parietal. .----_...-.-=--- 4.0 4.0 5.0 OV OTE es Ss ea Le Minimum frontal diameter 2_--__-__ (oes 89 OGY eee eee *97 92 (9) 85 (6) Ophryonice horizontal cireum---_-__-_--_-- 483 BIOW | ase nees. 2 | ee See een | eee Ee ae COM EAM CONG cae nee et ie ae. 105 HG See een ees 114 (10) 109 (8) Sagittal (nasion-opisthion) arc_------- 352 By | pee a el (LY eR ae eee (a ee Frontal (nasion-bregma) arc_----_---- 114 1 7 [A ee Se ee | ee eae ee Parietal (bregma-lambda) are_____---- 122 120) es. 23 T20 iE see ee ee Occipital @ambda-opisthion) arc__-_-- 116 1 Pl ee ee 1 OS (SEE Se ee See Transverseare $2222. Sas tt eee DEG) |e nae Se arse eee See ee ee RS ee eet RR ee Gnathion=nasionvhe tose oe ae Ee ee le eS ee | a SAE ee |e SE Upper alveolar pt.-nasion hgt_-__--_-- 72 (2) 70 (2) Bizygomatic maximum 140 (2) 129 (2) (Ore) yea) OVS) el ae Ty es eee eee tel Ak es Yl (eee Vege eee (are sete sages | beep peee = | ne eos ae ‘Orpitalheinnt ltz= 2) ss eee 34 M (8)} 35 M (1) Oxnbitalbresathifromidarcryon, Theses, | Vt) peOi hes oe es ae ee ee ook a ee oe Orbital breadth from dacryon, It_____- 38 M (3)} 38 M (2) INSSaUinel oh peo se eee Be ae Oe 49 (3) 50 (3) INE Co Ie Gh ee ee eee 26 (3) 25 (3) Maxillo-alveolar length. _.____________ 53 (1) 55 (2) Niaailo-alvoulan Wid thas a. ete el eee pb || ee eet SoS 9 a ee Sh ee eee 64 (2) HeivhtionsympbySsis: Gnandible)e- s2=- |) fe cateas eet Ne ee ES tet: WE. nn 2 |e ee = Motalanandibmiarienesth esse ee ae SLOG lace sonnei re LG) JR a e IH Sig tion pO Vas. oo eee wee elan ~~ 6 e2B iE ee eal ee ee SO) ee a ei ad Se PR BICRNIESSTO RDO OY) Ona eee eel a a TA ee eB Ae J) TAC RE ee Lee Ber an Wi Cu eek ee ete meet cee ee ON [a ks Ss Oe he ee Se ee Se IE oe Se ee GISH Olas bees ae ee ets] ek | BG) eam ees SNS Rese Pa BS IE 2 Rea ee SO Max, width, ascending ramus, It.7___- AQUA eS 2 oe 49 pO | RT ee a eS ee Min. width, ascending ramus, It______ Si eta 39 5p] Ee eee Eee ee Mandibular angle_...___-_-.._---_---- VR ike | a VV20 At. gia) | See oe SS (ete Cranialiindex. 2-0) Sea ey eS 73. 99 TOS4 7) eee. | eee ee 72.4 (5) 75.4 (3) Height-length index____-______________ OA Hl VME ee aa (SR (Ep A | ee EN |S RS oo Height-breadth index_________________ aL FAR TH bc we CUTIES! 3 | Oe Soot SOR (nt Sipe | eS ERD eee Ba ae Mean-height index__._______________-- S3sO5) i sOs08-1o. |b. eee (EE ee eee 88.0 (2) 90.1 (3) Auricular height-length index_________ 62. 14 bit ( De] ee oe aha [ie EN AD ba | BRE BS Cs MD | ae Cranialimodule: 22 ee 154 (2) 147 (2) inrontebindex tse wet Ve ee 77.6 (6) Fronto-parietal index_-_______________- ES Ul »| | Bee ee BE A OR Oe et 28) | ee eee ee Cranio-facial index_.___..___-_.___-___ = Zycorontalindex.— =e ae ea (ee Tee | ee SS | ee S| | eee es Mronto-ronial index. 2! Se Ses Le a TOD) Oh: ee Se ee a |e ee eS oie ee Se AyVeo-ponial. index 2.22.2. OS | et 1 Be Se See tee eae hd eee a TOE Vie (6 (Spe ea SRE a cae Sa SN ce eden aah RS ie AE eee ale 8 | OE eee | (2 eee Ee Uipperracialindexe 2 ee 52.1 (1) 53.9 (1) Orbital/index, mean_/__~-__-_---__2-- 89.6 (3) 89.7 (1) GASH st exes | ep i ane ee py : 53.5 (3) 50.3 (3) Mazxillo-alveolar index______._________ i (enema a 2) sa LAS eS OR) | eee 117.5 (2) *Estimated. +Approximated. 1 Stewart, in Wedel, 1943, p. 268. Figures in parentheses equal number of individuals represented. 2 Measured at 3 points 1 cm. above the temporaparietal suture and averaged. 3 Measurement taken in accordance with Morant (1923), measurement Q !. 4 No allowance made for tooth wear. 5 Single measurement taken with goniometer. 6 Measurement taken in accordance with Hrdlitka (1947). A eaeereren taken across top of condyles. = Mean. [BULL. 169 BUREAU OF AMERICAN ETHNOLOGY elem} ‘smpoh g¢-98 oq “TPUBUL 54 “FT BIT -XBUI “JA SBI ‘Og E2 anna neon nn =e- OU SF ey ase ee OD a ae OD eae ea ae OD naa oe BF ae 2s ODE ee eo bs ce a ee OT aeS Seno OfSAUIOD) |Fs> =~ +2 =se}OTAMO GH) 5 = = 07 0[GM0G 4|5=s5 > SORA MOO, |= = SOJOTAMIO)) (Peet SS So Ge uel09 MOTSNN00 [eULIEIXy a2 --------25- OS Chee Op Ss ae ae eon Ss ee pees | ea | eee Se oe DOpdaeT a--------- UM Paes Gas See MOD so|) sea S & SOP Gal SS SO eee Sale sa SRO ee ss 3 So bes se ae ese ScEsarS DIN POM son. eatUMTDeUIGNs) |5 SH so=a UN TPOULQ MR sas coe tine ee oe tals oe ee TET DOIUQLIN) Ig tae sg ae een ae uso ye | TR MOUO() mOTyBAIEg moy1eyd ‘seuog ueruTI0M ‘semnyng aks Oe ESE Sos ee Sima: "2S Seca ees BA ee se IS: | ee ee ee Oey ante n anne nn nnn nnn | on ann nnn nn na | =~ oo nn nn nnn =| $= nnn en a= | = en nn nnn no-one nnn nnn ne enn nena n= --- sno} Jo sdeyg BSS NUOSO We le ae: sane ae eODs er aeie say ae te ID, ae Pe IUCR. lige as UO Vliet Cho eee ee SSorEoaSOs UTA Ales | eee eka seme OSC Van eons ep ae tenn econ eae oman N UG |[ 0g epamen on TUL POI: | sutra teeta ra es OU} UO) 1OT) DlOpdurany ------- DeoUNoOUOTy |= a.2-- Sampo, |p -o-= see -=<---= | -—-- non unlouody |--—- --- pepuno non, yest] pee eee. - CAs WHCIMO sMOTSOI [BITd1990 wa n--2 2-72-27 Ope ee | ee eR ee gee ee Ree eS ee ee esOIdep: promod Pe Ee Osan | testes eens O Desa once ee ge ene aa) reas oa 9 eee SOND) PLOISHUIMICNG parse sia [EUG gl ares ae eee OAM me garea = See cece ee Ff aa oe oe SSdC Te lna rank ears ce OONUM toa agi is ee ae eee a> be TSDIOSSR IL SES ee AGT ae peice ae he UOT ig| ae ee UOT My pee 8 Oe eee a Se See BT eee pe ee ee SNOUT :WoTde1 [e10dula,y, Se eee QUO Niels par ee nee QUO Nig |o on een > ssa OUONG aise oes SaITMLUO Se paer sco ee SOUONM (Cmte ele ney oes eee pee eee Ue OU Te ALY Pa yALel nee sees na MEELIOO IA |pe mens an anes sOd0G7)) lier trey MUXELT DOT inne 7 pam CCL OOINTS tomer jo) 0 Wiccan ct cm SOSCOGT ~=-==------- AUOSC Va | samanena UNC Wi ilizan sae >= 0s }UOSC Vaalnunae coun as OD =actal n= aes oe PUUOSC IY «liemar a= ngs 1 uc mat cee OL SSAC Op) TUNOIO0ISO CT, ee aoa oe AUS So | see eea SS ulea: pom ase er SNS alana ce we ee COSY al enya Ol Soloman ea bo 0) = aencier a mane OF UAGTO CIN oo8) 1MOTdel [BOlIV SareeeSereccee GIO Niall anes scce sae ae cs coast" a= Deets a[e see as ase eae | cine a 5 aeRO NIG |e eae d sm a mE ONO TAT ------------- (ODmsessn | Gas see sre OO Nic aera ei= ¢ GUOND |b eae oe tke Ls eee ee aa | OD ee a Sat ees Taine eee eas = I SOLONOMIDATAT or ae hee ODpes 5 GSee ee ee aU pre os = STIMU | oe oe eee eee OD ers| ce le er ape Se SOSSO 2 ee [lems | een enna nee mannan | oo -=-qpeug [=o =O JOFAYSMOO [eITqI0Is0T a eee MULL DO Pe | sees sss SLED OIA Ses os as oe ese al ae as ae IDOI ee Se See eee ee OOS See TetShlbee bath = eee a sees toe Sots sa eee ee eal ene SSeS LUN ts eo ee oe ee a oe ee Dood meneajes---a55 oD.) elf aaa gee eel Naa gia a | ae ae cdi ilu abe orks geet decal’ (00 0 ieag ee an gs deena es Banca ea eu pera pet) 8) <1 ne-2=45=52-=- ODamae tare eoeced TUL Gs lige sos SMa abe eae oa oe sores gem OD eg gle eg eee oe ee ceed BID) ~------------ TIBUISP nite cee OORT E= sss 2 SS LUNN DOIN |= gen nn BUG) | cease sera: pnIICULS «| rims Caters cerege same Os IO, SOA DIT MOrG. =-=-------- TOP AT Cie fae cee ae LO DPA Clit me =e wae es DO OY AT Ge | fe nee aad | cee OP TAT) §| Sein Soe oa = eon mene OCLA QU NOS DICE MOTE SMOISOI [BJ UO1T = 222 ---------- OUO Nisan caeene CUO NM vrabs a eas: wOllO Ninn eas geet on CUO Ns | Pace pmyrecge ODO NIM pe gugrae pct pone been ont eerie pee LOT SUC LOS OG Se ie ae GSO oe fee IONOM Gea ea Bee DIOS ON le: > eee DIGS Onl ieeiene s MODIOAQ rae (0. ee Be yt ocean ee CUO ------------- AMATI | sear apn US ola eres ae SAS alice eke, pene USO pas eee ema aUelle| tar guan bea ele ehieg ees co Pg aes eee aa SULTON aa-=--2--- cmamipopy, |--------- -uTpey © SS UO: ara PDOOUNOUDIT ye ae on eyeurey ‘sived 91-99 ‘BluBATeO e[qipueul “Tq 6F (4) eyeut ‘srv04 go-9¢ ‘By IGATBO qqipuemt “Fg 99 eeu ‘s1voé eyeur ‘s1Bek gg-9g ‘Bajeo ‘Gq GO | 92-99 ‘BATuO ‘Ed Ly eyeur ‘siBoh 91-9¢ “BATBO ‘TE OF xos ‘a8e ‘muommtoads ‘ON “480 Po}BOTPUT JO SUOT}JBAIJOSGO [BoTZO[oyd1o Jy 72 SU0NDALasgO 1DItBO,OYMLOUL 3891428 FTI OdVI— 9 ZIAV], “OPBUL 9q JOU P[Nod suoTeAJosqO ft 73 Baad betel ald Cb eam! ie ace EL DOT Aral vee at es eer [Pe wee te Ns eh a epee cee so aoe aoe en SONOS [SENT ~----------- ques q:y—|==7—= o> -gmesq yj 55g esq y.- [29S tio sa seer acon feat eras soc sasannans|snaaesannensssc na qssssessess2==599 40019 [BSBUqNE -2----------- ICU Gr Seneroree aU DONA pas a ney OUBU Ry one sence real aay Sea | Gea Mae re CE oe Raa Ee EE EES TESST Bn a ee re a yr nn | ee a | nee Seen | oe XOAUO0-OABOUO() sos uuuunses uu Tia Gaus baa vane crac CU OLCITGSGNT ---------- EER GOLA AIF || Sa ESS ane Se ena sz Sie To pa ears haa asl pee SEES Sei ds Sd a aa eee DOdaG, ‘ospliq Test N -~------------- DOW a Peano aa tee, see aes nuw as = Caeser alts je see ee = he oo STOVE alia Sage ca eae ee eee eae oo TOTO DAUM GS ONT po Salas em AOTC NG, (eee aeGa ODEO Ber Selo RODE RS Rae ea SS SS ee STE SOP aa T EES as ae ae ae eS TRS Se SE PBA OOM TESE NE Bee ser = =a eae ANOTT arses 7 GRU DOIN, |oae = ee CMNEDOIAL fF a Penn fe pecs ae | Se oe UM POIAT [o= > 27 a= 2 SOU g Od TBSSNT w------------ AUSIG) |e aed Oona TUONG: | eee ee UCSC IY, jo - eeete comeres oo) spo ieee mR UNG) [por oro esO Top UOTSEAT o TMOTSAI [BSN Zi =e ee (any fial 2a eee ULSI al haces a aes See Sa | Gc ee | bee ee eS | eas ssoo01d o1}eul0sAZ eee eae OTC S c(i Ca ere en) a - < eOUOIN Gane cia ce ne enn Ig eg os oe VO a eee ev oe ee er eee Oo a RRO [e}rqtoqns s Tarrant oan aia em Se oo ae im mnie = al (Galetti ie SS | | PC eee eee ssod01d [BUISIe}AL a ee Baler CHa) (Stal ee Ot i ei ia cletinin || Shintaro) tal (iio eee beer 1a uorsefoid 10;10j0y Silene tersrersc, MALO lis caaeeeeaae ce RUS Gs| Ceacesa suns pee MaMa aap ce ees aoe yeaee a eses- sss r es oe HSC] USD Cegetom se chiba~ Oe Diba ape nny oo ae a wor}00ford [es0} BT se ee a. Gry cece TRUS c|seae a> mens ass GLAS acts ay ween QWOdICiha os 22s sas oan sos aeen| ayes see BOOT OL Nel tence erty eae ee y gE oy Sav[BUl JO 92TS SMOISII AVIV IAL [Rue chs fo RS Ar ifs ieee TC eeepc DGC Mt Reepeeeeerper me pairs az OECD taekbiostaescg ze beer see ToRwUTOUT [eITqIO oe) ae GIBIDG) |=s-- == sscae SIGROG)|c-ssacs4° oa OUGIG GU isacg—aa- a> s2s=eo cee gsa=e sesso DICT Ss pars Sopris oe ree eee odeys [B}1q10 S TUOTseI [VIIqIO hasan: SUISSIUL “41 id waannea-e Gaya (ge tla le AUGSON alli ses ss se sataans soos | Gamers ooo OUOU dic ||segeeaes quUCsold eg ra\| {== toes Speer a= Sao oyed omedursy pozs.10;10q ------------------=-|------------------=-|--------------------|----------~------~~-=-|--------"--- yuosqy |-~7 >> 77777" >= “47 —cemre10; snoutds-03410}g Co ad a ee os 41 —UouIe10} SnouTds-034.10} 7 Ss een nti (allt anoint le i |e (ee 9ye[d prlos£104d [vuIEJUy SI DGS | SSR SR SRSA CEST ee iain Wea | et are ecient cence peri \ eee UM POW |s- ses aes ses ese e eyed prosX10}d [eu10} xq B rseeeeneeees ABGSOUVO | Saoe yn cmnnn-Sucesen|nn ssc aaen se g—getas | kesagsneascn een senor lea qaede pee eee a ae= aa nee eo ches eee oe eee essoj [eoduAregd SSSoe ees eh 4 Si ieee a aa aE for Tol ais gsisysseuac oye t [GR en Mes a Op pn [Saree n= 7 i babar raeeea ania |e ol Ieee ieee «eS ele ag ieee eee ee | eae eee eee ee eeeeer se BULUIVIOJ 01B.1008'T lS ne A BPR ESSER Res E ae hae) Rmaieaie shies Fe eRe Sat ie 1-57 ol bee Se ee ee ee poe cee eee ee bet ase e owe Spro[Ayg Bl} Oke peepee |eecese-eecee aWighay |= ss=->==s-— SUG NU aamnaaesemecaniaebssan oa sass nee SHOW. A[ =| passaaeaaneaae OUOIN(=| Re s-ssze5c22- GeGyiigy [e==203- 38 gape ee |e asesaeaes eBay TGl-|F=55==2==555 Bedi aMle-yreceen tices cee oe eee snyeour £10};pnVy =» Saar (| F=ssseosasse WiC Wb OLY, lp sssa0n-ecewee aca |Senesacerer-a= Gti sicarcsona= eae 12140 (pa nn = a a ieee da eje[d ormed urd Tf, wana n-------- IBCs eaten Gites SNCULGs |p eanne eae TCAs beer secre ea SUIN tb re) a caer a] TOL Gy saan eatn eee sere ns == — = BONO POMS na O el eee eiierats i> -—=—-— 8 IMM POI |\e= 222-2 ae AOS Gd) |i = caesarean (O20 (Ci epee PSL 00 6 02) | a ae (OCG ll socaaics aed Ace i eR yjdep essoy prous[yp S| | = aga Si: s |Reeeseeacetes Qj - =| [pRSseReSSeoes SiO tii |aucelsnakmamteeitran |nasmonccanns sees snes |peeeccesscace (0) oat tei (page td i eaten oes BSSO} [BITdTN90 UBIpIPAT Ba J eee esease=-=-- QLLOING ll Snowaneae oe 654 MOS | oesan Wy ape oem eed | Lave ass em aa o> oll as Sse pee Se OL ONG a eaieraeeccc: pre et ae ell eseqAjeld AB — — Jannnn------------"--|--------------------|--------------------|--------------------|---------------_-5%_|_----------------------------- tolssdidep sno1j8g Pann nnn nn nn nnn n= -----|--------------------|------------------~-|--------------------]--------------------|------------------------------ WOISVG JO UOTBAL Bo 2 Inn nnn nnnn--------+-|--------------------|--------------------]--------------------|-------+------------|--------------------------- So[ApUod JO TOVAS[ !MOISaI IBEIsed IY a nn) ke al eects peat a ee ei ag GkOINe Day sense ssea SS cere ses srer cs ee OT en en eee (MOM), Fae seSaeseSso-ses Sess a| sacs sie aes es cceesss|sssessasresstss=shss|esscecsabsreschirceskhersenssnaensrers = Wsty ez 2 :MLIO} WOTI8}. PS stg atte zeeessee- OP~~>7 7 | -n2 arr nnn | once enna | noon nnn e-|------------- Opresss|pssaaser tenes aessnree 2ctee 1e91d1900-o10d u1aq, cS i a eeeeeeeencn Obese soeen coos tas ODmesed|paeaases EasacOUO | Gassmeatannae OP====5|F22=22=575237 Nt) Use aes oo ene Shas itg co - Paw es 18941389 Be Bt} f2es--22-==-2: OPaperc | cassseacssces ODsGas +|saasccaesaaraseets—|coaseaseco=—= Oprszss|Prsee55=sseo5 | eian ||ntn’ pene doar San ee [euo.10 +) (yey ca eee UO Na casein ies OELO NT Ge | Scicecinte ats eee been as eg eee OO NT eas OTE Ne ecceiediatasiacte moose ™aiciaar sakasialieniicin propque’T . ah :SomO0q UBTULIO MA on so non Sas Ce JFT2) (0 80 (OY Yin ihn a a I I a al ed te ea (Y) OF at. «|| i Se, Mig’ toe he 2k ene ee PI0}SBUL-OFJ OTIC T Prat OT ane oa EVI tae aD we Men OP] | ae pear cee ge | Reese Mae RRN e ent =| Sieemmeer coe OD een Pewee 2 so ee LG OlkEO-Ollodan Lene LA OD ata elie nc, NOUR ERE ete ec ao mek Noreen on ODES ears te ODEs alae or ia eran er eam OD CSTTOOT, 422141—58——-7 [BULL. 169 BUREAU OF AMERICAN ETHNOLOGY 74 *DOOK):{| 52s ase eee POOP) wits sas 2S E5=sRaF POO) \|F=s==5= Sie DOO Hye 2 SSeS a see aaa SSa Ase s zood AeA |-~~~--7--- iia Jes, canons Ent sa as ~-Ayyenb [erste ‘QUO Ni-|ane waaay AUSTIG» sana ae eae OULOIN -|'32255 27 hs Sean QUO Nin | paeerer mae 4003 30g HEV UES) || are PSO UNO COR dim |S ase staan 25 | eee ae mana ser Shag ||P > eames eee See ere liege 9 ear eee 5 | aera NSIS Rare ener “sios}our pedeys-[aaoyg DRE ca De a | De Be eid Les ec asl |e ee il (ie ee ET OS Oe oe ov "+ moMIpayy |--- ~~ ~~ peounouolg |------- peounouolg |------~ POOMMOWOIG | |se=a> seen eeeeemaet eels DOOUNOTONG:ATO)\\— | S528 3SSs 38 9995 tas canes sana S Rae Piet Ex pee inane | ts el acini soc ee | bon ei e en GRRE ores preter et See Coe sdsno 1e[O Wy SOUOIN GN | =s=s2 pecaes pan QUO Na seaaaewnenns OPRS=5 |isee=s5se2 cae ODS SaaS 2555 neg oe ee |e LS Dag ee ea sass aes sesseosqe JoquinN IOUON biases s— I Na ta taal aie eta ie i ee eo SUOTMOS JO UOISIOA Peek ae AUSSIE! | Sascha eA OS GIV) «| bia eee sar UO Worn ne Sera ae we mr eae Br | eos te ET ee STITT pots Siege D Dimerses neptee esas ULI DO Nie lanai er 2 sce aime a oe en ne ne et nS S| SORTS AE ae ES ee es So ce ee ees SOTTOS IT STOR) eo Rae ee RUBS eae amo as n§ USN Qulieece teen et ene eas |Peens sree as soso Seen) eee nee meee ene sane See A ea ass Seas es ets hz STO STON OO ONORETT) RAE R Fae UE DOIN | seeps ys ClO PG pee sme rm Teg sei ee eee fermen lee een es te k= mee ee ee Seep ra [Se > ene Sees a ein Sa ae em IOTSENT UT Oh SEIET TPO] (Aen | inane TLE CSIs | Gama INIT Oe | saan ee oF Cl aaa a aaa a Lr OZIS B[QIpUBIT ‘O[QIPUBYAT Ie Aiea SNe Se Ree se ana retaa mw gs re eae am Oe eT TST Se gee ea RR ee ee Ta See TR eee STL ES SERS RSET Oe sa| Seen SS SOR RSA Ren Sean aes 9UIds [eSBU 10T1e3SOg PTS ERE TOR | Snip as wae eae te lee cee Ghee eae a eeOS Ss Spates SLE R Cee A See e re te eee aaa es Ses Seema Ss esse aes area 9INJNS dsLIASUBL], Bern Cog eae a Le a ee EE a | ES GS ST fa a el cea ek a S| | ee a a UILO} SNOT, SSP Sates s AUGSG Veal apoe ela peat | een ro a ap a | ae eee ere ng Le re eo eee | EE OSS SEE RRO sonia sos Reno ae eT ONT? st} 02 Al baa areas ae LO) Oa ees manIpeyy | - of ORM TOW} |Pseac ae eseo es S52 555 | nasi teena es SoS aa trea Sae Sa asas ass see is eaeen SA Snssas JUSOY INVA O,v[eT Shear DAB UBH (i) oes aes Sea Ragen 9 oP SS TSAO Ses ae gabe | ST ase sates TTL RSE ss | sesaee tosses sree eers= | psnaaasSekennen cs reasesssass SOR ade me IG TOT, SUOIZO [V4 B[V Pie eeseee S20 Diese ih Sane en eveR Eee || bk eee Ss = oe me ES | Soe eee Rese = See Eel Smee eal ae ware pes ESE Aenea oe Saas ean = s=s=== TIS MIEN ELON SOULO Nia | iecenecaiermemeaen aa ODEs a pose eeee SUS s|csaanrscoaa, VUNG! | Fa e25e52 S2 eI aas Sa Te RRs Seo T ce sae aasss | se eesasssnssesanecssscesses Us| q}BVUso01d IB[OSATY PERSE SER ESS, QOS GRR ASaseaa sae cgenees |RESSS5 A 25“ SeSesmases |Pasaas cerca sscacsses|Sasaasaarannsa eran as |acenssadessasssSteess-=s=a ep BUsOdd TSOBIDUIAL TTISIQ}eUsOILg e[ euro} ‘sivok ¢¢-9¢ 91q ereurey ‘srvok (4) oyeum ‘s1ewoh eyeur ‘sieek eeu ‘sreo4 ereur ‘suvok -Tpueal 34 “FT LI 91-99 ‘elareATBo gg-9g “el TeAyeo Go-9g ‘BATeO “CE G9 | 91-99 “BATBO ‘Eg LF | 91-99 ‘BATed ‘Zq 0G -X®UI “41 SBI ‘OG EL e[qipuem ‘Te 6F | Weqipueur ‘F_ 99 xes ‘ose ‘WommTOEdS ‘ON “480 PE}YBOIPUT Jo SMOT}VAIOSGO [BOTSO[OUdIOJAL ponul,U0)—suoyvasasqo pooBojoydisow :8a1198 FT OdY¥[—'9 AIAVI, 75 TUTTLE CREEK DAM, KANS.—CUMMING Riv. BAS. SuR. Pap. No. 10] *poomnouolg ¢ {POC =) SS eat *puo *puo *pue *pue ‘roid pue yeyg | *xord pue yjeyg |-----~ gyeqg [77-77 yeqg |----- ASUS | = yyeqg | ‘xoid pus yyeyg | “xoid pus yeyg |~oyo[dmog |--------~---7->- nom =~ ueumedg “wory EO 6 (A ee ae el a eo | eal epee ee er eee Lt Tor QUSHON Sa qs |~~~~o1d *A | -10d soddn 04} Jo mors103 Jo saiseqy *peulqng. |-==>>---5 poeurqng |------ DOW | F=2= POINGa= aaa Pe |---peumqng |--------- POUIGN Gy Ease ss = \SEIWE jana ces Olg,:|a5 222525555 SSerer s e10dse Beul’y | DULG) y | manana eins pend |---~- pend |----~ Day) |e jDeLAKo) |RSS Mu) | ere |DeaKO) | ase [eta fe ha ON aa eee odeys 4eys e[pPryAL “Ose al epeense cic C0) Od Naas eae ett tere Site eer tl ee C0) 0 ite [reer Opa AlRenaaase7 pomqng |-----~ COE Gell T=5 SaaS Seen ass Joyueyo0s, PUL, SDOWilgeecetrsc = DOTATS | > as | a ae FE SSeS HST S| TSes OD 55) rare s ODRe Ha es AUDOSG VY |Peas Usa yy is—-- = ‘eoj10JUeyIOI1]0d Aq BSSOyy “auosdiy; |----s-> a3 RUGS Wal | Race te aia ae ta y| Sagem alse UGSC\Val |e saenacae AUTOSCVi || paaunn sme RUSS nlp GA Aig eee eS BolIoJUBYOOIJOU AY BASTIO -SUHOTIVAIISG O OO SL alinieser soe DRG Liilis SRes s HF ease RSS Se ep R es aloo ge oe hen > —al|es se 98:9 in| eee CORSA S| Sileetnaah ah |i) 2 5 leanne ODIOS i | amnmenn L/00LX9 BIseurA4e]d Jo xopuy Go| rear shen ee CGN | eR garer nr ess san peer haan ge eee ara one eee a Cork aes == See a1} Saeaeteneialenniat Pod = kei kainate Gg |" 9B “WBIp oFAGUByDOINWGN, ZL OG i sores asoa sos (<{ Agl lnieaeekae adel | aee Giel | ie eit ae aien io Cone Sa On| Pore oer cone gl etek koe 1:1 Cia bear d-8 “WIBIp d1IyUByIOI4GNE 9 -OP:Se alan ee ee DPSSR Ss 2+ aogier giniang ae Tear Go = [AL allie sage ae = sal Veale Mp palette | Bie “sie Mig “on CPZ | rte COMLEn | Sear meres 4/001 X¢ XOpUT O[PPIVAL 4563) |e Rae aR (10 as aos onal (Ceres a | ere 118 OES SA AL RS REA BREST es|| She Rae ee bY it sean acess NGS PT OR [@19}8[ “ULBIP JJBYS "PITAL ¢ SOGuleaate = te oe OCs Mae Boe |e as eee alg ape (Glia ee ES | TR eee ae rp as essa Pere eee (ti aeons ise Re eesepeeee d-e “UIBIp 3JVYs "PIV F as Maes Lien Of) ane ter Se all tos ue | [Paecer 512 2 Rilata re ee ER cone > ON ln ios ea ee cea =o |S ene || Backes eae s ks SoD MON RINE TDRXGIGD See re ea ars oes me alee es a eee be rel eee eg | oe eae eee arose pene ne SO ae ee an OS fo ements ecaammen? (LG) Tes bat 3 SSE STS | cea ah ig | | Rm Ty ARERR Sa | [ae a ala aa i lr a ne tee eo fr Co | seen Seen saeeULOUo xa rAnay ‘SOOIPUT pus (“UIM) sJUsMIOINSeOTAT “4.1 “i ‘12/6F “41 ‘a ‘%2/6r “MWe ‘29 | 9 ‘WT ‘99 | “9 ‘WN ‘Go | 4 a i) “44 “A ‘86 “40 “TAL ‘66 “41 TAL ‘26 T a eae | 1) Se ea ee Opis pus ‘xas “ON “480 poyVOIPUT JO SUOTIVAJOSGO PU S}UsMIEINSvoUMT [ejU10ISOg a a a a aa ae a eR Sire alee ceo eer gp bal atone ae IS flee rade y RE Shes rs LAS YAWaAA 1SUONJDALASQO PUD SJUAWALNSDAU 1DIUDIIISOg :8at4as TYIOQYI— W1AV\L, [BULL. 169 BUREAU OF AMERICAN ETHNOLOGY 76 *puo *pue “puso *puo *puo [eqsIp pues yeys Teqsip pus yeyg [e4sIp pus Weyg Te4sIp pue yeqg | weqsg | 7” eqodur0g TEAST DH DEBATE Gall ener terest aig nt erie eee ueuoedg PA TLORC Wa | enrim cae? ATIOSCV a. | eaxe sean A TOS (Vie cca so wie a QUOSO Wallen eee ra a eel ae NOS Vee os ee AUOSGIVIEH=-SaaSsaaes sseooid IejApuod eidng SO SUM tater nas mage Gyh 3a pil eee gee a a ATIOSU Wan (ees ae ee DOING | eet ee er amore OsIB |\tsai. et ORIG s ae eimjyiede jeyedag “O}IRUISice | canna SITE DULG] Clie | aaa DELL poaamee cs DENG) lisemaes = DENHalrae=H- Sas DENUFiligases eon PERO HSS see ae yyeys Jo edvyg :SMOTJBAIOSGO = G4 tal cael ema l Les? | CSi6Os | een E8025 | Saas OONT As eee 0A SS CON | KERR enema eae eee £/001/F Xopuy om Ot] Reece” a acs igee Oe | Se accipege veer RSE foes tee ap Ui lse | Beene me eae Wiles Pita eae een PA ae UIBIp “UIUI O[PPIVL F SACI (Ei Rape OSES cee: Say (a) in| FR reece ong oa Aad sak alae © wer £7 Ce haat ata Ce Pas ieee altel 0G;!| SaEseme peers ORs Pica ss eee WIP "Xv O[PPI € SCI ES EIEIO To terial brea Sahat The hice} Phe rnb 0b eames rosea eR LOOlp GX BU BOUT BIOL ETH ina in| Coa eae aan Cae ee ee o[PpluL 4B WeYs Jo odeyg :SMOTJBAIOSAO tot oalas Gab maperee Sika Spills see seo (OO) (0) 45 lepeaciecitiien Se tetaee || eaeaccal Gugeeleiemmnminccen obsess | igiaaaiasec cases eta ORM OM se to ee F/00LX¢ BTmeuDAye[d Jo xopuy SOG baie ees eas OGRA SSRSe S35 oo sears) [Gal eae gee ul Ac iar | to ts SRE ES” Soa Bl | Ea Sl Dee PON eaeewecerr a 4] “UIVIP “UUBIOJ OATITIJNN G (US| S9 cegepeie as Lee, ‘Gal Wiig nae’ Sk peace aie Sa (1) Sta] haar aes a> sas | ca in ereiegiie eens vor | RSS We 7 Be | A Soa lcs 3a ce d-8 “UIIp “U110J VATPIAINN F 37] 0)3 4) | eae ewe cage SORA Saas. ori ts Wet /AN | recone rghit iets call embok® Sram sine ce a5. |e nphaMagaaS oro se) ceases al aes BR Engr as eee Z/00LX€ XOpUr e[PPITAL OUI Ge uae e pee date reece = ba lpatentcan ia aang een ike [paabceteneN aobdenerl le ceeed Sooke seen. er ule || oes ee Reweeeyl lead od Gosaeee Seeker JBI “UIBIP OIPPTTAL TP ee Ra ese eke 900" Pe aaa tooo ee BE00 eee es cal aa TC a poe eed | le Be ee Ce ee ee d-8 “WIP e[PPTI Z FST ASS Fee SPSS aR Acas pace HoReee RTS cease Sea eso Somes Asean as | ise hse Se SSS sae a ses | Wake Ta SRE ornate Ss RRA Gar ae (Sessa sate ssseas Yjsu9] UNUTXe NY T iSeOIPUT pUe (“UIU) SJMOMEIMSevOYL “Ha “82/64 “Wal “Fe/6F “HS “COL "42 “TA ‘T9 i) “HI “TNL “TOT | el ee ees 1 eee etre ee cee ee Vidi penulyu0g— 8W01jMA1asqo pun syuawmasnsDawW JDIUDIISOg :8224a8 TIOdYI—L Z1IAV IL, 77 TUTTLE CREEK DAM, KANS.—CUMMING Riv. BAS. Sur. Pap. No. 10] *TBAO euod Jo jyey 1OTI04s0d “youT4stpUy JopIOg IO;IEAUT g “OSPII [8O]}19A AQ OMY UI POPIAIP oovyins JOpI03s0g 5 “mstid AIVUIPIO ¢ PANS =: |e aly “poyeUITyse Iyeys JO 1030SH, Se a ne nt a es FPUCTCISED, DUBIIGUSS os oe at as at ee te WETS oe iano. Mee 2 ee {NLS | pee een ie Bie ery pean ABU Ge Pec ems Ao Ie ee ee tomyoedg “OTPSUISTId [Bloyey |--------7- DOM BLOOERIMS | S[ POP eel saan see coms es eee DEQ e le casse eo eas ees ae ee DSU) ose no pen es oe cance edeys IJeUys-PLAL ‘SUOTIBAIOSGO Yysuay, WINUIIXe py :(UIUL) S}USUMIINSBOTAL “41 ‘4 ‘801 “41'S ‘06 “41 ‘TAL ‘90T “42 “T “61/6r —— a a ae eee a ee | ees a ee ee ee ee eee VIOdIa SI 8 (SM IS Sa ae OF0| nO Gl Sees aa ee pee ee ee ee OIG) CULO 6) = | anes emis wan goa on + eee a rere ee ueumoedsg OMICS Ae le at eee ae ae ee OMICS el wee tc nee aed teen a oe OJ GREAROUS TILES el Ueiaece cieen arate Gia et atl eia edeys WVYS-PIPL ISWOTIVAIOSAO ae we a (oy ae a oe Ge oe crs ie ae ae na em eae 1874 i> hos Bhai Poul TUNUIXe YT ?(‘uru) SJUIUIOIMSvI PAL SSS ee SN ie a ee ne ne ES en oe I Oe Fe “41 ‘TAL ‘98 “41 ‘TAL ‘88 “41 ‘A ‘ST/6F i ee ee ee ee VNIO SS ee oe Se ee es a ie, ee ee ee ee ee "pu [BIsIp pus yeyg |--~-------~ jhe) AE} 99) OOK A eh 0S SSS SS alata 949[Gu10D) | 5==-=-=--—- PCN TBS FO NI a | se weedy *XOAUTOO OOBJAMNS [PUIO}XG |-~-----7 XOAMOO ddRJINS [VeUIOIX| |~-~~--- XOAMOD OdBJAMS [VUIEIXY |--~---- XOAUODIGOGIILS | [6 W101 X hl | ee es Sone Sn ees edeys IJBYS-PIyT ‘SMOTJeAIBSGO a ais gta ee es ream eee S| ce ae ete ee mae on mae a OS GE. ee eee a eee iting iE: = Stn AL Aa LOT TANUIIX’ PAT >(‘UIU) SJUOMIeINseOyAL “41 ‘2 ‘16 “41 ‘(2) WW ‘48 “41 ‘TA ‘68 “41 ‘H ‘L1/6% ee eee ee eae Sniadva 78 BUREAU OF AMERICAN ETHNOLOGY [BuLu. 169 In June 1956, some 2 years after this report was completed and submitted for publication, William M. Bass checked the skeletal ma- terial from 14PO014 in the Division of Physica] Anthropology at the United States National Museum, Smithsonian Institution, and sub- mitted the following comments: The condition of the specimens, age, sex and a few measurements were checked. As a result of this study I feel that Cumming’s report is well done and would recommend only the following additions and changes: Cranium No. 66 (burial 4) showed intense copper staining on both sides of the skull, centering on the ears and right ascending ramus of the mandible. Less obvious copper staining was found on the distal end of the left radius and on both clavicles (Cumming, p. 51). There was a Slight suggestion of copper staining on the left supra-orbital margin of cranium No. 65. The three lower burials at this site, specimens Nos. 65, 66 and 73 showed a black staining on some of the bones, especialiy the arm and leg bones. This black staining may be due to manganese. Scrapings from the stained part of the bone yielded a positive micro-chemical test for manganese (micro-chemical test conducted by Edward P. Henderson in the Museum’s chemical laboratory). Webb noted similar stains on bones from Indian Knoll and upon chemical analysis concluded that under certain conditions where a limited amount of red ocher is used on the flesh of the dead body a chemical change occurs following decomposition, which results in a black staining of the bone (Webb 1946, pp. 247-48). Webb makes no reference to having tested the bones for manganese. The ages of all six specimens given by Cumming were too high. By careful analysis it has been possible to definitely establish the age of 2 individuals, specimens 49 and 66, and to lower the age of the remaining 4. Aging was based on tooth wear and epiphyseal union. The following are the ages assigned to each specimen: Specimen No. Age DOs Soe Ses ee ee ee ee 40-50 Aqesa a= i ee See ee eee 40-50 Gis a iets 6 tee Ce ee 380-40 ? GGL A Sees oh ia fe eee See ee 22-25 AQ! 2 dee el RE 2S ee Bee eee eee 17-21 ok EO 4 Wi ‘ Rd A ound Leloke exe bd J as RARE Ms " (SRR CO Ne , ated Gt PTI i 1 Van 7 tus Mae PO en ¥ Oe a ‘ ee ita th oi) a.) a ANY Pe » Wa oT eer Fs valk hb | : 2 ys 1 h Ws ” 4 h So BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 13 ATS TLE: a, A view of Sweat Bee Mound before excavation. 6, Sweat Bee Mound after clearing off the overgrowth. BUREAU OF AMERICAN ETHNOLOGY Re ee ae re 5s a, Sweat Bee Mound after removal of the soil fill. BULLETIN 169 PLATE b, Excavating Sweat Bee Mound. 14 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 15 a, Excavating and mapping Sweat Bee Mound. 43, View of the upper burials of feature 1 of the Sweat Bee Mound. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 16 b a, View of the lower burials of feature 1 of the Sweat Bee Mound. b, Views of the male skull of burial 2, No. 50, of the Sweat Bee Mound site. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE Views of the female skull of burial 1, No. 49, of the Sweat Bee Mound site. 17 BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 18 Sweat Bee Mound site. Projectile points: Nos. 31, 5, 29, 4. Blade: No. 38. Scrapers: Nos. 9, 131. Conch shell bead: No. 119. Catlinite object: No. 28. Disk shell bead: No. 117. Copper beads: No. 116. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 19 Sweat Bee Mound site. Scrapers: Nos. 123, 20, 32, 124, 1, 3. Cutters: Nos. 48, 7, 36. Modified core: No. 46. Anvil: No. 19. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 20 a, A view of the remaining portion of the Spillway site. b, Testing village site 14RY10. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 21 Spillway site. End scrapers: Nos. 1, 3,8. Side scrapers: Nos. 6, 9, 13, 5. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 22 a, Testing at the Reany site. Spillway site above cut in background. 6, The Spillway and Reany sites being destroyed by construction activities. Sweat Bee Mound in the foreground. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 23 Reany site. Rim sherd: No. 86/1. Body sherds: Nos. 28, 40/1, 27, 40/2. Strap handle: No. 85. Projectile points: Nos. 80, 50, 60, 49, 73, 61. Copper jingle: No. 83. BULLETIN 169 PLATE 24 BUREAU OF AMERICAN ETHNOLOGY ‘pL ‘ON vtoddoyg “SG ‘ON :opeTg “$9 ‘ZZ ‘ZTE “SON :sdedvios aplg “Ff ‘FE “sSON—Snoour]peostur $7 ‘99 ‘¢ *soN—¢ dnois ‘oT ‘TE ‘89 "soON—z dnoss t/T ‘T8 ‘8T “6 “SON—T dnois :siodeios pu ‘otis AuvoYy SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 169 River Basin Surveys Papers, No. 11 The Spain Site (39LM301), a Winter Village in Fort Randall Reservoir, South Daketa By CARLYLE S. SMITH and ROGER T. GRANGE, Jr. 422141—58——_8 79 Bis al Ae ne rae.) ‘i i A Bi i ie em! - : vk CONTENTS PAGH JENN Vel See ene es a ee ee ee Sa eee ea ee eran 83 PURE yo EG] ae eam en ale eater lens wetter So iia hela a OR Sly abe 85 Methodstofiexcayvation = 2 sean ae Ak: UO SIE es Se Toe 8 eg ee 87 Heaiimeds - cP OO Ole! es SUSI 72 NNT SE 9k ot ees ET 90 General description of features by number-____________________.___- 90 ROUSE Gee a a Rg ge Mi al a i 9 See a ge 91 PL GEURE EDOM ais 2 fess ee erg ir Og eS Mae ge ee ae ae ae el Be 93 | PE BY lg ec ed el SP tg Mat TPA Gp tpt eta es 94 Tomas Ware nas: fe blu SSeS, CAE OE See Ce eS SAE ee 97 RallkeinpeCraws Wiarecns2 teal See ae Seraph by gh 101 PUWTROG Wy ere ces sel 8 ee Re Re he ee a ni yee 102 IMGseelaneOuss 2-22 We eS ee ee ee OR oh Bae 103 G@iipned-stoneartiiacts. —- 2-222 oe ee eta te nee 104 | SIPC TECH HSMP 6.6) 01 2 a pipet ap by SJ Re peg NBER 104 DCrHpersuss fe 52) 22 ey 2, cee ee OR EOS We DETTE ON. 106 Feniverias ot ae Sen ets Soe ee IE ED ed he g 106 ROR hee Bp eee SE SS tt Cai ah ath Bas 107 Miscellaneous ‘chipped, stone... ae 22 oe oe et ge Be oe 108 Ce RO CNICIES GME AE VU RETOS 2 yt On eee ene aaa mre eee aes i 108 Soery smootherss 2 to. se Se, OUR, SEE Ape 108 Miseeliancos abraderse 22itwe See aie Or Peer oe 8 te Aye 108 LSP OloN (20 2S ee ees ec ee eee ee ee arn eee ae ple ae Eee ee 108 PURO RAAT) ONTO oe 2 Nee te ap ae es kg ee et ia 109 SLASTE LE VOTE 1 ee a eS PRI MS SY 5 el Ere, AAT em 109 ierie Gt Gani ec 2 Onset ne ten eye ee eee teen ee eee ee 109 Grmmmprecelul ns, 210. TRG Paes On Os) Ae ee OS Oe 109 Rangiestone antiiacis. ss i

b.-- 2. = li =. Ground bellenmnites -c5li 2-2 8 Rn UE ee 3 Gaming pieves= sae er ae Stenerdiskwemercs= Sot cee noe ere en SETS 1 Rough stone: ‘ 18 Fresh ie(s goR eee, San op rae Ee ee SLBIPLOMBAMMIerSLOneses ee = Leen eee ae ee Se Ree 12 Uammerstone, notched at side___._._-._____-.--___u.___. 1 Bismipnts: cette tite. Shee ee Limonite and hematite, fragments_-_..__._.._____-.___1_- x Shell: Sarapercss tsetse eee ee NEUSSOIS; COR OS\V ORME = san cee eters. sonne Senn see ee 14 NMiussel ied carmotcheG= ee ee es ee 1 Coppertaes + Steins ee eee eee Native copper bead, fragmentary__..__.___-_--____. .----2_ 1 Native copper sheet, fragmentary_-_..._......... _---_-_- 1 Bone and antler: Aolsvand!fiakers2+_-f. 2. =.= 22 ATICNArDUtHa WISE 6 ee Ue A 9 Roundedtpnth swiss tees s) 28 eee Soe ee LE re ee 2 bla Gea e e ae e Ee Oia ee 1 Bolinterawist i Ss shel 6 ee ages, Rea he 4 aniangulan shalt flakers® 9.0 2 U2 sg a Bs 3 Rib ilakorse «5 2 eee eae es SR. PSEA) PA) 3 soliton flakonsisic so) Boe es pee ie RE 12 at gs alae 13 Fragments of awls'and flakers--..2_/_.2 0 #2822. 22 20 Worked scapulae_-__-_.-... ~.-. FHoe}articular end unmodified .-. ¢8 te 12 TV OSHS EEA Pe 7894) Re ee AL Pace ean Ms PNR oe RSS 21 Fragments of hoes and knives... 2.5.) _ 2 ee 49 IMiscellancouse 222. 5 Peer a5! HUSHHOOKSES. wane ae Ed SINS PEN ee AR oe 2 Satrileee seas 2 eke he eee a we OL Bd 1 Piermont applicators: sath ee. Bert ea 1 Perforatediornaments.- oe <2 20 Set seg ae eee 5 Rib shaft wrenchest. 1s §t 25 Sid Ee ee 2 Cylindricalitipes: i= 4 8 Ss Ba ee ee ee 8 Wilma psi C ke. Meee SE A Se OE aL os Pe) 1 Cut hone eee ie a, ee SN ae ee 14 Cubiaritlen eka RS a SE aN) aD teed 3 COMPONENT B Pottery: ip Sherd 2225-6 == 22 ee iis: Cord lm pressed e. os eae se ee ee 1 Chipped stone: rojectile points=o2- ean ee Trianguloid, straight base, side notched______.______.____- 1 Trianguloid, straight base, corner notched_________________ 1 Trianguloid, convex base, side notched__________.._______ 1 Largetrianzuloid2 Lt SFE TOU 9s 1 Soe crit te) 4 Ganceolates straight bas@he.s ss. 022) o ubene en Ca eae Sau 2 Ground stone: UNS Elis pn. 2 liner Gia iia Same BV Vr EPOO VO akon oa = ean ee er Fe eee fe ee 1 UNASSIGNED Chipped stone: Projectile point__....2....__1--.- Lanceolate, straight base, ‘‘Yuma-like” _________.______.- 1 Occupation of the site only in the winter months by the people responsible for component A is suggested by the location, sheltered by tall trees, on an easily floodable low-lying remnant of a terrace. 118 BUREAU OF AMERICAN ETHNOLOGY [Bunx. 169 The scarcity of bird bones, so common in other earth-lodge villages in the region, argues for the use of the site in winter. Two bird bones were found. One is identifiable only as “duck”; the other is from a passenger pigeon, possibly an early spring migrant. The scarcity of the bones of immature individuals among the mammals suggests winter also. Bushnell (1922, p. 143) states that the Mandan and Hidatsa, “go, in winter, into the forests on both banks of the Missouri, where they find fuel, and, at the same time, protection against inclement weather. Their winter villages are in the thickest of the forest, and the huts are built near to each other, promiscuously, and without any attempt at order or regularity.” Bushnell uses for illustration Karl Bod- mer’s “Winter Village of the Minatarres” in his plate 44. 4 x x Toothed=honewesherw | 2 2 ko See) Se ee ee eae Strap -handles.on pottery: ti2. 2o e ee eeee aes onivpotteny.- asa ee ee Ee ie eae oe as pee ee eee Side-notched stone projectile point._--/ =. 2.2. 2 See eee eae BONG iSHhOd kisses Sere ae ate Ce NS ag ey a eee eee See fo eee Ne eS Bonetttibet? -22 7082 tee Or EUs 20s De ge Le ee ees cee Re Oe ens Smallewinter villages. 03s! a2 2. ee ec ae Ee eae Sere Wumerous;: small closelyiset, wall postsii23.. £22. Wiha ee A ee eee Se ee Shell-tempered: pottery associated. = 22. 2. 2 ee ee eee Lnop-handles/on-potteryn. SL ei Se es eee ae | a eee Stone gaming piece. = 22-22 e ee ee ee ee ee eee Lozengé-shaped knife. 2.3202. 498 l3_ 2eR eC ee ee) Ie Nia eee T-shaped miniature pipe__-...__-___- eee ee ee ee ee eee | Eee at ee eee Pebbie:pendant-— S552 Cee Fe ee eae ee eel ee Se eee eee Woorked=-fossil bellemnite:- 45-5 es Se ee | es eens XX KXKKKXKXKXXKKXKXKX It is impossible to weigh one trait against another and judge, for example, whether the presence of numerous, closely spaced wall posts is more important as a determinant than loop handles on the pottery. We conclude that the Spain, Wheeler, and La Roche components are rather similar in content. In our opinion this indicates that they all belong to the Chouteau Aspect, but not necessarily to the same focus. Component A at the Spain site appears to be the manifestation of a focus other than La Roche to which Hurt has assigned the Wheeler and La Roche components. The term Shannon Focus, named after Pvt. George Shannon of the Lewis and Clark Expedition, is assigned to the complex. In the course of our investigations in Lyman County a large surface collection was gathered from the Clarkstown site (map 5; pl. 25, a, b) and a small collection was obtained at the Deerfly site (map 5; pl. 25, a, b) by excavation. Both of the sites are described in the Appendix. The occurrence of these sites on high terraces suggests that they may have been summer villages. The collection from the Deerfly site is too small to permit assign- ment of its component B below the level of the Chouteau Aspect. The pos tg SPAIN SITE—-SMITH AND GRANGE 123 collection from component B at the Clarkstown site is assignable to the Shannon Focus along with component A at the Spain site. Addi- tional work was done at the Clarkstown site in 1954 by a field party from the Smithsonian Institution under Paul Cooper. The data gathered are not available for comparison. Other sites may be compared with the Spain site but not in such detail. This is because of lack of published data or inadequacy of the sample. No comment can be made on the White River, Bowman, Medicine Creek, and Evarts sites mentioned by Meleen (1948, pp. 26-27) because no detailed information is available. Hurt (1952, pp. 61-62), in his discussion of the La Roche Focus, includes the McClure site (map 5) which is known to him from the unpublished work of W. Raymond Wood of the University of Nebraska, and the Ericksen site, a village situated on the White River, but not specifically located. Stephenson (1954, p. 18) includes the Pishelville site, Knox County, Nebr., in the same focus. The Bennett Focus of the Chouteau Aspect has been suggested (Stephenson, 1954, p. 18) as a name for the cultural manifestation characterizing the Black Widow B and Meyer components (map 5) in Oahe Reservoir. The descriptions of the pottery and other arti- facts from the Meyer site (Hoard, 1949) are inadequate for compari- son, and the data from the Black Widow site, excavated by Paul Cooper for the Smithsonian Institution, have not as yet appeared in print. It seems best to reserve judgment in regard to comparisons between the Shannon Focus and the Bennett Focus. The age of the Chouteau Aspect must be estimated by indirect means. Meleen (1948, p. 31) submitted samples of wood from the La Roche site to George F. Will for dating in terms of a dendrochronological chart established in North Dakota. A date bracket of A. D. 1484-— 1475 was determined for the site. Lehmer (1954, pp. 137-138) has questioned the applicability of the master chart for south-central North Dakota in central South Dakota on the basis of the marked variance in the rainfall records of the two areas. Hurt (1952, pp. 61-64) finds that the dating does not fit the archeological evidence in relation to other culture complexes, notably the Thomas Riggs Focus. It seems best to disregard this application of Will’s dates and to approach a solution to the problem in other ways. First, we are safe in assuming that the Chouteau Aspect is prehistoric on the basis of the absence of European trade goods in the sites that have been excavated. It is doubtful if any trade goods reached vil- lages in the area under consideration before the closing years of the 17th century. Present in the area are manifestations demonstrably older than the Chouteau Aspect such as component D at the Talking Crow site (Smith, MS.) and the Arzberger component (Spaulding, 124 BUREAU OF AMERICAN ETHNOLOGY [Bun. 169 1956), assignable to the Aksarben Aspect. They are characterized by pottery similar to that from western Nebraska dated about A. D. 1300-1517 on the basis of dendrochronology at Ash Hollow Cave (Champe, 1946, pp. 48-50). We would place the date of similar mani- festations in South Dakota at about 1500 because the culture seems to be intrusive there and to contribute traits to other cultures known to be later on the basis of stratigraphy and seriation at the Talking Crow site (Smith, MS.). The absence of cord marking as a surface finish on the pottery of the Chouteau Aspect indicates that it falls later than the Aksarben penetration in time. This means that the Chouteau Aspect has a maximum time span of from some time after A. D. 1500 until some time before A. D. 1700. A conservative estimate for any one of the components is about A. D. 1550-1650. The Chouteau Aspect, with its distribution along the Missouri River from central South Dakota into northeastern Nebraska on the time level just prior to the advent of European traders, may represent the culture of the traditionally related Arikara and Pawnee. This inter- pretation is offered as a speculation subject to modification in the light of later findings. TABLE 5.—The archeological sequence in Fort Randall Reservoir 1900 Historic Dakota (Talking Crow A, Oacoma A, Clarkstown A, Deerfly A). 1800 ? 1700 Pahuk Aspect (Talking Crow B, Oacoma B, Oldham A). Great Oasis Aspect (Oldham B). 1600 Chouteau Aspect (Spain A, Wheeler, Clarkstown B, Talking Crow C, Deerfly B). 1500 Aksarben Aspect (Talking Crow D). 1400 ? 1300 Chamberlain Aspect (Swanson). 1200 ? 1100 Woodland Pattern (Scalp, Ellis Creek, Spain B, Talking Crow E). ? Paleo-Indian (?). In table 5 the sequence of cultures in the Fort Randall Reservoir is set forth in simplified form. The table is based on discussions at the 11th Plains Conference for Archeology (Stephenson, 1954) and on data from the Talking Crow site (Smith, MS.). The dates are rounded estimates. No attempt is made to classify the cultures below the level of the aspect. Letters following the names of the sites designate sep- arate components based on stratigraphy, on seriation, and on the basis of separating known complexes in mixed collections. 2 At the 11th Plains Conference for Archaeology (Stephenson, 1954) it was suggested that the archeological complexes previously separated as the Upper Republican and Nebraska Aspects be grouped together and called the Aksarben Aspect. Cultural mani- festations in South Dakota with Central Plains affinities often share traits with both the Upper Republican and Nebraska Aspects and the term Aksarben is a useful one in this situation. ‘The validity of the new term must be tested by means of a review of the data from both areas, ee ig SPAIN SITE—SMITH AND GRANGE 125 The sequence in the vicinity of the Oahe Dam, designated as Fort Pierre Branch by Lehmer (1951, 1954), lacks manifestations at- tributable to the Paleo-Indian and Woodland. The Anderson and Monroe Foci are part of the Chamberlain Aspect. Sites comparable to Thomas Riggs appear to be absent in the Fort Randall Reservoir. Although not part of the Fort Pierre Branch the Aksarben Aspect is present in the form of the Arzberger site south of Pierre. The Chouteau Aspect is represented by Lehmer’s “Meyer and La Roche” grouping. The Great Oasis and Pahuk Aspects are absent in the Fort Pierre area. Hurt (1954) places the Spotted Bear Focus prior to Lehmer’s Stanley and Snake Butte Foci and after Chouteau. Find- ings at the Cottonwood site (Hurt, 1954) indicate occupation of the area by the Dakota on about the same time level as in the Fort Randall Reservoir. SUMMARY AND CONCLUSIONS The latest and principal occupation at the Spain site is attributable to the Shannon Focus of the Chouteau Aspect. A minor occupation attributable to the Loseke Creek Focus of an unnamed aspect of the Woodland Pattern is also present. The possibility of a still earlier occupation in the region, if not at the site, is indicated by the presence of the basal portion of a Yuma-like projectile point. The latest occupation of the site seems to have taken place in winter. The component represents a community of earth-lodge-dwelling people situated in a place protected from the wind and handy to a supply of wood. A conservative estimate for the period of occupation would be between 1550 and 1650, perhaps about 1600. APPENDIX THE CLARKSTOWN SITE (39LM47) In the course of the season in which the Spain site was excavated we visited the Clarkstown site (89L.M47), situated in the point of land between the White and Missouri Rivers on the second rise of ground at an elevation of 1,350 feet, about 214 miles northeast of the Spain site (pl. 25, a, by map 5). The legal description of the locality is NW, SW, sec. 22 T103N R72W. The Lewis and Clark Expedi- tion passed the mouth of the White River on September 15, 1804. Clark (Thwaites, 1904, p. 148) states that “. .. in the point is a butifull situation for a Town 3 gradual assents and a much greater quantity of timber about the mouth of this river than useal, .. .” Because the archeological site is the only settlement in evidence we have named it Clarkstown. The site had been freshly plowed and we were able to gather a large collection of artifacts from the surface. The pottery is identical with 126 BUREAU OF AMERICAN ETHNOLOGY [Buue. 169 that from the Spain site. The collection comprises 1,060 potsherds, of which 945 are body sherds and 115 are rim sherds. The body sherds include 540, or 57.2 percent, simple stamped; 204 or 21.5 percent, plain; 201, or 21.3 percent, decorated shoulder. Eighty-one of the sherds from decorated shoulders are classifiable as follows: Opposed diagonal lines of incising, 35.8 percent; incised herringbone motif, 39.5 percent; incised parallel lines arranged ver- tically, 22.2 percent; punctated, 2.5 percent. Rim sherds are represented by 115 specimens, 91 of which are classifiable by type: Iona Indented, 39, or 42.9 percent; Iona Diag- onal-Incised, 3, or 3.8 percent; Iona Horizontal-Incised, 19, or 20.8 percent; Iona S-Rim B, 5, or 5.5 percent; Iona S-Rim C, 1, or 1.1 per- cent; Talking Crow Straight Rim, 16, or 17.6 percent; Grey Cloud Horizontal-Incised, 8, or 8.8 percent. The group of 24 miscellaneous rims includes 8 plain rims, 1 strap handle, 2 miniature vessels, and 14 unclassifiable sherds. Among the rim and body sherds are eight that are shell tempered. The total number of horizontally incised rims is 33, or 28.7 percent. This amount is close to that found at the Spain site. Artifacts of stone and bone from the Clarkstown site are few in number and do not appear to differ significantly from those present at the Spain site. In the eastern part of the site objects of 19th century White manufacture and pieces of catlinite bearing marks from metal saws occurred. Perhaps most of these specimens pertain to the occupation of the region by the Dakota. Mattes (1949, p. 518) notes that the south half of section 22 was the probable location of Medicine Bull’s camp and a mission house in about 1890. The later occupation is designated as Clarkstown A, the earlier as Clarkstown B. THE DEERFLY SITE (39LM39) Excavations were carried on at the Deerfly site (39L.M39) from June 13 to June 22, 1954. The site is situated on a sloping terrace a mile north of the Spain site in the NEV, sec. 29 T103N R72W (map 5; pl. 25, a, b). The site was so named because of the prevalence of deerflies in this locality in contrast to others. Occupation A consists of a series of log cabins and sheds attributable to the Dakota of the last quarter of the 19th century. These structures have left small, shallow depressions thought to represent the remains of earth lodges until excavated. Numerous items of late 19th century White manu- facture including glass beads were associated with catlinite, cut with metal saws, and a perforated pebble on the floor of one of the structures. Test trenches in the area yielded meager evidence of occupation B, characterized by 18 potsherds and a few artifacts of chipped and rough stone. The pottery is identical with that found at the Spain Riy. Bas. Sur. Pap. Noli}. SPAIN SITE—SMITH AND GRANGE 127 and Clarkstown sites. The three rims found are identifiable as Iona Indented, Talking Crow Straight Rim, and unclassified plain. Five of the body sherds are simple stamped; 10 are plain. LITERATURE CITED BUSHNELL, Davin I. 1922. Villages of Siouan, Caddoan, and Algonquian tribes west of the Mississippi. Bur. Amer. Ethnol. Bull. 77. CHAMPE, JOHN L. 1946. Ash Hollow Cave. Univ. Nebraska Stud., N.S., No.1. Lincoln. Cooper, PAUL L. 1949. Recent investigations in Fort Randall and Oahe Reservoirs, South Dakota. Amer. Antiq., vol. 14, No. 4, pt. 1. DuNLEvy, Marton L. 1936. Comparison of the cultural manifestations of the Burkett and the Gray-Wolfe Sites. Chap. Nebraska Archaeol., vol. 1, No. 2, pp. 147-247. Lincoln. GrorRGE, EDYTHE L. 1949. Report of the investigation of the Robinson Site, Hughes County, South Dakota, Peoria Bottom, 1948. State Archaeol. Comm., Archaeol. Stud. Cire. No.1. Pierre. Hii, 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. 3-71. Lincoln. Hoarp, Lyon J. 1949. Report of the investigation of the Meyer Site, Stanley County, South Dakota, near the mouth of the Cheyenne River, 1949. State Archaeol. Comm., Archaeol. Stud. Circ. No. 2. Pierre. Hurt, WESLEY R., JR. 1952. Report of the investigation of the Scalp Creek Site, 89GR1, and the Ellis Creek Site, 39GR2, Gregory County, South Dakota, 1941, 1951. South Dakota Archaeol. Comm., Archaeol. Stud. Cire. No. 4. Pierre. 19538. Report of the investigation of the Thomas Riggs Site, 39HU1, Hughes County, South Dakota, 1952. South Dakota Archaeol. Comm., Archaeol. Stud. Cire. No.5. Pierre. 1954. Report of the investigation of the Spotted Bear Site, 39HU26, and the Cottonwood Site, 39HU48, Hughes County, South Dakota, 1953. South Dakota Archaeol. Comm., Archaeol. Stud. Cire. No. 6. Pierre. Kivett, Marvin F. 1952. Woodland sites in Nebraska. Nebraska State Hist. Soe. Publ. Anthrop., No.1. Lincoln. LEHMER, DoNALD J. 1951. Pottery types from the Dodd Site, Oahe Reservoir, South Dakota. Plains Archaeol. Conf. News Letter, vol. 4, No. 2, pp. 3-29. Lincoln. 1952. The Fort Pierre Branch. Amer. Antiq., vol. 17, No. 4, pp. 829-336. 1954. Archeological investigations in the Oahe Dam Area, South Dakota, 1950-1951. Bur. Amer. Ethnol. Bull. 158, Riv. Bas. Surv. Pap. No. 7%. Mattes, MERRILL J. 1949. Report on historic sites in the Fort Randall Reservoir area, Missouri River, South Dakota. South Dakota Hist. Coll. and Rep., vol. 24, pp. 470-577. Pierre. 128 BUREAU OF AMERICAN ETHNOLOGY [Buxy. 169 MELEEN, ELMER EB. 1948. A report on an investigation of the La Roche Site, Stanley County, South Dakota. Univ. South Dakota Mus., Archaeol. Stud. Cire. 5. Vermillion. SMITH, CARLYLE S. 1951. Pottery types from the Talking Crow Site, Fort Randall Reservoir, South Dakota. Plains Archaeol. Conf. News Letter, vol. 4, No. 3, pp. 82-41. Lincoln. 1953. Digging up the Plains Indian’s past. Univ. Kansas Alumni Mag., vol. 52, No. 4, pp. 4-5. MS. [Data in manuscript on the Talking Crow site on file with the Univ. Kansas, National Park Service, and Smithsonian Institution.] SOLECKI, RALPH §. 1953. Exploration of an Adena Mound at Natrium, West Virginia. Bur. Amer. Ethnol. Bull. 150, Anthrop, Pap. No. 40. SPAULDING, ALBERT C. 1956. The Arzberger site, Hughes County, South Dakota. Occas. Contr. Mus. Anthrop. Univ. Michigan, No.16. Ann Arbor. STEPHENSON, ROBERT L. 1954. Taxonomy and chronology in the Central Plains-Middle Missouri River Area. Plains Anthrop., No.1, pp. 15-21. Lincoln. STRONG, WILLIAM DUNCAN. : 1935. An introduction to Nebraska archeology. Smithsonian Mise. Coll., vol. 93, No. 10. 1940. From history to prehistory in the Northern Great Plains. Smith- sonian Mise. Coll., vol. 100, pp. 353-394. THWAITES, REUBEN GOLD, Eprror. 1904. Original journals of the Lewis and Clark Expedition, 1804-1806. New York. TOLSTOY, PAUL. MS. Some Upper Missouri pottery types. MS. on file in the Dept. Anthrop., Columbia University. WEDEL, W. R. 1986. An introduction to Pawnee archeology. Bur. Amer. Ethnol. Bull. 112. are _ tne rn RECA. wy et [ A. we RAY (ii oy iu iy rae ae 4 a) iat i) ae BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 25 Aerial views of the Spain site and its environs. a, Oblique view to the northeast; b, vertical view by U. S. Department of Agriculture. A, Spain site; B, Deerfly site; C, Clarkstown site; , Point from which oblique view was taken. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 26 Views of excavations. a, Excavation unit 2, to north. 5b, Ten-foot square in Feature 1 to east. c, Feature 1 completely excavated, to east. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 27 Bgetisiens tee 73 SECS 53 RAS eeinlt ec ali . Se Seon Views of excavations and artifacts in situ. a, Excavation of refuse south of house 1. b, Blades in situ on the floor of house 1. c, Fireplace in house 2. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 28 House | during and after excavation. a, Starting cross trenches, to north. b, Completely excavated, to west. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 29 Iona ware. a-d, lona Indented; e-h, Iona Horizontal-Incised. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 30 Iona and other wares. a-d, Iona S-Rim; e-g, Iona-Diagonal-Incised; h-j, Talking Crow Straight Rim; &-m, Grey Cloud Horizontal-Incised. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 31 l Unclassified pottery. a, Indented rim; 5, incised collared rim; c-d, plain rims; e, loop handle; f-g, shell-tempered sherds; h-k, m, decorated shoulder sherds; /, simple stamped body sherd. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 32 AAA 8 Se Chipped stone artifacts. Upper, Nos. 1-18, Projectile points; Nos. 19-28, end scrapers; No. 29, side scraper. Lower, No. 1, side scraper; Nos. 2-15, knives. Scale in lower group applies also to upper group. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 33 Chipped, ground, and rough stone artifacts. a-c, Drills; d, blade; e-h, shaft smoothers; 1, J, abraders; k-m, hammerstones. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 34 Stone, shell, and bone artifacts. a, Pipe; b, pendant; c, worked fossil bellemnite; d, disk; e, notched shell; f-/, awls; m, flaker; n-p, tubes; g, ornament; 7, fishhook; s, pigment applicator; zt, flaker; u, knife; 7, spatula; w, shaft wrench. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 35 Miscellaneous artifacts. a-g, Woodland projectile points; h, Yumalike point; i, Ellis Cord- Impressed sherd; 7, Woodland ax; k-m, Shannon focus hoe, flaker, and pick. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 36 Photomicrographs of native copper. a, Microstructure of cross section of bead xX 100). Taph PI » + Z b, Microstructure of cross section of amorphous fragment (X 100). SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 169 River Basin Surveys Papers, No. 12 The Wilbanks Site (9CK-5), Georgia By WILLIAM H. SEARS 129 422141—58——_12 ee a HORPUEDEYT ArAowuenee - - teolvad) si neatromtk tn waste. wey Wak edict sire) thy fave Gwe ianitrinct as ead eer ny i ‘i Miavatrectutd Tet Meeker a marks ils: Yreatee CONTENTS PAGE Puiieeriuction 94728 es 4 a Se os ee es SE ee ee 133 hersite and excavations - pM 2 = wee Eee ee le ot eee 138 Hie earth lodge. <5. - = S| Ai eee te rer i I Pe te he a 143 PEne COW RN DCrOG. 25.54 See oe eh ee eee 149 Phase Ae. s2 oes 22S ae Sa eee ee ee we See Soe os eee 149 IP ASer en ee oak ee Oe es et SS ee ee Le 154 ‘Ehercartnlodges..2-- ey eee ee ei ae 160 Biowal: period summaryowme + 2 ee See ee ee ee eee 163 The-Wilbaoks period... mes... BYE so eo eae 172 CHO VAIar periods <9 2). an a ee Le ee ee eee Se eee 177 Bummary, and conclisions: ae... Sask Nie eee! Se Pe SSL Le 178 Witerature Cited: 24. = Sin. Pe he oa Se be eee eS Soe 183 Appendix A. References to published pottery type descriptions citedin text. 187 Appendix 8. Pottery type deseriptiong® ..2%. =. 2 ye 189 Htowabeblain2 2+. es soos] eee CO Dee heim Se oe 189 Etowah Complicated Stamp: 23 “Akay. Rw ee te ea ee 189 EiowalBurnished® Blain & 1) Sue See ea ee ee LE De 190 Etowah: Polished, Bisek! <2. te =U SO i ele 4 ee ee Ele 191 iowa ed: Hilmecs. 1208" . xtyy! IAM 9 Re 192 WilbanksPlain. © & 23° 4c. 56 ee SR 192 Wilbanks Complicated Stamp-_-._ 2222 ee 5. === 193 DIXeS ME IAI es 45h iee: 202 02) Soe 22 See See on eee 194 ILLUSTRATIONS PLATES FOLLOWING PAGE 37. Etowah Complicated Stamp (diamond motifs)_.--._---------------- 194 38. Etowah Complicated Stamp (line block and filfot cross motifs) ---_--- 194 39. Wilbanks Complicated Stampe= #42 _..- =... 5 194 40. ‘Sixes Plain and EtowahbjIneised a) 42 2-8 te tet 194 41.4Plate' and bottle rims, various wypes#=. 522220 2 Si. 2 SUS Se Susi Sse 194 42. Notched rims, Etowah Burnished Plain, engraved sherd, sherds from effigy tvessels <5 25 Bee LF A oh ee Fee See SoBe 194 43. Hiwasee Red Filmed, Etowah Red Filmed, Hiwasee Red on Buff__-- 194 44. Pipes; chisels, celt, aud pettery/disks. ..._..-_-_...-..-..--.--=---- 194 45. Pottery ladle and bead, earspools, bone awls, shell spoon___-__------- 194 TEXT FIGURES PAGE 5. (Frontispiece.) The earth lodge (artist’s reconstruction) --._..------- 132 6. a, Profile, main cut north of earth lodge; b, profile, main cut including roof fill and wall construction south side of earth lodge____---- facing 140 fPRloor plansedriiodge we pe. Aeoe fe. cet eee cee facing 142 8. Chart showing proportional occurrence of major motifs used in Etowah Compltatedi Sham | fe se te OS ol EO 165 9. Chart showing development through time of angular motifs used in North Georgia Complicated Stamps_.........£_-2.2-----------=- 168 MAP eeLoecation map of ©K pol ht ao ee ee ce ee 139 [BULL. 169 BUREAU OF AMERICAN ETHNOLOGY 132 “punoj se sjivjop Jo} / puke 9 soin3y 999 [PM Sey oprsul isnf Ppeuor{ses ssols ‘a8 po] Yiive S-MD 2y) Jo uonsnIj}suose1 sashiy—s¢ aun y THE WILBANKS SITE (9CK-5), GEORGIA’ By Wiuram H. Sears INTRODUCTION The Wilbanks site, 9CK—5 in the nomenclature of the River Basin Surveys, Smithsonian Institution, is located on the flat bottom lands of the Etowah River in north Georgia. As this river is in the foothills of the Appalachians, it flows in a southwest-northeast direction, fol- lowing the line of the ridges and valleys, rather than flowing north and south as do most of the rivers in Georgia. At Rome, near the Alabama border, the Etowah joins the Oostanaula to form the Coosa, and eventually empties into the sea through the Alabama River. The area generally is quite hilly, and heavily timbered even today. The only flat areas of any size are the river flood plains, which are composed of the most fertile soils in the area. Archeology in the Etowah valley has a long history, beginning in 1883 with excavations by Rogan in mound C at the Etowah site for the Smithsonian Institution (Thomas, 1887, pp. 95-109). Fifty years later, W. K. Moorehead returned to the Etowah site and conducted far more extensive explorations, completely demolishing mound C and thoroughly postholing the other two large mounds. Large sections of the village area were also opened up (Moorehead, 1932). The work of Rogan and Moorehead aroused much interest, sustained profession- ally to the present day, through their finds of exotic artifacts, now known to be paraphernalia of the widespread Southern Cult (War- ing and Holder, 1945, pp. 1-34). The large repoussé copper plates in particular have been illustrated and described many times, and at least their motifs—warriors in bird costumes—are noted in every discussion of the Southern Cult or of North American Indian art. During Moorehead’s excavations at Etowah, Margaret Ashley worked at other sites in the neighborhood, particularly Carters Quarters and Pumpkinvine Creek (Ashley, 1932). After Moorehead’s departure, the valley quieted down again for a few years, but woke up again in 1938 when Robert Wauchope con- ducted a survey for the University of Georgia, this time a survey with a large number of test excavations in a large number of sites. His preliminary report on this work, and a following paper on com- 1 Submitted December 1952. 133 134 BUREAU OF AMERICAN ETHNOLOGY [Buny. 169 plicated stamp-design development (Wauchope, 1948 and 1950), have done much to clear up the assortment of puzzles and contradictions left from the earlier years. Following Wauchope, J. R. Caldwell did further intensive survey work for the Smithsonian Institution in the Allatoona Basin, about to be flooded by a power and flood control dam near Cartersville, a few miles upstream from the Etowah site. For a time it appeared that Federal funds would not be available for salvage work in this basin, which extended 20 miles upstream to Canton. To afford a partial remedy for this unfortunate situation, the De- partment of Anthropology at the University of Georgia excavated the most promising site in the basin, under the direction of the writer. This report describes the results of that work. To bring the history of archeology in the Etowah Valley up to date, the Smithsonian Institution investigators returned in 1949 and 1950. A large number of sites were excavated. Reports were pre- pared by the archeologists in charge, J. R. Caldwell and C. F. Miller. The Wilbanks site appeared particularly promising for our one- shot salvage attempt for several reasons. First, Wauchope reported, from his test trench, a rather thick stratigraphic column (Wauchope, 1948, p.205). Although he had worked out the basic ceramic sequence for the valley from the sequence at this and other sites, there was need for checking and expanding the outline. Second, a gully a few yards from the mound, which was the main feature of the site, had been producing, through the washing out of graves, artifacts re- lated to the Southern Cult. Most impressive of these was a pair of carved stone earspools, one of which is illustrated on plate 45, 4. We had a chance then to get a large ceramic sample in good stratigraphic context, and perhaps to tie the Southern Cult in to some specific seg- ment of the cultural continuum represented by the ceramics. As indicated in the preliminary report (Sears, 1950), we succeeded in a part of this program, particularly in working out a ceramic se- quence with a large sample from well-defined strata. It was also possible to relate one segment of this ceramic sequence to the building period of a ceremonial structure, since the mound, the excavation of which consumed most of our time, was a collapsed earth lodge with a large amount of pottery on the floor. The Southern Cult relation- ships have proved to be a bit more difficult to handle, but it is reason- ably certain that the cult had its peak during the second half of the total span of Etowah culture occupation of the site, as defined on pages 163-171, and almost completely certain that the cult in this area pertains to the Etowah culture rather than to the later Wilbanks (formerly Savannah) or Lamar Periods. A tentative outline of the cultural sequence, concerned largely with ceramics, has been presented in the preliminary report. On the whole, this outline has been retained after study of the complete col- Riv. Bas. Sour. — Be. Raed WILBANKS SITE, GEORGIA—SEARS 135 lection. A few changes have been necessary, but these are largely terminological. The present sequence, using a terminology for the Etowah valley agreed on by the writer and J. R. Caldwell, is as follows: ErOtonistorie@: Perlod see ee ee eae Lamar mate: “Mississippi 2) se ee aes _ eee Wilbanks Mature Mississippi? 4 er tontaloes) Sie ae Etowah Period IV Tit II I There are several departures from standard usage in the termi- nology used in this sequence. First, the Wilbanks period has been called the Savannah period by Caldwell (1950, pp. 11-18) and Fair- banks (1950, pp. 148-144, 147). The writer engaged in a bit of circumlocution in the preliminary report (1950) and called it the period characterized by Savannah Complicated Stamp. A detailed discussion of the problems will be found on page 172. It should be pointed out here, however, that it has been necessary to recognize the complicated stamped pottery involved as a type definitely distinct from Savannah Complicated Stamp, although it is closely related and is diagnostic of the same time period. Secondly, even if the stamped type had been Savannah Complicated Stamp, the remainder of this north Georgia complex would still be much different from that char- acteristic of the Savannah period as it was originally defined on the basis of materials excavated at the Irene site (Caldwell and McCann, 1941). The terms “protohistoric,” “Late,” and “Mature Mississippi” are of course the writer’s subdivisions of the more generally accepted term “Late Mississippi” as used by Griffin (1946, pp. 37-95) and others. It is felt that this period needs subdividing, particularly for work in Georgia, because of the large number of discrete complexes at many sites, all classifiable as Late Mississippi in the present system. Gener- ally speaking, such extant classifications as those of Griffin (1946) and Ford and Willey (1941) segregate the earliest Mississippian culture type, readily separable in most cases from the preceding Middle Woodland-Hopewellian manifestations. However, two readily dis- tinguishable, although genetically related, culture types are classed together in the second Mississippian period, called Late Mississippi or Temple Mound II. This lumping is probably a product of the chronological scales used until very recently, which would not bring Early Mississippi to an end until about 1500, leaving so little time for other developments that spatial rather than temporal concepts were usually employed to account for observable differences. As noted above, it is immediately obvious in Georgia that spatial differentiation 136 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 is only part of the answer. In the writer’s belief, that same situation obtains in other parts of the Southeast. “Mature Mississippi” is a term introduced by the writer, inspired by James B. Griffin, intended to be descriptive of the culture type and period indicated as the major type and period represented at the Kolomoki site in southwest Georgia (Sears, 1951 a). At that site, and in other cases, as we hope to demonstrate, we are dealing neither with Early Mississippi, well defined by Griffin (1946), Ford and Willey (1941), and others, nor with the protohistoric period which is invariably represented by the Lamar Complex in Georgia. Mature Mississippi is then considered to be the peak period of Middle Mississippi culture, the period of greatest artistic and ceremonial development wherever Middle Mississippi culture may be found. In Georgia and elsewhere it is the period of the Southern Cult and of the flamboyant expressions of a different ceremonial complex at Kolomoki (Sears, 1951 a, 1951 b, 1951 c). In Tennessee, Alabama, and neighboring regions it is the period of the Southern Cult again. It is separated, in broadest terms, from Early Mississippi by the development of elaborate ceremonial complexes; from the proto- historic developments by the reverse situation, the decline of this ceremonialism, lack of the great mound sites, and probably by villages that were smaller but much commoner. In any event, in most areas of the Southeast it is not too difficult to separate a climactic Mature Mississippi period from the related but ceremonially and artistically simpler culture of the protohistoric period. In this report, we have continued to use the term “Late Mississippi” to cover the Wilbanks period, since it seems to fit with neither of the other periods, actually being an intrusive culture in the Etowah Valley. Perhaps when we know it better, ascription to one of the other periods, mature or proto- historic, will be possible and will simplify matters. The writer feels that he should point out here that he is fully aware of the inherent dangers of classification in archeology, particularly in that it often becomes an end in itself. Yet we must all engage in classifi- cation, in breaking our subject matter down into units that can be handled conceptually, if we are to progress. For analytical work in archeology, we need, and all create, terminological pigeonholes, each term representing a class. Without them, obviously, every manifesta- tion in the Eastern United States would have to be taken into account for all comparative studies in the area. The point here is that as the bulk of known material increases, finer classification is needed to keep the bulk broken down into units that can be handled for comparative studies. Middle Mississippi is a culture type readily recognizable in many varying forms throughout the Eastern United States. The divi- sion into Early and Late eventually became a necessity, to provide Riv. Ss. ait Bag, Sue WILBANKS SITE, GEORGIA—SEARS 137 pigeonholes for some of the variations as soon as it became obvious that the total culture type had two major sets of temporal, and con- sequently cultural, constants through the course of local variation. This fact was recognized some time ago by the provision of pigeon- holes labeled Middle Mississippi Phase and Upper Mississippi Phase in the Midwest Taxonomic System (McKern, 1939). It now appears that it is possible, and equally necessary, to subdivide the Late Missis- sippi horizon, since sets of temporal and cultural constants for lesser divisions have become apparent. The total terminology reflects not only classification per se, but also cultural development in the South- east over a long time period, recognizing the facts that while there is a basic culture type, Mississippian, shared by many ethnically discrete peoples, this basic type also changed, and the changes were par- ticipated in and contributed to by these diverse peoples. Sixteen pottery types are recognized in this study, resulting from the analysis of 25,351 sherds. Except for a few sherds representing types traded from the Tennessee area, they are local products. Also recognized and documented are ceramic styles, modes of decoration, and vessel shapes which are found in other pottery types foreign to the Etowah Valley, and which consequently indicate relationship with the cultures productive of these other types. In broadest outline, it may be stated that the Etowah culture, al- though subdivided into a number of periods, is in the Mature Missis- sippi period, and is roughly coeval with the Dallas (Lewis and Kneberg, 1946) and large-log town house (Webb, 1938) cultures of Tennessee; Kolomoki in southwest Georgia (Sears, 1951 a, 1951 b) ; Late Weeden Island on the Florida Coast (Sears, 1951 a); Mound- ville in Alabama (Moore, 1905) ; Spiro in Oklahoma (Orr, 1946) ; and others less well known but equally important. The Wilbanks cul- ture, too, obviously shared in Middle Mississippi culture, as indicated by the basic vessel form and a few other items. It was not as spe- cifically and definitely a participant at Etowah culture, however, at least not in the light of present knowledge. Lamar, while typically a South Appalachian Province product (as are the other two at CK-5) is very definitely a part of the rather uniform protohistoric culture of the Eastern United States and the Plains area—uniform as compared with the situation in earlier periods, that is. As did the other cultures in this time period, Lamar appears to have participated in a general leveling out process, a tendency toward cultural uniformity that is particularly evident in the ceramics. References for pottery type descriptions available in print will be found in Appendix A. New descriptions are in Appendix B. I wish to express my thanks to the many individuals and organiza- tions, listed below, who have aided in making the excavations and this report possible: A. R. Kelly, who initiated the project and held down 138 BUREAU OF AMERICAN ETHNOLOGY [BuLt. 169 the home office; the district office of the United States Corps of En- gineers for much helpful advice, the loan of a vehicle, and surveying equipment; Robert Wauchope and J. R. Caldwell for advice and sug- gestions on Etowah valley archeology; Smith Johnston, Jr., of Wood- stock, for a place to live, for finding equipment, and for helping to find a crew; the crew, particularly for doing their best through some very bad weather in November and December of 1948; Miss Doris Pogue and Mrs. M. Albertson for their excellent work on the text figures included in this report; and J. Mahan, E. Klein, and students at Reinhardt College for washing the pottery. THE SITE AND EXCAVATION Since this site, about midway between the towns of Canton and Cartersville, on the south bank of the Etowah River, was in the center of the Allatoona Reservoir, it is now under 80 to 90 feet of water. The river made a bend at this point, and the area of occupation was nearly in the center of this bend, about at the mid-point of a crescent- shaped stretch of flat bottom land 14 mile wide and 114 miles long. During the population peak in the Etowah period, the village covered approximately 80 acres. Near the center of this bottom-land crescent, on the hillside, a small stream issued from the hills and flowed along the base of the ridges to join the Etowah. However, the bottoms are so flat that in flood stage the Etowah often reversed itself and flowed across the center of the bottom land crescent and into the small stream. As this occurred half a dozen times in the past 50 years, a channel, normally dry, was cut that at times bisected the arc of fertile river bottom. A large number of graves were washed out in this stream bed or gully, some of them yielding the Southern Cult material noted in the Introduction. The larger portion of our time and available funds were spent on the mound, which was the most prominent feature of the site. It occupied a dominant position in the center of the river side of the site, just a few yards in from the riverbank. It was hoped that this struc- ture, approximately 220 feet in diameter and 6 feet high, would pro- duce more cult artifacts, this time in tight archeological context. This hope was not realized, although the excavation did yield important information about both ceramic sequence and the type of ceremonial structure used by the Etowah culture. ‘The mound and some surrounding area was first staked out in a 5-foot grid system, oriented with the cardinal directions. Since the outward appearance of the mound gave us no indications of its type or inner structure, exploratory trenches were run in on the main grid axes. If we had carried these to completion, the mound would have been quartered. However, the appearance of a possibly undisturbed Riv. Bas. Sur. Pav. NO. 12], WILBANKS SITE, GEORGIA—SEARS 139 BALL GROUND r, Zoid SWAMP Go. *CANTON OCTOR 9-CK-5 CARTERSVILLE Y ehe logs ’ rol saute yp ALLATOONA ' RESERVOIR Map 7.—The location of CK-5. core structure stopped excavation part way into the mound in the north, east, and west trenches. All trenches were excavated in 5-foot squares. Usually excavation was done from the top down in 6-inch arbitrary levels, although at some points the appearance of sterile or otherwise distinct strata caused us to drop the use of arbitrary levels and to sack materials according to their relationship to these distinct strata. The stratifica- tion was then recorded in scale drawings of the profiles after a section had been completely excavated. In the first trench (see profile, fig. 6, a) several facts became ap- parent. First, the refuse midden was stratified. Superimposed 140 BUREAU OF AMERICAN ETHNOLOGY [BuLL. 169 midden deposits of slightly different physical appearance yielded dif- ferent types of pottery. The lower levels yielded sherds belonging to the Etowah ceramic complex (Wauchope, 1948), and the upper level yielded sherds diagnostic of the Wilbanks (formerly Savannah) com- plex (see pp. 172-176). There was also a representation, in wash or plow mixed layers, of the Lamar ceramic complex (Kelly, 1938). Alternate layers of sterile yellow river silt and black gumbo clay appeared, the layers angling in toward the mound center. Farthest in, we encountered a layer of soft gray clay (wavy end to profile, fig. 6, a). The nature of this stratum was not understood at all. Since it had not been disturbed by the plow, we attempted to leave it undis- turbed until we did understand its function. Conceivably it might have been some sort of undisturbed core mound. It was clear that it had never functioned as a structure in and of itself for any period of time, since the outer surface retained the lumpy appearance produced by slightly patted down basketloads of soft clay. Other trenches were run out along the edges of the mound, east and west, and then short north-south trenches were cut in from these, through midden and black sterile layers, until they contacted the gray clay again. It was finally determined, through a series of such trenches, that the construction was uniform all the way around, a core construction serving an unknown function; the core was made of soft gray clay, buttressed by alternate layers of black clay and yellow silt. Most of our artifact sample was recovered from the midden ex- cavated in cutting these trenches. After definite stratigraphy was recognized, particularly that an upper layer of softer brown midden was producing pottery stamped with curvilinear designs, a certain amount of work was done entirely in the midden area, designed to increase our ceramic samples from definite stratigraphic context. In particular, a large section of Wilbanks period midden, the upper soft brown deposit, was stripped off. Before cutting in to the center of the mound, an attempt was made to get further information as to its nature by peeling off from its top all plow-mixed material. This paid dividends immediately, since the core of the structure then showed up as square rather than round. A central square of mixed soils became apparent, surrounded by con- centric rings composed alternately of the sterile yellow silt and black clay soils that we had already encountered and sliced through in the trenches. With this information in hand, it seemed highly probable that we were dealing with the remains of a collapsed earth-covered structure, the black and yellow soils representing wall buttressing, the gray clay the inner wall construction, and the square of mixed soils in the center the collapsed roof fill. Several other features showed up after the plow zone had been stripped off. One of these was a 5-foot-wide band of disturbed earth nine ON DREE aa @; Stained oesic eumund yop, snos wil 22 } Sond -sitt woah, sierite ‘ se . Ce Chay, brown and biack compect {mvyib aes ce Shop Coy i" (sevog2) nobbiM fe imcttted fit Dat cape tlie yord | Burned sath, incl. doub fragment#tia yord bno asbbiM) bexim RRGGe Yellow soi data Joni) toivstom signee beyooeb bao bamue rahi “taro. sebaeate aioe ain aagoit 422141 O—58 (Face p. 140) be SN RERRE Se J -_ SS RUS NENG SSR NOS Se Ypipsilddllldanaaarmmnme : SS ee VN 2», La A\ \Ypiow zone , gray humus ee Stained basic Midden (dense) nH Sond-silt wash, sterile 0 5 pee Midden (medium) Dll Clay, brown and black compact Sess Midden (sparse) = Groy clay FEET ea Mottled fill Groy silt And Mixed (Midden and Gray silt) Burned and decayed organic material (incl. thatch) ea Burned earth, incl. daub fragments Yellow silt Ficure 6.—a, Profile, main cut north of earth lodge; b, profile, main cut including roof fill and wall construction south side of earth lodge. Riv. Bas. SUR. 1 — ay Ras WILBANKS SITE, GEORGIA—SEARS 141 about 50 feet long, running north and south. The north end of the band was close to the center of the mound, and the south end extended out past the outer edge of the wall buttresses. This was obviously the filled-in trench excavated by Wauchope (1948) in 1938. We cleaned the trench out, discovering that it had cut completely through the wall of the structure. One end was almost directly over the lower edge of the outermost buttress and the inner end was close to the center of the lodge floor. Inspection of the profiles thus obtained (fig. 6, 0) verified our hypothesis that the mound was definitely the remains of a square, earth-covered structure which had collapsed. Two other intrusions were found, each about 4 by 8 feet, placed well back toward the rear of the collapsed roof fill. They were handled very carefully, with the expectation that they would prove to be in- trusive graves of the Wilbanks or Lamar periods. Final excavation showed one of them to be completely sterile, with its base on sterile basic soil. The base of the other was similarly situated, but it definitely was not sterile. It contained one pool ball, No. six! The writer has not been able to discover who dug these holes originally, but rather suspects that they date from Warren K. Moorehead’s work in the Etowah valley. The fact that they stop sharply on undisturbed basic soil after cutting through the floor of the lodge indicates that they were not the work of a complete amateur. Much as the “six ball” is appreciated, one wonders why it wasn’t the “eight ball.” Once we were sure of our ground, the collapsed roof fill was re- moved en masse, using Wauchope’s cleared trench as a wheelbarrow route from the inner squares and throwing dirt from the outer edges out onto the mound slopes; from here it was removed by a Ford tractor using a scoop. Since it appeared, from the profiles in Wau- chope’s trench, that the outer edges of the square of collapsed roof fill were just over the inner edges of the gray clay buttresses (fig. 6, 0), cutting away the roof earth was started in the center of the floor and carried to the outer edges. In the beginning we stopped cutting at the first. appearance of the gray clay buttresses, but finally it was necessary to cut into these until the molds of the large horizontal logs used as a core for the walls were encountered (see frontispiece, fig. 5). This became necessary because large sections of the wall had collapsed after the logs had rotted. Several inches of undisturbed roof clay were left on the floors as cutting proceeded outward. This clay served to protect the floor for a final overall cleaning. Through the study of the profiles in Wauchope’s trench, and cuts into the gray clay buttress from inside, we discovered that the con- struction of the edifice had started with a square outline of very large logs, averaging 2 feet in diameter. These were propped into a level horizontal position with slabs of rock. After placement, the logs 142 BUREAU OF AMERICAN ETHNOLOGY [Buut. 169 were covered with the gray clay, which sloped off gradually outside but formed an almost vertical wall, 2 to 3 feet high, on the inside (fig. 6, 5, and frontispiece, fig. 5). As we started to clean the last few inches of roof fill off the floor, two sets of objects showed up. First, there were large numbers of rocks scattered around with no apparent arrangement. Second, lying on the debris constituting the lodge floor before the roof collapse, molds of horizontal poles, 4 to 6 inches in diameter were found. These forms were outlined by and preserved by the hard clays that fell from the roof. Most of them contained not only mold, but also large fragments of wood that had not completely decayed. Since these must have been the rafters that were covered by the clay roof after they collapsed, the fill was sliced to trace them out. They are plotted in black on figure 7. It will be noted that with one exception they run north and south. Many smaller fragments, only 2 to 3 inches in length, were also observed. Usually we were not able to plot them because of difficulty in working in the hard clay, which had a tendency to lift in blocks from the soft debris on the last aboriginally utilized floor surface. Without exception, however, these small fragments also ran north and south. Clearly, then, the roof was of the gable type as illustrated in the artist’s reconstruction (frontispiece, fig. 5). There is little doubt that these rafters had fallen almost straight down and had broken as they hit the floor. In a few cases we were even able to tell which particular rocks had caused specific breaks in a rafter. At this time another feature of the method of roof construction be- came apparent. As the excavation walls were sliced down in an at- tempt to approximate the original inner-wall line, holes slanting out and down were observed at a number of points at the top of the gray clay inner-wall buttress. These demonstrated clearly that the rafters had been held in the proper angled position across the top of the large horizontal logs used to outline the structure, and then had been plastered into place with the gray clay (frontispiece, fig. 5). In the few cases where we could check accurately, the location of these holes coincided with the position of rafters on the floor. The function of the rocks on the floor must remain uncertain. Although there were enough of them to construct an encircling bench as was done at Peachtree (Setzler and Jennings, 1941, pl. 6, A), there is no evidence at all for such a reconstruction. I rather incline to the view that some of them were used in piles as extra props under the rafters. Probably in this way they formed a partial inner- wall lining. Apparently scattered rocks are of common occurrence in structures of this sort (Webb, 1938, pp. 34, 41, 42, 44, 101, 153, 154, 155, 162; pls. 18, 19, 108, 109, 112), although their function has never been precisely determined. (ZI ‘d a08q) 8S5—O IF IzzF 422141 O—858 (Face p. 142) Approximate Location ee Fa Intrusion Surface / \ \ \ \ \ \ from ; \ 1 \ \ \ } ae cr / 2 / U / 1 u ' \ \ Ficure 7.—Floor plan of earth lodge. (Scale: Distance between 2 symbols (®) at top of figure equals 40 feet.) basikstpcewsbol aiias tebnaty “wolI-—\ asvord B.glewp2 sugil to Riv. Bas, Sur. ' Leet Pav. No12] WILBANKS SITE, GEORGIA—SEARS 143 The aboriginally deposited material on the floor differed in compo- sition from the midden deposits outside of the building, probably indi- cating considerable difference in the type of usage producing the deposits. There were, of course, large numbers of sherds on the floor, including most of our sample of the ceremonial type, Hiwassee Red on Buff. Most of our pipes also came from here (pl. 44,7 and 2). Ex- cept these artifacts, the bulk of the floor deposit was wood ash, with a fair amount of charred bone. Very little vegetable matter had ever decayed on the floor while the structure was in use. The middens outside, on the other hand, were characteristically a rich black in color, indicating high percentage of decayed organic matter. A final note on the construction of this edifice follows: No vertical posts were used in the walls. This was checked in a half dozen areas, although the total wall buttresses were not removed. A|] post- holes for vertical posts were inside the lodge and inside the buttress walls (fig. 7). These may represent vertical posts used to support the roof, perhaps propping up a ridgepole. Even this is not certain, however, although an inspection of the floor plan (fig. 7) will indicate that it is possible. Other posts, particularly the smaller ones around the edges, may have supported benches, etc. A more complete descrip- tion of the structure, as reconstructed, will be found in the next sec- tion. THE EARTH LODGE Because of structural peculiarities, its dominant position in the site, and relationship to other earth-covered ceremonial structures excavated archeologically and observed during the early historic period, the earth lodge merits special attention. Features of con- struction observed in excavation were described in the previous sec- tion, and are illustrated in figure 7, the floor plan, and in two profile drawings, figure 6. A pen-and-ink reconstruction is included as the frontispiece (fig. 5). Construction of this edifice may be summarized as follows: Rather large logs, averaging 24 inches in diameter, were laid on the clean-swept ground surface. They formed a square approximately 45 feet to a side, with the logs held in position and leveled with slabs of rock. If a ridgepole was used, it must have been put up next, supported by two posts at opposite ends of the structural midline. However, there is no certainty that these vertical support posts were used. Rafters, made of poles 4 to 6 inches in diameter, were placed in the properly angled position, lying across the large horizontal logs with their outer ends on the ground surface and their inner ends touching or crossing in an east-west line that was to be the roof peak. If a supported ridgepole was not used, a pole may have been lashed into place, secured to each pair of rafters as they came in 144 BUREAU OF AMERICAN ETHNOLOGY [Buxn. 169 from the north and south sides of the building (see frontispiece, fig. 5). This would line up the points of juncture, prevent slippage, and aid in taking the thrust out toward the butt ends of the rafters. It did appear that, structurally, no vertical posts were needed to sup- port the roof. The design was so efficient that the thrust of the roof weight was out and down the main axis of the rafters. A light ridgepole lashed into place would serve mainly to prevent slippage. After all the rafters were in place, the buttress of gray clay was built, cementing the rafters into place against the ground and against the horizontal wall logs. The inner side of this buttress was kept smooth and vertical since it served as a low interior wall for the building. Presumably light poles, canes, or reeds were woven between the rafters at about this point in the construction. The rafters at this time would have had sufficient weight of gray clay buttress on their outer ends to support the weight of the workmen. Certainly a solid mesh of lighter materials was made to support the final earth cover- ing. The debris resulting from the decay of this light mesh formed a 1-inch brown layer on top of the artifact-containing debris of the final aboriginally utilized floor. The next addition was yellow river silt, completely sterile, free from organic materials and brightly colored. This formed an em- bankment on the outside of the lodge (figs. 5 (frontispiece) and 6) and covered the roof with a layer about 30 inches thick. When the roof finally collapsed, the portion of the yellow silt which fell inside took on a different appearance from that in the outer embankment. Con- tributing to this difference in appearance were fragments of organic matter, probably the remains of grass roots and pieces of the roof mesh, and some mixture with gray clay similar to that used in the main wall buttress. This clay may have been used as a ceiling plaster, since a fair amount of it was mixed with the silt, in units of various sizes. Obviously it had fallen with the yellow slit, when the rafters gave way under the load. Immediately after the yellow silt, another layer was added, com- posed of tough, dead black clay whose only source is half a mile away across the river. This layer added another embankment to the wall and again extended partly over the roof. Apparently the layer over the roof was only 3 or 4 inches thick, although it is difficult to be certain because the top of the mound was planed off at one point in its history to level it for the erection of a barn. Two more layers were added to the outer walls of the structure, another layer of yellow silt and a final layer of black clay (figs. 5 (frontispiece) and 6), each completely sterile. Owing to the removal of the top of the mound noted above, it is not certain whether or not these layers extended over the roof, although I am inclined to believe Riv. BAs, SUR. — Bry. Bas. Sy WILBANKS SITE, GEORGIA—SEARS 145 that they did not because of certain minor peculiarities in the stratifi- cation observed in the corners of the structure. Quite probably, the interior of the gray clay inner walls was par- tially lined with the flat stones shown on the floor plan. They do tend to be concentrated near the walls, and might have helped to support the rafters. If built into compact benches, they would not have scattered as widely as they did when the building collapsed (fig. 7). It will be noted that no mention has been made of a door. No evidence for one was found. Entrance was either over the buttresses and through the lower part of the roof, or Wauchope (1948) came in through the door with his trench in 1938. The first hypothesis is the most probable, particularly since definite doors were missing in the large-log and small-log town houses of the Norris Basin (Webb, 1988). Apparently, square ceremonial lodges covered with earth were relatively common throughout the Southeast during the Mississip- pian period. There are also structures, apparently related, that had thatched roofs and walls banked with earth. As noted below, the construction of such edifices without using rows of vertical posts to form the walls is decidedly unusual, being paralleled in only one instance. The earliest earth-covered structures in the Southeast, at least east of the Mississippi River, are those at the Macon Plateau site. Fair- banks (1946) reports briefly on these, describing one in some detail. This building, circular with molded clay seats and altar or chiefly seat, used a four-post center arrangement and buttress walls as much as 21 feet thick. No wall posts were used. Other, similar lodges on the site, apparently later, did add wall posts to the basic circular structure. The Macon structure then provides a parallel for our type of construction without wall posts. However, the total Macon structure differed considerably from that at CK-5 in the use of four center posts, a long entrance passage, the circular shape, and the molded clay seats. Since the Macon lodge was built in the Early Mississippi period whereas the CK-—5 structure falls toward the end of the Mature Mississippi period (see p. 136), a lineal relationship may be inferred. Another center for earth-covered ceremonial structures is Ten- nessee. Of the two types reported in the Norris Basin (Webb, 1938), the earlier small-log town houses are related to the Macon structures, the relationship being attested both by structural similarities and by the associated ceramic complexes. The small-log town houses are similar to the Macon structures in the possession of baked-clay “furniture,” usually just one large seat and an altar in Tennessee rather than the elaborate setup found in 4221415813 146 BUREAU OF AMERICAN ETHNOLOGY [BuL. 169 Macon. The buildings at Macon and in the Norris Basin were finally burned for ultimate disposal. They differ in that all the small-log town houses are square, as are all such early edifices recovered archeologically in the Kastern United States except the Macon speci- men. The Norris Basin houses also differ from the Macon building in the lack of a four-post center arrangement, a lack which again allies them to the CK-5 town house and others in the Southeast. As Fairbanks (1946) suggests, this four-post arrangement is a typi- cal Caddo feature, and may thus indicate that Macon drew first from the probable homeland of the earth-covered house. Other ceremonial structures that belong in this Early Mississippi period are those at Hiwassee Island, which belong in the Hiwassee Focus (Lewis and Kneberg, 1946). Although they are generally similar to the small-log structures in shape, wall construction, and in the use of molded clay fireplaces and seats, Kneberg and Lewis do not believe that these buildings were earth covered. There is a relationship evidenced between Macon, Hiwassee, and the small-log town house cultures through the baked-clay furniture, even if earth were not used for a roof covering at Hiwassee. Participation of these three cultures in a common basic culture is also indicated by shared ceramic traits, particularly noded loop handles. On the whole, these earth-covered or earth-walled structures of the Early Mississippi period in the Appalachian region of the South- east differ from the CK-5 structure in the use of clay furniture and in destruction by burning. Then too, all of them, except certain Macon examples, used posts in the exterior walls as contrasted with the CK-5 construction method which utilized earth buttress walls. Ceramic evidence already briefly outlined (p. 184) and to be pre- sented in detail in following sections places the manifestations at CK-5 in the Mature Mississippi period. Thus, we might expect that the earth-lodge structures most closely related to the one at CK-5 would be the large-log town houses of Tennessee (Webb, 1938). These buildings, successors to the small-log edifices in their own area, vary from their Tennessee predecessors in the direction of our Etowah Valley example. The large-log town houses do make use of vertical posts in their outer walls. However, these houses were square, lacked any clay furniture, and were finally destroyed by collapse or were pulled down rather than burned (ibid.). Since there is some evidence for cultural connection, in this general time period, of the Etowah culture with that of the Tennessee-Cumber- land area, and, considering similar features of lodge construction, there seems to be little doubt that the CK-5 earth lodge and the large-log town houses represent mutual participation, with local variations, in a common culture pattern. Riv. BAS. Sur. es pce et WILBANKS SITE, GEORGIA—SEARS 147 Two other contexts for ceremonial buildings might be mentioned here. First is the Peachtree Mound (Setzler and Jennings, 1941). Judging by the descriptions, drawings, and traces of earth-buttress construction in several of the photographs (ibid., pls. 2, 6; 3, 6, c), the writer has little doubt that the basis for the Peachtree Mound was a collapsed earth lodge of the type found at CK-5 and in the Norris Basin. Feature 29 includes most of this structure, although the writer believes that buttress walls should have been added to the description in the Peachtree report. It is seriously doubted however that the reconstruction of the stones found in Feature 29 of the Peach- tree Mound into an encircling bench is correct. The stones as found (ibid., pls. 5, 6; 6, 0) do not appear to warrant such a reconstruction, and the construction of such features from imperishable materials is apparently confined to the Early Mississippi period. There is little doubt, however, that there were later additions to the Peachtree Mound, which may quite possibly have been made to adapt it to use as a substructure mound at some point in the mound’s history. The ceramics from the Peachtree site are a mixture, includ- ing everything from Deptford Check Stamp through Lamar types to the apparently dominant late Check Stamped type (ibid., pl. 35), which may quite possibly represent a 17th- or 18th-century Cherokee pottery type. This latter pottery would belong to the period of substructure mound usage, while Lamar or pre-Lamar types would represent the ceramics of the period of earth-lodge usage. In attempting to relate this Etowah Valley ceremonial structure to obviously related earth-covered structures known from the early historic period, we are troubled by the same fact that caused Webb a great deal of grief (Webb, 1938, p. 222). That is, while our latest archeologically excavated structures are square, with the exceptions to be mentioned, all of the earth-covered council houses or town houses described in the Southeast during the 17th and 18th cen- turies were round. We need not repeat here the often-repeated quota- tions from Hawkins, Bartram, Adair, et al. It is sufficient to point out that round or octagonal earth-covered lodges or council houses were a central feature of villages occupied by Muskhogean- and Cherokee-speaking peoples in the 17th and 18th centuries. Very few cases of round ceremonial structures are known from archeological context. Probably the earliest one is that found on level D of the mound at Hiwassee Island (Lewis and Kneberg, 1946, pl. 19). Although it was not earth covered, this building was in all probability a council house for public gatherings. However, this structure was in use at the very beginning of the Dallas period and was apparently replaced later by square structures (ibid.). Another building of this type was found at the Irene site. This structure 148 BUREAU OF AMERICAN ETHNOLOGY [Bunn. 169 also was not earth covered, but it is almost certainly an example of the type of council house described by early travelers, and is a rather large specimen. However, it belongs in the Irene ceramic period (Caldwell and McCann, 1941, pp. 30-31), and thus it is practically historical, and is not too much help in our present problem. Since most earth-covered structures of the fully prehistoric period are square, except for the very early small ones of the Macon site, and all such structures in the historic or late protohistoric period are round, we may infer that a shift in type took place at the end of the Mature Mississippi level, or sometime during the protohistoric pe- riod best characterized in Georgia by the Lamar Horizon. This shift from the early round or square lodges, rather small, with definite seating arrangements for all of a very small group, or at least for a few individuals in a small total group, to larger structures without the definite seating arrangement and a larger total capacity, culminating in the very large structures of the late protohistoric and early historic periods, has rather interesting social implications. Assuming that all these buildings represent the structures wherein the deliberations and ceremonies associated with governmental proc- esses took place, the specialized small structure at Macon indicates participation in government by a very limited group, the delibera- tions and ceremonies not being available to the general public, not even at the audience level. One individual, for whom a special seat in a commanding position was supplied, must have been dominant. Such a period of autocratic rule, probably supported by supernatural sanctions controlled by the priest-chief, was succeeded by one in which there was less recognition of a paramount place for one individual, no special position or seat being marked out. The total structure was also larger, implying that more people took part in affairs, either as participants or as an audience. Finally, in the latest period of all, structures are large enough to accommodate a very substantial portion of the population residing in a given community. We know historically that prominent individuals, politically speaking, had seats in dominant positions, but the entire community was admitted to, and to a degree participated in, the deliberations of the ruling body and the ceremonies performed for the good of the total society. The total development then, as indicated by these earth-covered build- ings, is from a theocratic government participated in by a small group, to a much more democratic government, in which most adult males at least had a voice, deliberations being open to all. The CK-5 lodge may be placed at the midpoint in this sociopolitical development. A definite group of rulers is indicated, probably a superior class, who at the same time were not completely out of reach of the people. Riv. BAS. Sur. tase WILBANKS SITE, GEORGIA—SEARS 149 The developmental step after this was taken by the Creeks with the development of the square ground, having places for each clan. As utilized by the Creeks, this arrangement allows for participation in ceremonies by the total populace. THE ETOWAH PERIOD Although groups representing three distinct cultures used CK-5 at various times, it is dominantly a site of the earliest of these three, the Etowah culture. The following Wilbanks period is represented by a much less intensive occupation, and the Lamar culture by a lighter one still. There are three units which must be considered in a discussion of the Etowah period and its ceramics. In chronological order the first is phase A, physically represented by the lowest levels in the midden deposit outside the mound (fig. 6, @) Phase B, the upper levels of the midden containing Etowah ceramics, and, finally, the materials directly on the floor of the earth lodge. This last unit is actually a representation of phase B ceramically, but needs separate considera- tion. We shall take these up in chronological order, starting with phase A, the oldest. Percentages and charts used to define the pottery complexes and to illustrate their similarities and differences will be found at appropriate points, supplemented by a comparative chart at the end of this section. Formal descriptions of pottery types not heretofore defined are in a special section at the end of this report. For other types, appropriate references will be found. PHASE A In the first cut taken into the slopes of the mound, from the north side, the lowest stratum containing artifactual materials was physi- cally indistinguishable from higher levels of the same midden. This was discussed in the preceding section and may be observed in figure 6, a. However, since this stratum sloped across the trench in an east- west direction, and since we lacked time and funds to strip out a pure sample, there is some mixture with later materials. The mixture is slight, fortunately, as checked by sherds taken from a number of small refuse-filled depressions into basic soil that were totally within the confines of the trench. The ceramic complex for this earliest period having been determined from this stratigraphically isolated lower level sample, both as to types and styles represented and their relative frequencies, it was possible to group collections from other units with this original sample. In all cases, these other units were from the lowest levels in their particular portion of the site. Apparently, the earliest occupation was scattered, so that while refuse of this period is 150 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 earliest when found, it does not appear in all stratigraphic cuts. The figures given below are based on the total combined sample. Eight pottery types, as follows, are represented in this earliest period of occupation: Etowah Complicated Stamp ** Hiwassee Complicated Stamp Etowah Red Filmed * Hiwassee Red Filmed Etowah Polished Black * Etowah Burnished Plain * Htowah Plain (Wauchope’s Etowah Plain Smooth) ** Sives Plain ** The types marked with an asterisk are defined for the first time in this report. (See Appendix B.) Those marked with a double asterisk were introduced into the literature, but were not formally defined, by Robert Wauchope (Wauchope, 1948). Formal type descriptions are included in Appendix B for these types. Etowah Red Filmed, Etowah Polished Black, and Etowah Bur- nished Plain have not heretofore been recognized, but I have used “Etowah” as the first term in their titles to hold the series together terminologically. It is certain that all of these types occur together as a definite complex during the middle period of Etowah Culture at least, having been recognized by the writer in a number of collections, including collections from the Etowah site itself and from Ball Ground or Long Swamp, Wauchope’s CK-1 (ibid.). The total sample from Phase A was 3,958 sherds, definitely lopsided proportionately in the direction of Etowah Complicated Stamp, 56 percent of the total sample. The plain types together totaled 1,536 sherds, or 39 percent of the total. The situation is even more lopsided when we consider only decorated pottery. Of the 2,395 sherds deco- rated by one technique or another, the types represented are Etowah Complicated Stamp, Etowah Red Filmed, Hiwassee Red Filmed, Etowah Polished Black, and Hiwassee Complicated Stamp. Of the 2,395 sherds in five types, 2,208, or 92 percent, are specimens of Etowah Complicated Stamp. Within the type Etowah Complicated Stamp in this phase we are able to distinguish five stamp motifs, the specific motif being dis- cernible on 560 of the 2,208 sherds. Only those cases are included in which we were absolutely certain of the motif. The variation in the occurrence of the specific design motifs within this pottery type 1s as important in the study of Etowah ceramic chronology as is the variation through time in the occurrence of the several specific types. In round figures, the following tabulation outlines the proportionate occurrence of the several motifs in Phase A. Percentages are of the 560 sherds bearing recognizable motifs (X =5 percent to the nearest 5 percent) : ie Roo WILBANKS SITE, GEORGIA—SEARS 151 Motif: Percent 2-bary, diamond... * — he Eats fl vw ‘ited: Wage ‘ule at) tea iit mae ly v phe 'sreatly - Alt “at pee Bigepe lana i +g ras fy a art BE lt 7 hae eh . ' Lasend sis} tate { sere Cty 0% ot a Me faith ven \: peels : ' . it. a Nhy ai bes 7 ‘ Ferns ys 4 sap: SOV pAOCH odd ea iaaielet MA Cartan Ae eA OEM REE ae a i Oe - BOR “ag , 7 , ‘ A i yh hay LP ! a ¢ \ ns a APPENDIX A LITERATURE REFERENCES FOR FORMAL DESCRIP- TIONS OF POTTERY TYPES MENTIONED IN TEXT Crystal River Negative Painted____----~ Willey, 1949, p. 391. Dallas Negative Painted__-__-----__--_- Lewis and Kneberg, 1946, p. 96 (discussion). Dallas Decorated (Includes Dallas Modeled) se s= ass ae Ibid., p. 105. JOP 3 Oh i ee ee Ibid., p. 95 (discussion). Dunlap Fabric Marked_------------~-- Jennings and Fairbanks, 1940, p. 7. Etowah Complicated Stamp__---------- Wauchope, 1948 (discussion): Ap- pendix B. Htowall incised22=2 22 eu ee ee Ibid. Etowah Plain (Plain Smooth) _---------- Ibid. Halsterd Plaines 2 Sek os eee Jennings and Fairbanks, 19409, p. 3. Hiwassee Complicated Stamp__-------- Lewis and Kneberg, 1946, p. 104. Hiwasseethed ion But. 2 = = Ibid., p. 104. iEiwasseenhed Hilmed 2222255. Sas Ibid., pp. 103-104. lamar Boldsincised&= 22 Jennings and Fairbanks, 1939. Lamar Complicated Stamp_____________ Ibid. Napier Complicated Stamp____________ Jennings and Fairbanks, 1949, p. 8. Pickwick Complicated Stamp__________ Haag, 1939. Savannah Burnished Plain____________ Caldwell and McCann, 1941, pp. 45-46. Savannah Check Stamped_____________ Ibid., pp. 44-45. Savannah Complicated Stamp__________ Ibid., p. 45. Savannah Fine Cord Marked__________ Ibid., pp. 48-44. Tampa Complicated Stamp____________ Willey, 1949, pp. 486-437. 187 a ‘reg Oban senttr er el? pamiel / i at. weno serotoa ae “eirat Seesenemenemierennenitesciiciet _Catotesu gels} lo Satay itor Did vile. haat ice as ibe ee ~- {uobsarmeie RO: itt (ia ales bi he area eel = J a7 0hG? pioattaial Non eamionel iy qh. :(uoiemanelisy Sheet squifodeW |. A xiftioeg DME saan #1 AOL Hampi taniainicest pat tot a GGT windsnt boa eiwal ae. : Mb ag DIT oo AAME-BOL agg DHT 2... Maer Panadtiet han egaianal oo BE =, bitin atasdalet ban eyalisnet ; am 7 LIAL indetoldibive loin’) prerieas debt ag bidt — BP qitdl . qnte bE RB gq bid |. TES ng CRE SNe Liesl, I Tal APPENDIX B POTTERY TYPE DESCRIPTIONS ETOWAH PLAIN: Paste: Method of manufacture: Coiled. Tempering: Sand, often micaceous, rather coarse. Moderate amount. Texture: Slightly sandy, well amalgamated, no distortion or lamination. Hardness: 2.5. Color: Ranges from medium brown through to black. Most sherds are in the darker brown range, interior, exterior, and core. Surface finish: Good to indifferent smoothing on interior and exterior. Ranges from a near burnish to quite sandy, with smoothing tool marks evident in all cases. A roughened, crackly finish is found on some sherds, probably the result of poor drying before firing. Decoration: None. Form: Rim: Flared, slightly to heavily, most moderate flare, ca. 45 degrees. Lip: Usually carelessly rounded, occasionally flattened. Body: Elongate jar, narrows toward round base from moderately slop- ing shoulders. A few plates and open bowls are also included. Base: Convex. Thickness: Rim: 9-13 mm., average 11.5 mm. Body: 9-13 mm., average 11 mm. Appendages: Strap handles, with or without flat buttons, occur rarely. (See plate 40 for handle types. ) Geographical range: As remainder Etowah complex. See Etowah Com- plicated Stamp. Chronological position: As remainder complex. See Etowah Complicated Stamp. Probable relationships: As Etowah Complicated Stamp. ETOWAH COMPLICATED STAMP: Paste: As Etowah Plain. Surface finish: Well smoothed, near polish in many cases on the interior. Exterior apparently treated in the same way before the application of the stamp. Decoration: Complicated stamps, usually drafted and carved with consider- able skill. Application of these stamps ranges from very careful and pre- cise to sloppy with a great deal of overstamping. Only the one motif is carved on a stamp, although in some cases parallel horizontal lines fill in the area of the stamp outside the main motif. Motifs are illustrated in figure 9 and on plates 37 and 38. The commonest motifs in the several 189 190 BUREAU OF AMERICAN ETHNOLOGY [BuLn, 169 ETOWAH COMPLICATED STAMP—Continued periods of Etowah development are as follows (see text for precise data and documentation) : Cowan se ee Ladder-based diamond and line block. Etowah II (CK—5 Phase A)___. 2-bar diamond. Etowah III (CK—5 Phase B)__. 2-bar diamond, line block, and filfot cross. Ktowah Ves2* 22 As in period III, but with very sloppy execution in drafting, carving, and placement. Form: As Etowah Plain. Geographical range: Total complex of Etowah Complicated Stamp, Etowah Plain, Etowah Burnished Plain, Etowah Red Filmed, and Etowah Polished Black, is well known only in the Etowah drainage thus far. However, available surface collections containing sherds of the diagnostic stamped type indicate that it was the dominant pottery type in the Georgia pied- mont during the Mature Mississippi period. Scattered sherds have also been noted in collections from the area around Macon and Milledgeville in the center of the State, which is probably indicative of the southern ex- treme of the types range. However, sherds of Etowah Complicated Stamp were also found at the Irene site. Typical Etowah motifs are there in- cluded in the illustrations as part of Savannah Complicated Stamp (Cald- well and McCann, 1941, fig. 18). Chronological position: In the Etowah valley, the stratigraphic position of the type and complex is between the Woodstock Complex, Early Missis- sippi, and before the Wilbanks-Savannah Complexes, which in turn im- mediately precede Lamar, protohistoric, and early historic. Position in the Mature Mississippi period is confirmed by the presence of ceramic traits shared with other cultures demonstrated to be in the Mature Mississippi Horizon, as Dallas, which participates in the Southern Cult. Probable relationships: Apparently the type developed in the north Georgia tradition of complicated stamping, using angular motifs from Woodstock Complicated Stamp. It, in turn, enters into the ancestry of Lamar Com- plicated Stamp, as that type was made in north Georgia. This is based on stylistic seriation of major motifs used in the stamping (see fig. 9) as well as on known relative temporal positions of the relevant types. ETOWAH BURNISHED PLAIN: Paste: Method of manufacture: Coiled. Temper: Fine sand, small to moderate amount. Texture: Fine, well compacted and amalgamated. Hardness: 2.5-3. Color: Reddish buff to dark brown. Usually many fire clouds are present. Surface finish: Interior well smoothed, although concentric rings left by the smoothing tool are apparent on most sherds. Exterior burnished, probably while pottery was leather hard. Finishing has produced a fine crackle as well as the burnish. Decoration: None on the commonest form, the compressed globular bowl. Some bowls have notched added rim strips, applique human faces, in Dal- las tradition. Shallow open bowls and plates, very rare, have notched rims in a few cases. ay a a= WILBANKS SITE, GEORGIA—SEARS 191 ETOWAH BURNISHED PLAIN—Continued Form: Rim: Direct, unmodified on the normal bowl form. Flared at angle from body in plates. Lip: Usually carefully rounded, occasionally slightly flattened with slight exterior extrusion of paste. Body: Compressed globular bowl decidedly the predominant form. Other forms are open bowls, with or without scalloped rims and added rim strips, bottles with wide or narrow necks, jars, and plates. Base: Convex. Thickness: In compressed globular bowls, 8-11 mm., average 9.8 mm. Variable in other forms. Appendages: Four flat horizontal lugs, semilunar in form, on a few open bowls. Geographical range: As remainder of the Etowah Complex. See Etowah Complicated Stamp. Chronological position: As remainder of the complex. Probable relationships: The major vessel form is too unspecialized to be indicative of relationships. The finish and rare specialized vessel shapes as plates, bottles, and open bowls with added notched rim strips indicate relationships to the Dallas Plain and Modeled types (Lewis and Kneberg, 1946, pp. 80-107) and in turn to similar specialized wares characteristic of the Mature Mississippi period in the area from Tennessee to Arkansas. ETOWAH POLISHED BLACK: Paste: Method of manufacture: Both coiled and modeled from mass of clay. Sherds from bottles and more elaborate forms, possibly including effigies, indicate that some segments were modeled from the mass of clay, welded into place, and then carved to appropriate thickness on the interior. All bottle necks were shaped separately and then welded into place, the welds only being smoothed off on the exterior. Temper: Small amount of fine micaceous sand. Texture: Fine. Hardness: 2.5-3. Color: Core and interiors medium to dark gray. In rare sherds, the core is black. Surface finish: Polished Black. Black, a definite slip in many cases, in other instances may be a deliberate smudging or a thinner paint. Fired on to a well-smoothed surface, and polished, in all cases. Definitely slipped pieces have a higher luster than others as well as greater thickness of pigment layer. Decoration: Black slip, polish. Form: Rim: Variable. Number of elaborate vessel shapes. Lip: Asrim. Body: Mostly rather elaborate vessel shapes. Bottles with wide or narrow necks, bowls with flared rims, sometimes notched or with flat horizontal semilunar lugs, open bowls with added rim strips, and apparently some effigies. A number of sherds are thick, carved on interior, and oddly shaped. Also a few plates and a few small jars. Thickness: Varies with vessel shape and part. 192 BUREAU OF AMERICAN ETHNOLOGY [Bury. 169 ETOWAH POLISHED BLACK—Continued Appendages: Flat lugs, horizontal, semilunar, on open bowls, notched added rim strips. Geographical range: Probably as remainder Etowah Complex. See Etowah Complicated Stamp. Chronological position: Apparently confined to Etowah periods II and III. Probable relationships: Related to black filmed and polished wares in similar shapes characteristic of the Mature Mississippi period. Mound- ville Engraved is the best known of these; names have not been published for others although they are generally known. Range generally is as that given for Etowah Burnished Plain. ETOWAH RED FILMED: Paste: Method of manufacture: Coiling and modeling from mass. As in pol- ished black type; most of vessel is shaped by coiling technique. Other segments, in more elaborate vessel forms, modeled from mass of clay and welded into place. Temper: Small to moderate amount fine sand. Texture: Fine to sandy, depends on amount of temper. Hardness: 2.5. Color: Cores and interiors light buff to medium brown or gray. Surface finish: Interior poorly smoothed, even roughly, in many forms; well smoothed in open bowls, with tool marks eradicated. Exterior well smoothed before red film applied. Decoration: Red slip or red paint, usually the latter, a few sherds have thick, polished red coating, fired on. Form: Rim: Varies with specialized vessel shape. Direct bowl rim is probably the commonest. Lip: As rim. Body: Variety of forms. Definitely represented in the CK-5 collections are bottles with wide and narrow necks and possibly blank faced effigy bottles, plates, and bowls of two types, open and compressed globular. Many elaborately modeled sherds are presumably from effigies of some sort. One small human effigy, with the open mouth as orifice, was recovered by an amateur from a grave washed out in the gully on the site. Base: Variable. Thickness: Variable. Appendages: Vertical lugs on wide-mouthed bottles known in two cases. Geographical range: Presumably as remainder of complex, only known in Etowah valley at present. See Etowah Complicated Stamp. Chronological position: As total Etowah complex. See Etowah Compli- cated Stamp, but is probably confined to the middle periods, II and III. Probable relationships: Considering only formally described types, it is obviously a close relative of Hiwassee Red on Buff, with which it shares vessel forms and red filming. Other than this, to a widespread series of Mississippi red filmed wares most of which have not yet been described. Some of the vessel forms, particularly the plate and the bottle, are indica- tions of partcipation in Mature Mississippi Culture. WILBANKS PLAIN: Paste: Method of manufacture: Coiled. Tempering: Large amounts coarse sand. srt vie WILBANKS SITE, GEORGIA—SEARS 193 WILBANKS PLAIN—Continued Paste—Continued Texture: Coarse and sandy, lumpy fracture. Hardness: 2-2.5. Color: Reddish brown through medium gray. Most sherds fall in lighter reddish brown tones. Surface finish: Well-smoothed interior, rarely shows tool marks except as slight textural differences. Exteriors are sandy, apparently were not well smoothed after partial drying as were most interiors. Decorations: None. Form: Rim: Moderate flare, often with sharp angle on interior at shoulder (see profiles below). Lip: Carelessly rounded or squared. Body: Hlongate jar, rounded but distinct shoulders, in many cases with definite angle on interior setting off neck region. Base: Convex. Thickness: Rim: 11-13 mm., average 12.33 mm. Body: 10-14 mm., average 12.25 mm. Appendages: None. Geographical range: Only known in Htowah valley. Chronological position: At CK—5, Wilbanks strata is over Etowah and under Lamar. Probable relationships: See Wilbanks Complicated Stamp. WILBANKS COMPLICATED STAMP: Paste: As Wilbanks Plain. Surface jinish: As Wilbanks Plain. Decoration: Complicated stamping covers the entire vessel exterior except the base. As a rule, one simple motif is used on a stamp, without addi- tional fill elements to take up space not used by the main element. With one exception, the known motifs use three to four lands plus a solid central dot, the exception being the rather elaborate “elongate U with cross bars.” Lands and grooves are broad, lands averaging 3 mm. wide, grooves slightly wider. Stamping is deep and definite, but with much overstamp- ing. Motifs known to date are (pl. 39): CLO] ee oe a se 15.8 percent at CK-5. Bull’s-eyen es Seas ee 44.1 percent at CK-5. CE Te Pe el NES TO Ste 32.4 percent at CK-5. Quatrefois Sees ewe ee er 5.9 percent at CK-5. Form: Rim, lip, body, base, thickness as Wilbanks Plain. Appendages: Semilunar, flat horizontal lugs just under the lip in two cases. Geographical range: Only known in the Etowah valley. Chronological position: As Wilbanks Plain. Probable relationships: A Late northern extension in the development of complicated stamping using curvilinear motifs which has the southern part of Georgia as its center. The closest relative is Savannah Complicated Stamp on the basis of vessel form, overall stamping, stamp size, and simplicity of stamp design. The type develops from Harly Swift Creek in a line running through the Swift Creek style and then through Kolomoki Complicated Stamp. The tradition culminates in Lamar Com- plicated Stamp, as that type is found in south Georgia. 194 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 SIXES PLAIN: Paste: Method of manufacture: Coiled. Tempering: Finely crushed shell, moderate amount. Texture: Slightly contorted, lumpy fracture. Hardness: 2.5-3. Color: Exterior surface medium to dark gray. Interior surface medium brown to dark gray. Cores gray. Surface finish: Well smoothed, near burnish interior and exterior. Tool marks are obvious. Decoration: None. Form: Rim: Moderate to strong flare from abruptly constricted neck. Pulled out to four rim points in several cases. Lip: Rounded. Body: Uncertain. Rather strongly rounded shoulders, probably globular jar. Base: Probably convex. Thickness: Average 11 mm. Appendages: None. Geographcal range: Known only from CK-5. Chronological position: Phase A, Etowah period ITI. Probable relationships: Local representatives of the widespread Middle Mis- sissippi shell-tempered plain wares. The four pointed rims, plus the affili- ations of other Phase A material at CK-—5, point to the Dallas Culture of Tennessee as the home of the most closely related type (Lewis and Kneberg, 1946, pp. 80-107). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 37 Etowah Complicated Stamp. / and 2, Two-bar diamond motifs. Etowah periods II and Ill. 3, One-bar diamond. Etowah periods II and III. 4, Three-bar diamond. Eto- wah periods II and III. 5, Ladder-base diamond. Etowah periods I and II. 6, One- bar cross diamond. Etowah periods II and III. All rim sherds are normal form, mod- erately flared. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 38 Etowah Complicated Stamp. J, 2, 4, Line-block motif. Etowah periods I-III. 3, 5, Filfot cross, Etowah type. Etowah period III. Rim sherds are normal form, mod- derately to strongly flared. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 39 Wilbanks Complicated Stamp. J/, 2, 3, Bull’s-eye motif. Overstamping about average for this type. Normal rim form for type on /. 4, 5, Quatrefoil motif. 6, Scroll motif. 7, Elaborate motif. Elongate U with cross bars. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 40 | Sixes Plain, shell-tempered trade ware, Etowah Plain, and Etowah Incised. /, Sixes Plain. Four-pointed rim sherd. 2, Strap handle, shell-tempered trade ware. Heavily shell tempered. 3, Strap-handle fractures on Etowah Plain rim sherd. 4, 5, Etowah Incised. Strap handles with buttons. 6, Etowah Incised. Plain strap handle. 7, Etowah Incised. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 41 = Plate and bottle rims. /-4%, Plate rims. Etowah Plain. 5, Rim from wide-mouthed bottle. Etowah Plain. 6, Rim from straight-necked bottle. Etowah Burnished Plain. 7, 8, Etowah Plain blank-faced effigy sherds. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 42 : Minority types. /, Pinched rim. Etowah Burnished Plain. 2, Scalloped rim. Etowah Polished Black bowl. 3, Notched added rim strip. Etowah Plain bowl. 4, Etowah Burnished Plain rim. 5, Engraved Polished Black sherd. 6, Human face adorno, probably from bowl. 7, Fragment of pottery dog or bear effigy. &, Fragment of pottery dog or bear efhigy. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 43 Painted types. /, Hiwassee Red Filmed. Body sherd with small vertical lug. 2, Etowah Red Filmed. Irregular plate (?) rim. 3, Etowah Red Filmed. Narrow jar or bottle neck, small vertical lug. 4, 5, 6, 7, Hiwassee Red on Buff (4, upper surface, plate rim; 6, normal bow] rim). BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 44 6 Artifact types. J and 2, Pipes. Etowah period. 3 and #4, Chisels. Etowah period 5, Celt butt. Etowah period. 6-8, Pottery disks. Wilbanks period. BUREAU OF AMERICAN ETHNOLOGY BULLETIN 169 PLATE 45 Artifact types. J, Ladle. Wilbanks period. 2, Pottery bead. Etowah period. 3, Fragment of stone napkin ring earspool. Etowah period. 4, Stone earspool with Southern Cult motif. From grave near earth lodge. 5 and 6, Bone awls. Etowah period. 7, Perforated mussel shell. Etowah period. SMITHSONIAN INSTITUTION Bureau of American Ethnology Bulletin 169 River Basin Surveys Papers, No. 13 Historic Sites in and around the Jim Woodruff Reservoir Area, Florida-Georgia By MARK F. BOYD 195 422141—58——_17 Aa: Ne 1 hi Bynes, a ve ee elisha oe rel ly phe edgrtramee Spee 1 a lp ate 7 , Shy emt fears eateuite by ner ry rhe eons a aa * ry marc Tak if j , : ¢ ' Pe anna Be eeneereieey ‘wuiwoeperntall a Te ee ate 4 Nola Rian tdi AY wanes ! = oe ee ee || . Selsd cai oy - at on yeaa ariel nian vi ae i “ Rashoo yt mt ith) finite fae at | see Hod format yHPOn 2 ‘ aM sl = ee yd aban a pith canon vee: colitis yeas mes oer os sor our CONTENTS PAGE UG OO NEYO: KO 1 ee Ln eae aac eae ee pean eeu aUEL: 9s 8 2) ZR ae 199 Part 1. Historic sites that will be submerged in the Woodruff Reservoir__ 204 BieconchanimMico7s Wesel vaAwOn e291 sas ae em ge he eee at 204 IE MeO WH Diocese ays els eg ear he ret ate epee ences a aol hoch rs 209 Sabacola and Cherokeeleechee’s Town and Fort___________________ PAN AL Homigecolie 22 22S = Be ee ee ee 20: a ae 221 Part 2. Historic sites adjacent to the Wcodruff Reservoir_____._________- 252 Bikanachstte or Red Ground 22 oo 2 ae he 252 Pypurie rl eo a Re Ek ea a te Oe a) el 257 INICOHS BONG pOSt=...s220 2.8 2 Bee ee 260 Eilicotizs: Observatonyess..2 4. te Seles ae oie Ne 272 Winited estates) Arsenate. 22. ars Ae Oe 276 Canmipritecovery oto = see ere = BO Pes oer ee Ae ty Ae 287 Okiahimec'or'Coxmpur. 22-0 pevet ts fe Uy te 290 piwik ow, (OA pee SPOR 8 ed og LE A Ne OE edi a glk Eee lee 291 [era aI Cele oe: See ed Se | Res Oe ae See Ie eRe fe Seah Se Ce Ee 298 IBUTECSAOC UTES COW S248 Je. oo vote ae 300 RECOMMMEREAUONS 6 =: 4° Bae fee Ree Coe 3 ee ee ee 304 ACKHOWICEETMCHUS: | = 455 ew pee eet SY oe ah Ses ge 505 Pibboprap by 20) hess" ef ih Dares eee ok ek ey ee ven ae im ad, hele 306 erLORrAp hy 5225 ses oh eS Re DY ty of aa eae SE pit, A Beep a 313 ILLUSTRATIONS PLATES FOLLOWING PAGE 46. Platrot Miconchatimico’s Reservation.-.-...-.=.-2 see 2225-254. 314 47. ‘Village Indien sur l’Appalachicola.’’ (After Castelnau, 1842.)_______ 314 48. Plat of certain Georgia Land Lots, Districts 14 and 15_____-________ 314 49. a, Profile of midden, Fowl Town No. 2 on Chattacoochee River bank. b, Turkey patch opening before clearing. c, Turkey patch after clearing, operations 22h Me 032.2 88 ee St Se ee 314 50. Plat of certain Georgia Land Lots, Districts 15 and 21______________ 314 51. a, Fort Scott gun monument in original location. b, Camp Recovery gun monument. c, Fort Hughes gun monument. d, Florida State Hospital: view of the Arsenal building 2202240 2s. Ye 522 2 314 52. a, Fort Scott site after clearing. 6, Florida State Hospital. View of DRE Maa zine a... oes se lot eelo Ste ee ee 314 53. Region about the fork of the Apalachicola River. (Purcell-Stuart HUUSRENOT L dO) es eat a ee cm See 314 54. Plat of the reservation of the United States Arsenal at Chattahoochee, 1 eae A pe ca pal PR es te RGR A ae ae) eas Bea oe 314 55. ‘‘Arsenal 4 Mountvernon.” (After Castelnau, 1842.)___.____________ 314 TEXT FIGURES 10. Plat of archeological reconnaissance in the fork of the Apalachicola | 0 (2) ON 2 pC oe BE NS oe Np an ne ls YRS! ee ed 212 11. Reconstructed ground plan of Arsenal quadrangle..________________ 277 MAP 8. (Frontispiece.) Area adjacent to the Jim Woodruff Reservoir________- 198 (Butt. 169 BUREAU OF AMERICAN ETHNOLOGY 198 ‘NOAIOSAY YNIpOOM wif oy} 02 Jusoelpe voly—'g dvjJ YIOANISIY 44NYGOOM WIT OL LN3DVF0V ONY Ni S3LIS SINOLSIH ey a ae a Om J90INENIVE ~ .S20UNG ‘Ol S3HONK LYOJ 6 / FINNIVINO 2 AYFAIITY AWYD 9 WHISHY S$ “SHO $§,4L091TIF LSO0dLNO $.TIOIIN E SOTUVI NYS Z PLLVHIVNEHF “1 ANFIIVEOY SILIS “HaAId = VIODIHIYIVdY s e a = Wh dd¥ ='M HY WLE'S = 19 000% = KOMI INO Wf, +0 £4095 4407 0 ~ VI0ovYeVsS “D z NMOLIMOZ “8 ‘SIY S,ODINILVHINOII “VV ‘NI SFLIS pote ne , o WBNS HISTORIC SITES IN AND AROUND THE JIM WOODRUFF RESERVOIR AREA, FLORIDA- GEORGIA * By Marx F. Boxp INTRODUCTION The Apalachicola River is wholly a Florida stream, formed at the western extremity of the boundary line between Georgia and the colonial Kast Florida by the union of the Flint and Chattahoochee Rivers. The Flint is the shorter and more pellucid member of the union, while the Chattahoochee has greater volume and marked tur- bidity. “Rugged” is not a term that with strict propriety can be applied to Florida topography, but if understood to be used in a relative sense it may be said that the terrain about the confluence is perhaps the most rugged to be found in the State. Swinging from west to east in an arc with a northern concavity is a height of land that originates in Florida west of the Chattahoochee and, extending into Georgia, curves northeastward east of the Flint at least as far north as Camilla, which at one time may perhaps have afforded an impediment to the passage of the waters to the southward. Certainly it continues to be responsible for the deflection of the course of the Flint to the southwest as it approaches the point of confluence. After their confluence, the conjoined rivers pierce the height of land. The terrain of the Georgia counties lying between the rivers, Seminole and part of Decatur, is low and prairielike, with a gentle slope from north to south. The west bank of the Flint from below Bainbridge to its mouth, and the east bank of the Chattahoochee below Early County to the confluence, are devoid of marked elevations either along the immediate riverbanks or at a short distance back therefrom. This is in marked contrast to the east and west banks of the respective rivers, where their flood plains are sharply delimited by relatively high bluffs. The Jim Woodruff Dam is located on the Apalachicola River about 1,000 feet below the confluence, and will be about 6,130 feet in length, 1 Revised manuscript submitted August 1956. Original report submitted to Region One office of the National Park Service in December 1953 and accepted and approved in March 1954 by the Regional Director as fulfilling the agreement between the National Park Service and the Florida Historical Society. 199 200 BUREAU OF AMERICAN ETHNOLOGY [Bunn. 169 and except for its extreme eastern extremity, lies wholly within Flor- ida. The normal pool contour or waterline of the lake will be at 77 feet above sea level, with provision for an additional surcharge con- tour 2 feet higher, which will provide a maximum head of 33 feet. It is constructed for the dual purpose of aiding navigation on the rivers above and providing electric power. When the combined waters of the rivers are held back by the dam, they will spread out over 37,500 acres, creating a lake with a highly irregular shoreline, calculated at approximately 243 miles in circumference (frontispiece, map 8). The slack water created on each river by the dam will extend much farther upstream than the area of flooding. It will be evident 52 miles above the dam on the Chattahoochee and for 47 miles on the Flint. Regarding the upper limit of the pool as marked by the limit of slack water, this on the Chattahoochee will be encountered a short distance above the Henry County—Houston County line in Alabama, but the river southward will remain within its original banks as far as the Alabama—Florida line. The upper pool limit on the Flint will extend a few miles north of the Decatur County line, between Baker and Mitchell Counties. For 5 or 6 miles north of Bainbridge there will be a narrow, irregular, and disconnected belt of flooding, beyond which the stream will remain within its original banks. The only tributaries of consequence to be affected are Fish Pond Drain and Spring Creek, both in Georgia, which in their lower courses will disappear. It is likely that before the deforestation inevitable in the agricul- tural development of woodland, the Chattahoochee waters were much less turbid. This stream, however, has always been renowned for its floods. The waters of the Flint, except when in flood, are relatively clear, and the stream receives much accession to its volume from nu- merous springs on its banks from Albany downward.? The banks of the Chattahoochee in the vicinity of the fall line and below were occupied in prehistoric and historic times by numerous towns of the Lower Creek Indians. Abundant evidence of proto- historic occupation is found on all high points along the banks as far as the fork. Many of these sites are revealed by ancient shell middens exposed on eroded river banks, which indicate that these waters were much less turbid at the time when these mussels thrived. Although prehistoric aboriginal settlements were numerous on the Flint, in historic times for reasons now unknown, that river was not so extensively favored. Historic villages appear to have made their 2Since the reservoir behind the Woodruff dam has been filled, the water of the lake exhibits the turbid appearance of that of the Chattahoochee River, even up the valley of the Flint to the neighborhood of Bainbridge. Capt. Henry S. McCord, USAFR, senior instructor at the Southern Airways Contract School at Bainbridge, Ga., states that from the air, the springs along the Flint stand out strikingly against the brown silt-laden water, as clear blue spots. Riv. Bas. Sur. 7 OT — Biv ke is) JIM WOODRUFF RESERVOIR AREA HISTORY—BoYD 201 appearance along the Flint about the middle of the 18th century, as colonies from the older towns on the Chattahoochee. Events in the region of the confluence have played an interesting even though minor role in the history of the Southeast, particularly in colonial times. The rivers themselves, and the footpaths on either bank of the Chattahoochee, afforded ready communication between the aboriginal towns and villages on either bank. After Spanish penetra- tion into Apalachee, a course of overland traflic from the southeast intersected the river path, and occasionally a rare Spanish party pene- trated to the westward of the Apalachicola. After English intrusion on the Chattahoochee at about the fall line, the development of Spanish intercourse received a check. When Florida was united to the British dominions there was a revival of intercourse through gulf ports, in particular St. Marks, and after the restoration of Florida to Spain, the principal traffic flow continued on a north-south axis. The periods of Spanish or British intrusion were interludes of short duration and did little to alter the face of the region, as it con- tinued remote from the occupied and settled frontier, and relatively speaking, for extended periods it was not heavily populated by the Indians themselves. Prior to the delimitation of the international boundary line, the Indians of the region had things pretty much their own way, and successfully played off the Spanish and United States Governments against each other. Creek power, however, waned after the Treaty of Fort Jackson in 1814, which resulted in the cession, but not the occupa- tion by settlers, of the Georgia part of the region. Probably an unan- ticipated consequence of the treaty was the occupation of the region on both sides of the line by the survivors of the defeated hostiles, which was expressive of its remoteness. Even after the cession of Florida to the United States, development of the region was slow, and it was not until the expulsion of the Creeks from Georgia in 1826, and the establishment of Columbus at the fall line on the Chatta- hoochee in 1828, that development became accelerated. A thriving trade arose with the ocean ports of Apalachicola and St. Joseph by means of steamboats, the first of which reached Columbus in 1828. Unless the river was low, in which case the boats could not ascend above Kufala, the navigation to Columbus was open the entire year. The Flint was irregularly navigable as far as Albany, but it never supported steady traflic above Bainbridge or Newton. Reliable trans- portation stimulated settlement along the waterways, and steamboat landings were opened about every 3 to 5 miles on either bank. The Civil War interrupted all except local river traffic within the Con- federacy, and subsequently the extension of railroads to the river towns caused an increasing diversion of traffic, which became com- plete with the introduction of motor highway transportation. 202 BUREAU OF AMERICAN ETHNOLOGY [BULL. 169 It appears unquestionable that since the Civil War there has been a gradual withdrawal of population from much of the territory close to the river, for which a variety of factors, impracticable to consider here, were responsible. Relatively little of the land that will be inundated was actually in recent cultivation. Most was cut- over and in second growth, although a little near the fork appeared virgin. Consequently the preliminary clearing operations have been unusually extensive. In the following sketches, 14 historic sites will be discussed. Only 4 of these will be submerged, and consequently only these 4 come within the scope of the responsibility of the National Park Service. It appeared desirable, however, to consider a number of others that will not be flooded, not only because of their own intrinsic interest, but because 3 at least, namely, Camp Recovery, Fowl Town 1, and Fort Hughes, are tributary or secondary to Fort Scott, which will be submerged. The following sites, which will be submerged, are distinguished on map 8 (frontispiece) by consecutive capital letters, viz: A. Econchatimico’s Reservation B. Fowl Town 2 C. Sabacola and Cherokeeleechee’s Town and Fort D. Fort Scott Adjacent to the impounding, and distinguished by consecutive numbers on the map, are the following: . Ekanachatte or Red Ground . San Carlos . Nicolls’ Outpost . Ellicott’s Observatory . United States Arsenal Camp Recovery Oklafunee or Coxpur . Fowl Town 1 . Fort Hughes 10. Burges’ Town CHOARDARYNHE The foregoing are regarded as historic sites because of available documentary records, either written or printed, which preserve a greater or less amount of information revealing their significance. Certain of these, such as Ekanachatte and Oklafunee, are purely aboriginal. There are numerous other aboriginal sites in the area that are ignored because of the absence of historical data regarding them. While some of these may be protohistoric the majority are prehistoric, and their investigation and description pertained to the archeologist. There are two sites deliberately omitted from this report. The first is that of the Confederate Ship Yard on the Chattahoochee River, where two ironclads were constructed during the Civil War. Bry Ris. Sh" JIM WOODRUFF RESERVOIR AREA HISTORY—BOYD 203 A position on the east bank of the Chattahoochee south of Saffold bearing this name is shown on sheet 8 of the 1943 aerial mosaic map prepared by the United States Engineers. The authority for this identification has not been discovered. It was thought it might have been derived from the plat of the 1872 survey of that river by the United States Engineers, when there is every reason to believe the surveying party must have included a number of men intimately familiar with the river who could have identified the site from per- sonal acquaintance, and when vestiges of the installation likely were still present. This plat, however, is not discoverable in either the Mobile or Washington offices, and it is believed to have been lost when the Mobile office was destroyed by a hurricane in 1916. Al- though the site was visited, no remains were discernible, although it was alleged that a line of piling parallel to the bank is revealed at low water. The site is situated about one-half mile south of U. S. Highway 84, where the river pool will be confined within the existing banks. No living local resident could be found who recalled the significance of the site. The second site omitted is a river landing on the east bank of the Flint River about 1 mile below Fort Scott, at or just below the convexity of the river bend. It bore the name of British Burial Ground Landing. The earliest utilization of the name encountered is on an anonymous list of Flint River Landings published in 1901. The existence of this landing is largely forgotten, and of the few persons who could recall the name either expanded as given above or contracted as British Landing, none could afford an explana- tion for its application. Its distance from the fork would appear to preclude any connection with Nicolls’ Outpost. The general area of the landing was not visited until after clearing operations were com- pleted, when nothing significant was discoverable. It remains an intriguing mystery. Early Indian paths and pioneer roads are shown on map 8 (frontispiece) in solid lines. The position of the path along the west side of the Chattahoochee, and that of the Pensacola-St. Augustine road are based on the Purcell-Stuart map of 1778; the others are from the 1820 plats of the surveys of the Georgia land districts. Railroads and modern roads are shown in dotted lines. Although one or more visits were made to every site, with the exception of the Arsenal, the only one on which tangible structures remain, it has been impossible to pinpoint the position or extent of the sites with exactitude. Indeed, the position of Fowl Town 1 has so far not even been approximated. The best that for the present can be said for this site, is that it probably lies somewhere along a 3- to 4-mile are having a radius of about 3 miles, centering on Fort Hughes. Local memory of their significance has disappeared as com- 204 BUREAU OF AMERICAN ETHNOLOGY [BuLt. 169 pletely as their vestiges. The prosecution of the fieldwork con- sequently gave rise to a deep sense of frustration. This, however, is to be expected when dealing with structures erected from easily perishable materials. The sites are systematically discussed in the following 14 sketches under the headings: Location, Identification, Condition, Authenti- cation, and Interpretation. PART 1. HISTORIC SITES THAT WILL BE SUBMERGED IN THE WOODRUFF RESERVOIR ECONCHATIMICO’S RESERVATION This was the northernmost of three small Indian Reservations located on the west side of the Apalachicola and Chattahoochee Rivers, which were created by the terms of the Treaty of Moultrie Creek in 1823, and was assigned to Econchatimico. The latter is identified by his title, meaning Red Ground (Town) King, rather than by his name, which is unknown. Location.—On the west side of the Chattahoochee River in Jack- son County, Fla. It occupied part of fractional township 5 North, range 7 west, comprising all of fractional sections 16, 21, and 28, and parts of fractional sections 9 and 33, all lying along the riverbank, and parts of whole sections 8, 17, 20, 29, and 382, comprising alto- gether less than 4 square miles. Identification—From the plat of the above-described fractional township in the office of the Field Note Division, State Department of Agriculture, Tallahassee (see pl. 46). The reservation was laid off by Col. James Gadsden, in accordance with the terms of the Treaty of Moultrie Creek (1823) in the spring of 1825, and the fractional town- ship which embraced the reservation was subdivided into sections (outside the reservation) in the same year. After the reservation was relinquished by Econchatimico, its interior was laid oif to co- incide with the interior divisions of the embracing township in 1848. This plat shows that at the latter date the extensive Indian “old fields” were already occupied by squatters. Colonel Gadsden’s field notes of this survey cannot be located in the National Archives. Ownership.—With the exception of about 80 acres, the entire area of the former reservation has been purchased by the land acquisition office of the United States Engineers, and ownership is now vested in the War Department. The greater part lies below the 78-foot contour, and will be flooded. Condition.—Prior to the clearing operations, the entire tract was either in woodland, abandoned fields, or pasture. Several knolls of red clay will not be submerged. Riv, Bas. SCR. JIM WOODRUFF RESERVOIR AREA HISTORY—BOYD 209 Authentication—The duration of the occupation of the reservation, and the position of its boundaries are well documented. The area has been well explored archeologically by Ripley Bullen (1950), who found numerous sites indicative of an extended aboriginal occupancy, which he numbered for identification. Two sites, numbered 16 and 21, could not be correlated with any recognized archeological period. Two others, Nos. 25 and 31, were identified as of Weeden Island period, and one, No. 22, as of Fort Walton period. One, No. 25, presented sherds of both Weeden Island and Lower Creek types, while three, Nos. 28, 29, and 30, presented sherds of the Lower Creek types. The last clustered in the vicinity of an old river landing known as Port Jackson, and although too far north from the south line of the reser- vation for the approximate position of Econchatimico’s residence, nevertheless probably mark the site of his village of Tock-to-ethla, appropriately meaning “River Junction,” from the Creek Etoh, mean- ing “together,” and Wethlocko, meaning “big water” or “river” (Simpson, 1956). Interpretation—In the later interpretative section treating of Ekanachatte, it will be shown that by 1821 Econchatimico and his followers had abandoned that site and moved lower down the river to a position about 10 miles above the fork, the name of their settle- ment being variously given as Tock-to-ethla, Totoawathla, or Toto- withla, and that Econchatimico secured from the United States Com- missioners at the Moultrie Creek council on September 18, 1823, the concession of a personal reservation on the west bank of the Chat- tahoochee River. The limits of the reservation, as described in the special article appended to the treaty, were as follows: Commencing on the Chatahoochee, one mile below Hconchatimico’s house; thence up said river for four miles; thence one mile west; thence southerly to a point one mile west of the beginning; thence east to the beginning point. The United States promised to guarantee the peaceable possession of the reservations to the chiefs and their dependents only so long as they continued to occupy, improve, or cultivate them, and in case of abandonment, they would revert to the United States. The chiefs were empowered to select the individuals who should remain with them and were required to submit a list of the names of these to the Superintendent of Indian Affairs, and it was agreed that none other should be received or permitted to remain without permission of the Superintendent. Voluntary removal to the southern reservation would be permitted, and the United States reserved the right, in the event of misconduct, to compel their removal to the southern reser- vation. Subsequent to the portentious treaty of Payne’s Landing in 1832 to which the Apalachicola bands were not parties, separate negotia- tions were conducted with these bands for the cession of their reser- 206 BUREAU OF AMERICAN ETHNOLOGY [Buut. 169 vations. Col. James Gadsden, United States Commissioner, could not get Econchatimico and Mulatto King to participate in the nego- tiations which resulted in Blount’s relinquishment of his reservation (the lowermost) in 1833, but after further conferences with Mulatto King and Econchatimico, he secured their marks to separate treaties at Pope’s, Fayette County (a short-lived dismemberment from Jack- son), on June 18, 1833. By these treaties they agreed to surrender to the United States all their interest in their reservations, in return for which the Government would allow them to exercise either of two options. In any case the United States would, within 3 years, cause to be surveyed off three sections of land embracing their im- provements, and patent it to them. Within this period of 3 years the Indians might dispose of the land conveyed to them at private sale and emigrate, in which case the expenses of removal would be at their personal cost; but should they elect to remain, the protection afforded them by the United States would be withdrawn, and they thereafter would be subject to the territorial laws and government. Alter- natively they could become parties to the treaty of Payne’s Landing, and receive $3,000 for the land relinquished. The treaties were ratified by the Senate, April 12, 1834. Despite the circumstance that Econchatimico and his followers ap- pear to have conducted themselves with all possible decorum, a series of events finally took place to disturb their tranquillity and threaten their safety. Econchatimico possessed several Negroes whom he had inherited under Indian law many years before Florida was acquired by the United States. This aroused the cupidity of a prominent planter of Jackson County. In 1833 a bill of sale was produced, allegedly executed by a relative of Econchatimico as purported owner of these slaves, by which title to 10 of them with their increase was transferred to a Creek Indian who shortly thereafter, for a consideration, resold the slaves to the planter. The latter instituted action for an attachment against Econchatimico in the Superior County Court. This move, coming to the attention of the acting Indian Agent, was brought to the notice of the acting Territorial Governor, John D. Westcott, who referred it to the Secretary of War. The Secretary issued instructions that a judicial proceeding should be instituted before the United States District Judge. Judge Cameron halted the proceedings while he instituted an investigation of the intricate case, which required 2 years for completion, and found that the planter had no shadow of title to the slaves, who rightfully belonged to Econchatimico. Be- fore this opinion was delivered, however, the planter voluntarily dismissed his attachment, but it was soon known that he had sold his claim to other persons who apparently were professional slave stealers from Georgia and Alabama. Hostilities having openly broken out Riv, Bas. Sc. JIM WOODRUFF RESERVOIR AREA HISTORY—BOYD 207 with the Indians of the peninsula and with the Creeks, it was easy for these slave stealers to alarm the whites living adjacent to the reservation by spreading reports that Econchatimico and his band meditated mischief. Their neighbors were thus induced to enter the reservation and disarm the Indians. While they were in this defense- less condition, the slave stealers appeared with accomplices on March 10, 1836, assaulted Econchatimico, and forcibly seized the Negroes and removed them from the territory. Although their identity was known and they were promptly indicted by the grand jury, these slave stealers were never apprehended nor were the slaves recovered. In view of the hopeless outlook, Econchatimico futilely petitioned Congress for indemnificaton (Message: H. Doc. 271; 24 C.-1:24). During the period of the lawsuit described, Econchatimico sought the advice of Gen. Wiley Thompson, the Agent, in 1834, as to the best course to free himself from the annoyance caused by lurking persons who sought to kidnap the Negroes. Thompson pointed out that the distance of the reservation from the Agency made it difficult to provide protection, and suggested that Econchatimico remove to the vicinity of the Agency. He offered him the privilege of using the plantation connected with the Agency to make a crop. Econcha- timico appeared interested, and doubtless had General Thompson found the time to press the matter, Econchatimico would have removed that year, as the project received the approval of the Office of Indian Affairs. The Indians along the Apalachicola appear to have been the only Florida Indians interested in the educational opportunities provided by the treaty. Econchatimico sent several boys from his band to the Choctaw academy in Kentucky. Although Econchatimico appears to have been convinced that emi- gration offered the only solution for an escape from the pressure the Indians experienced, his actual steps in this direction are obscure. That he was friendly to the United States in 1836 appears certain, as in his petition to Congress of April 2 he stated that he had sent some of his warriors to aid the United States against the lawless Seminoles; this was verified by the former subagent to these Indians, William 8S. Pope, who stated in April 1836, that actually 11 warriors had left to be with the troops. These recorded dates indicate that in 1836 Econchatimico was actually residing on the reservation. In supporting Econchatimico’s petition in 1836, former Governor Duval stated that the chief was an old man, now destitute of assistance to cultivate his land, and must with his family suffer want if adequate remuneration for his loss was not speedily made to him. The Comte de Castelnau (1948), who passed up the Chattahoochee in 1838, mentions passing the Indian village of Hitchetan (pl. 47) 208 BUREAU OF AMERICAN ETHNOLOGY [BuLt. 169 on the west side, the chief of which, Hcouchatemico (sic=Econcha- timico), is also called king of the red lands. He further states that Econchatimico was an old man, bent with age, who had had his nose and ears cut off for the crime of adultery. The source or derivation of the name “Hitchetan” is unknown, but it is evident that the name “Tocktoethla” had fallen into disuse about the time the reservation was created. It is likely that in the early years of the war with the Indians of the reservation down the peninsula, the Government overlooked the agreements reached with the Apalachicola Indians. This oversight, however, was corrected in 1838 when on July 18, one Daniel Boyd was appointed to supervise and conduct their emigration. He made an appraisal of their improvements, to which he gave a valuation of $3,000. An effort was made to secure payment of this amount with Congressional sanction, from the sum of $15,400 due the Seminoles proper under the treaty, which was denied. These Indians did receive payment of the $3,000 due under the terms of their cession treaty on the day of their departure from the reservation (House Rep. 746; 27 C.-2: 5). The final history of these Indians is soon told. In a report to the Adjutant General, dated Tampa, July 1839, Bvt. Brig. Gen. Z. Taylor related (Sprague, 1848, p. 221) : With a view to compel the emigration of the Appalachee [sic] Indians, who had engaged to leave the country on the 20th October, and who had made some objections to a removal, I left Tampa on or about the 2d of October (1838), for their towns on the Apalachicola river, which I reached on the morning of the 12th, with two companies of mounted men (one dragoons and one infantry), the whole under command of Captain BE. Backus, 1st infantry, and found on my arrival a portion of the 6th infantry under Major Noel. On the day fixed for their removal, I succeeded without the use of compulsion, in embarking the whole of the Appalachee [sic] Indians, about two hundred and twenty in number, to their new homes beyond the Mississippi. Under the erroneous designation “Appalachee” Indians, Taylor in- cluded not only Econchatimico’s band, but also the band of Mulatto King. According to Foreman (1932, p. 368) : The steamboat Rodney was engaged to transport the whole of the Apalachicola tribe numbering 300, and 34 Creeks who had taken refuge on Dog Island. These Indians refused to embark until they were paid the pittance promised them for their land and this was done as soon as they boarded the boat at Pensacola, .... Gen. Zachary Taylor advised against the risk of taking so many Indians to sea on one steamboat, and Daniel Boyd, the conductor of the emigrants, chartered two schooners, the Vesper and the Octavia to carry part of the emigrants as far as New Orleans. They sailed from Pensacola October 29 and arrived at New Orleans four days later. There the passengers on the schooners were transferred to the Rodney and passage up the Mississippi river began. By the eleventh they were opposite Princeton, Mississippi. The emigrants, said Boyd, had suffered very much from sickness. Six have died since we left the Chattahoochee and Biv. No'13], JIM WOODRUFF RESERVOIR AREA HISTORY—sBoyD 209 more than twenty are now on the sick list. The weather has been unusually cold for this season, which has no doubt increased the number of invalids. The water in the Mississippi river is very low; we lay two days upon a sand bar about twenty-five miles above Vicksburg. Boyd with his emigrants arrived at Montgomery’s Point at the mouth of White Tiver on the thirteenth, where he was obliged to wait for an increase in the stage of the Arkansas river, which brought him to Little Rock on November 22. As there was not sufficient water for the Rodney to ascend higher, the emigrants were transferred to the North St. Louis, a steamboat of lighter draft; they took their departure soon after and had gone about fifty miles when further progress was prevented by a bar in the shallow river. The Indians were then put ashore; their conductor hired teams to haul their sick and provisions and started with his people for Fort Gibson. It is not known whether Econchatimico was alive when the emigra- tion began, or, if so, whether he survived the journey. FOWL TOWN 2 The reason why this locality, regarded as the original Perryman family homestead, was contemporaneously called Fowl Town, has not been ascertained. Location.—The plat of the survey of fractional lot 196 of Land Dis- trict 14 of the original Early County, Ga., displays the words “Old Fowl Town” inserted so as to occupy the entire west side of the lot, beginning where the Chattahoochee River cuts off the northwest corner. The river area is about a mile northwest of the buildings on the Fairchild plantation, and more or less on the site of the Fairchild river landing. It is in the present Seminole County, Ga. (pl. 48, ¢). Identification—Survey Book DDD in the Surveyor-General De- partment, Atlanta, indicates that this lot was surveyed on January 25, 1820. Although this Fowl] Town is represented on several maps of Georgia made in the two or three subsequent decades, no earlier carto- graphical representation of the site is known. It seems probable this can be identified with the “Perimans” shown on the Early map of Georgia, 1818, reproduced as plate 9 in Swanton (1922). No other direct references to the site have been encountered. Condition.—Obliterated. Before preimpoundment clearing opera- tions the site was in woodland, traversed by a nearly overgrown trail to the landing. Authentication.—Detection of an exposed shell midden on the cav- ing riverbank led to the discovery of the site in 1948 by A. R. Kelly and Frank S. Jones (pl. 49, a). Recognition would otherwise have been impossible, as the original humous topsoil of the river bottom is covered by an overburden of sand-clay silt from 2 to 4 feet thick, on which the recent forest cover was growing. Some excavation was carried on in 1951 by students of the University of Georgia Summer School. More extensive excavations were carried on in 1953 by Joseph Caldwell of the National Park Service. Plate 49, a, shows the vertical 210 BUREAU OF AMERICAN ETHNOLOGY [Buut. 169 section he made through the clamshell midden on the crest of the riverbank, revealing the depth of silt deposit above the midden. The sherds found in the early phases of the explorations indicated, according to Dr. Kelly, the presence of four cultural horizons, which from the most superficial downward were: (a) Pure check stamp of the Montgomery Fields type (b) Some phase of the Weeden Island (c) Swift Creek (d) Pre-Swift Creek, unassociated with shell deposits. Caldwell related (personal communication) that in the later stages of his fieldwork, exploratory trenches were extended to the upper por- tion of the site, which revealed some sherds of the Chattahoochee brushed or scratched pottery, iron nails, a fragment of china, and a piece of green glass, all of which are consistent with late 18th-century occupation. These, of course, belong to a horizon higher than “(a)” on the previous list. interpretation—While the most superficial of the artifacts de- scribed are correlated with occupation in historic times, and to this extent confirm the notation of the land plat mentioned, the unavail- ability of other historic data hampers productive discussion. A point of some interest on the 1820 district survey plat is that the Indian path shown running from Fort Gaines to the fork passed by the site, which was also connected with Fort Scott by two direct paths. This suggests that the occupants were on amicable terms with the Fort Scott garrison. This further suggests that the latest occupants were some members of the numerous Perryman family. Some con- firmation of this inference is afforded by Doyle (1938-39) , who, speak- ing of the radius of operations of the friendly Indians in the service of the United States at Fort Gaines, said that in pursuit of marauders they could seldom venture lower than old Perryman’s former dwelling about 40 or 50 miles below Fort Gaines. Actually this site is about 53 miles in a direct line below Fort Gaines. His allusion to “old” Perryman, and the implication that the site was then abandoned, suggests that the original Perryman occupancy might have gone as far back as the white trader, Theophilus Perryman of the 18th cen- tury, or his halfbreed son Jim, onetime resident of Okatiokana (Okiti- yakani), and progenitor of the numerous Indians of this name (Wood- ward, 1859 p. 107). It may be recalled that the celebrated adventurer, William Augustus Bowles, married the daughter of a Lower Creek chief named Perryman, and doubtless owed much of the influence he wielded to this alliance. A statement by Woodward (1859, p. 157), in describing the contents of the letter that he drafted to General Gaines before Fort Gaines in January 1818, when under the suspicion, shortly proved erroneous, that the fort was in possession of the Indians, stated that in the event of his repulse, he would try and “descend the Riv, Bas. S®- JIM WOODRUFF RESERVOIR AREA HISTORY—BOYD 211 river [to] below the Perryman Town and go across to Fort Scott,” affords further confirmation of this inference. The direct path from this site to Fort Scott, suggests that it contemporaneously might have been the residence of George Perryman, onetime caretaker of Fort Scott. All this, however, hardly seems consistent with the applica- tion of the name “Fowl Town” to the site, unless this has the implica- tion that the occupants were Hitchiti speakers. SABACOLA, AND CHEROKEELEECHEER’S TOWN AND FORT These closely adjacent and possibly overlapping sites were not simultaneously occupied. The second settlement occupied the vacant fields of the first settlement after the first had been abandoned. Lieut. Diego Pefa (Boyd, 1949) is the authority for the opinion that if the sites are not identical, the latter is clearly adjacent to the former, from his statement that the latter was situated in the “old fields” of the former. Location —The MSS Colonial Office map (N. A. C., General, 7) of North Carolina, South Carolina, Azilia, Florida, Carolana, Louisiana and the Southern Indian Country (circa 1721-27) is, according to Crane (1928), the first detailed English map of the southern frontier. It is a compilation of data derived from various informants, among whom Colonel Barnwell was likely prominent. The positions shown are roughly relative and consequently, as well as from its scale, the map is devoid of value for exact site identification, at least in this area. It is of interest because in the point of land in the fork of the Apalachicola River it shows a large scale symbol for a fort, accom- panied by the following legend: In 1716 the Palachucolas Removed to this Place and Built this Fort under Cherekeele (ch) ee their Leader. Wright’s map (1763) does have the merit of representing the Flint River to change from a westerly to a southerly course as it approaches the confluence. It shows, just below this bend on the west bank of the Flint, an “Indian Fort.” Neither fortification or village is shown at the fork on the Purcell map of 1778. Prior to the completion of the clearing operations in the peninsula at the fork of the Apalachicola, thorough archeological exploration was virtually impossible, as the area was covered by a dense forest growth, which, probably because it had been cut over, had a very nearly impenetrable undercover of saplings and shrubs. Despite this obstacle a reconnaissance by A. R. Kelly in 1951 revealed a low mound (site 9 SE 27) about the center of Lot 42 of Land District 14 of the 1820 land survey plat of the original Early County, now forming part of Seminole County, Ga. In 1953 he discovered scratched and brushed Lower Creek sherds in a turkey patch (a small clearing in which turkeys are baited) in fractional Lot 36 (site 9 SE 20), close to the 4221415818 212 BUREAU OF AMERICAN ETHNOLOGY [Buny. 169 west bank of the Flint River, about a half mile southwest of Gaulding’s Landing (pl.49,,¢). Itisthis site which bears the identifying letter “©” on the frontispiece map. Plate 49, 6, shows a view into the turkey patch, looking east, before the clearing of the surrounding timber; the earth spoil to the right within the entrance is from exploratory trenching. Plate 49, c, shows a view over the former turkey patch, looking east, after the clearing operations have destroyed the forest cover; the further extremity of the pile of spoil (left) is shown on the right side in plate 49, 6; the trees still standing in the distance (right) are adjacent to Spring Creek. These observations were re- sumed in association with Roland Bowers subsequent to the clearing of the turkey patch. Mr. Bowers relates that at that time several postholes indicative of some structure were encountered, as well as sherds of smoky green glass, a fragment of the stem of a clay pipe, and a broad-bladed steel hatchet or tomahawk. Upon completion of the clearing operations, a thorough recon- naissance of the entire peninsula was made by Joseph Caldwell, arche- ologist of the National Park Service, who has personally communicated the substance of his discoveries (fig. 10), which he courteously has per- mitted me to summarize. ARCHEOLOGICAL SITES IN THE FORKS OF THE FLINT AND CHATTAHOOCHEE RIVERS AS SURVEVED BY J.R.CALDWELL, NPS, 9/29 ~ 10/12/53 THIS MAP 1S BASED ON Jifa WOODRUFF RESERVOIR a= CLEARING SHEET NO.3, DATED 9/29/52. EVEN NUMBERED BAG CONTOURS SAID TO BE APPROXIMATE ONLY. 000 a NUMBERED CONTOURS WERE ROUGHLY SKETCHED 1 TO SHOW SITE DETAIL. (ie ARCHEOLOGICAL SITES 87, 9SEIO WHALEY'S MILL 9SZ20 TURKEY PATCH 9SE2/ BELOW TURKE Y PATCH 95&22 GAULDINGS LANDINS 9SE23 ON SPRING CREEK POSSIBLE 9SE24 {ica TION OF CHEROKEELEECHEE T.* 9SE25 SPRING SITE 9SE26 ABOVE SPRING SITE 9SE27 YNDIAN MOUND” # AJUDGEMENT BASED ON THE CURRENT DATING OF THE INDIAN CERAMICS FOUND THERE. SCALE IN FEET Figure 10.—Plat of archeological reconnaissance in the fork of the Apalachicola River. Pav. No is) JIM WOODRUFF RESERVOIR AREA HISTORY—BOYD 213 Referring to figure 10, it will be observed that along the convexity of an arc roughly marking the sweep of the 60- and 70-foot contours be- tween the rivers, no less than 11 aboriginal sites were discovered to the west of Spring Creek. The heavy silt overburden in the river swamp below these contours prevented exploration in this area; nor was any evidence of intensive occupation discovered to the north of this zone. The surface collections made at these sites show that only two (sites Nos. 9 SE 20, the turkey patch, and § SE 24) have a historic phase. The pottery complex from 9 SE 20 more clearly resembles the usual historic complexes from the general area as described by Bullen (1950), which could be readily demonstrated to be later than 1716. At site 9 SE 24, on the other hand, superficial scraping with a power scraper revealed postholes pertaining to some edifice, adjacent to which was a pit or trench yielding about half a bushel of sherds and European artifacts. The surface finish and decoration of the sherds showed numerous specific similarities to those found on the presumed site of Palachucolas on the Savannah River, where Cherokeeleechee resided prior to 1717 (Caldwell, 1948). Such pottery has previously not been found in southwest Georgia, and hitherto has been known only from the Savannah River, Macon (Ocmulgee Fields Hitchiti), and the Kasita site near Columbus. The associated trade artifacts at 9 SE 24 are of English rather than Spanish manufacture. Thus of the 11 sites, 9 SE 24 is the only one which tentatively can be regarded as that of Cherokeeleechee’s town. This lies from 2,000-3,000 feet west by south from the turkey patch site (9 SE 20), represented on the map by the letter “C,” and probably is situated in fractional lot 113, to the west of lot 42. Should this identifi- cation of site 9 SE 24 be correct, then as will be later shown, there is reason to believe that Sabacola el Menor was on a different site, which by exclusion, would appear to point to 9 SE 20 as its former location. The 1820 plat of the land district shows the Indian trail on the east side of the Chattahoochee River to terminate about the center of lot 42, which lies directly to the west of lot 36. This suggests that there had been some occupation of the locality, probably by hostile refugees, up to about the time of the establishment of Fort Scott. Identification.—Presumptive. Condition.—Obliterated. Authentication.—The nature of the artifacts discovered by Caldwell on the site 9 SE 24 is of such a character as to afford a strong presump- tion that this is the site of Cherokeeleechee’s village. We are not aware of any recorded visit to the site in the late 17th or early 18th centuries by any Carolinian, and their references to Chero- 214 BUREAU OF AMERICAN ETHNOLOGY [Buty. 169 keeleechee’s occupation are doubtless based on reports from Indian informants. Interpretation.—Some preliminary orientation is necessary before proceeding with this sketch. On the map of 1721-27 previously re- ferred to, mention was made of the Palachucolas. This word is an English elision of the name Apalachicoli, applied to one of the 13 Lower Creek villages on the Chattahoochee on the earliest town list known to us, that of the Bishop Calderén, later discussed (Wenhold, 1936). When first known, the Spanish believed this town exercised preeminency over all other Lower Creek towns, in recognition of which they called the Lower Creek settlements on the Chattahoochee, the Province of Apalachicoli. Some, at least, of the people of this town early became inclined to the English, and moving across Georgia, established a settlement on the right bank of the Savannah River about 1680 (Swanton, 1922, p. 131), roughly contemporaneous with the Yamassee exodus from Guale. This town was known as Palachucola, and was established several years before English influence was mani- fested on the Chattahoochee. This became an important stopping point for the English traders on their inland journeys.