°—o10 BUREAU OF AMERICAN ETHNOLOGY [Buty. 185 
Bismarck, N. Dak., which is identified on native testimony as a site 
abandoned by the Mandan about A.D. 1764. The site was partly 
excavated by W. D. Strong in 1938. The preliminary description of 
the pottery (Strong, 1940, p. 364, and pl. 5) is based in part on a 
statistical analysis made by Carlyle S. Smith. (In November 1954 
Dr. Smith furnished me an annotated copy of a chart showing corre- 
lations of rim decoration and rim form for 720 rim sherds, and of 
surface finish and rim form for 315 body sherds, from Slant Village; 
and Dr. Strong, in a letter dated March 31, 1955, granted me per- 
mission to use these data. But I will not introduce the correlated 
statistical data here because this is the only comparative collection 
I am utilizing which has been treated in this manner.) 
2. The Sheyenne-Cheyenne or Biesterfeldt site, on the lower Shey- 
enne River in southeastern North Dakota. This fortified earthlodge 
village, partially excavated by Strong in 1938, is identified on native 
testimony as a Cheyenne site which was attacked and burned by the 
Chippewa. Its destruction occurred, according to Strong’s estimate, 
circa A.D. 1770. Descriptions of the pottery obtained from the site 
have been published by Strong (op. cit., pp. 373-374, and pl. 8) and 
by Wood (1955). 
3.. The Buffalo Pasture site on the right bank of the Missouri River, 
about 8 river-miles above Pierre, S. Dak. This fortified earthlodge 
village and two other earthlodge sites upstream, the Fort Sully and 
Lower Cheyenne River sites, were excavated or sampled by a Colum- 
bia University expedition in 1939 under the direction of Strong, who 
identified them as Arikara “in origin,’ on comparision with the 
slightly later, well-documented Leavenworth or Lewis and Clark site 
on the right bank of the Missouri above the mouth of the Grand 
River. The pottery from the three sites has been described briefly 
and collectively by Strong (op. cit., pp. 380-381, and pl. 9). It should 
be noted that further, extensive excavations conducted by the Missouri 
Basin Project in 1951-1958 at the three sites have revealed that two 
of them, the Fort Sully and Lower Cheyenne River sites, are multi- 
component sites, and that only the latest major component in each 
case seems to be identifiable, like the single component at the Buffalo 
Pasture site, as “protohistoric Arikara.” Reports of these investi- 
gations are now in preparation. 
The five ceramic complexes, numbered 1 to 5, are compared with 
respect to 22 attributes in table 5. An “x” under a given com- 
plex means that the attribute is present therein. Occasionally the 
“x” is modified by “rare” or an arabic numeral in parentheses. The 
latter indicates the few actual specimens recorded. 
