NO. 2 SALVAGE PROGRAM, I95O-I95I — COOPER 53 



debris, apparently associated with the military post, found on a lower 

 terrace to the southeast will probably be covered by a proposed chalk 

 blanket. 



Other sites which were revisited and in which minor test pits were 

 dug include Whetstone Agency, established in 1868 for Brule and 

 Ogallala Sioux and occupied until 1872; the Lower Brule Agency 

 at the mouth of American Crow Creek, which was in existence from 

 1868 to 1892, at which time the Agency was moved to its present 

 location near the Big Bend of the Missouri River; and Fort Hale, 

 the military post attached to the Lower Brule Agency, but located 

 several miles above it. At each of these sites there was abundant evi- 

 dence, in the form of surface irregularities and debris, of the former 

 existence of the establishments in question, although most of Fort 

 Hale had been destroyed by a shift in the course of the Missouri 

 River. A search was made for the site of the Brule Agency, a very 

 temporary predecessor of Whetstone Agency presumably situated 

 just below the mouth of the White River. No evidence of the site 

 was observed, perhaps because there was little construction during 

 its brief existence in 1867. 



Owing to its importance in the history of White penetration of the 

 region, notably its connection with the Leavenworth expedition of 

 1823 to chastise the Arikara, the site of Fort Recovery, possibly 

 identifiable with the Cedar Fort referred to in earlier sources as in the 

 same vicinity, was made the object of a protracted and intensive 

 search, but without success. Mattes (1949, pp. 522-528) has made a 

 convincing case for its location within a rather restricted area on the 

 right side of the Missouri near Oacoma, but repeated examinations of 

 the present surface and numerous test pits failed to produce indica- 

 tions of its presence. It may be that the post was on a lower terrace 

 and has been covered by flood deposits, since there is reason to believe 

 it was occupied long enough to have left ample traces of its existence. 



The most intensive excavations of the 1950 season in the Fort Ran- 

 dall Reservoir were at a site (39LM53) superficially marked by low 

 mounds and depressions and by fragments of burned earth. The site 

 lies along the border of a flat, uncultivated terrace immediately adja- 

 cent to the timbered bottoms on the west side of the Missouri River 

 about 3 miles above Chamberlain, Horizontal stripping of two of the 

 mounds uncovered two structural units, each consisting of two rooms 

 separated by a roofed space. Fortunately the structures were burned so 

 that there were observable certain details of construction which are usu- 

 ally obscure in the sites of destroyed wooden buildings. Thus it seems 

 certain that only the northern room of the more southerly unit had 



