SEVENTIETH ANNUAL REPORT 19 



and the outlines of the buildings that stood inside the fortification were 

 followed. Exact dimensions of the fort and buildings were obtained, 

 as were some of the constructional features of the interior of the build- 

 ings. All wooden structures had been burned, and evidence indicates 

 that the post was destroyed shortly after abandonment in 1872. About 

 90 percent of the site was excavated and no additional work will be 

 required there. A number of discrepancies found between the various 

 features revealed by the digging and a plan of the fort drawn in 1871 

 raised a number of puzzling historical problems. About 500 yards 

 northwest of the fort the remains of a ''Missouri Dugout" were found 

 and excavated. At the end of July the party moved to the Fort 

 Randall site (39GR15) on the west bank of the Missouri River half a 

 mile southeast of the Fort Randall military post. Work there showed 

 that the remains were those of a brick kiln, which probably belonged 

 to the period of Fort Randall I. The remains of the kiln and features 

 associated with it were completely excavated and the party left the 

 Fort Randall Reservoir area at the end of August, proceeding to the 

 Kirwin Reservoir in Kansas. 



During the 1952 field season work was also carried on in the Fort 

 Randall area by the Nebraska State Historical Society and the Uni- 

 versity of Kansas under cooperative agreements with the National 

 Park Service. The Historical Society party under the direction of 

 Marvin F. Kivett continued excavations in two sites (39LM26 and 

 39LM27) located along the highway a short distance east of Oacoma 

 and about 2 miles west of Chamberlain, S. Dak. Some digging was 

 also done at a site (39LM81) lOi^ miles upriver from Chamberlain. 

 The work at the first two locations, which was completed, showed 

 evidence of a historic Siouan occupation underlain by an earth-lodge 

 village belonging to what has been called the Fort Thompson focus. 

 The third site was found to have three components, historic Siouan, 

 a level producing a simple-stamped type of pottery which has not yet 

 been culturally correlated, and an earlier Woodland occupation. The 

 University of Kansas party under Dr. Carlyle S. Smith spent a third 

 season at the Talking Crow site (39BF3) about 3% miles below Fort 

 Thompson, S. Dak. During the three seasons at the site 9 houses were 

 completely excavated, 4 were partially excavated, and 14 were tested 

 to obtain their dimensions and samples of materials from them. 

 Stratigraphic tests were made in three refuse mounds, trenches were 

 dug across the surrounding fortification on four sides of the site, two 

 long trenches were cut through areas between the houses, and numer- 

 ous other test pits and trenches were dug. From the data obtained it 

 appears that the site had four components. The latest was Siouan 

 dating from shortly after the Civil War. Prior to that was the last 

 occupation by earth-lodge-building people, probably the Arikara, 

 during the period when European trade goods were beginning to 



