10 BUREAU OF AMERICAN ETHNOLOGY 



of arclieological sites. Although 10 of them were excavated or tested 

 by the field group, results were not encouraging. Apparently the 

 area was occupied quite extensively but cultural debris is scanty and 

 the habitation sites thin. 



At the Gregg site (39HY222), one of the largest in the pocket, 

 portions of two earth lodges were excavated and five interhouse areas 

 were tested. A single circular lodge and several tests were dug at the 

 Fry site (39HU223) , two lodges were exposed at site 39EnD'224 nearby, 

 and a lodge and two large cache pits were cleared at the Hawk site 

 (39HU238) . Architectural features were not found at the remaining 

 sites, but several clusters of exterior cache pits were cleared at the 

 Saint John site (39HU213) and artifact collections were made at 

 sites 39HU225, 39HU230, 39HU231, 39HU249, and 39HU250. 



The earth-lodge structures excavated within the area of the cul-de- 

 sac are all quite similar. Each was circular, with an irregular pat- 

 tern of wall posts and four central supports. Entrance passages, 

 where found, opened to the south or southwest. A small central 

 hearth was characteristic and there were usually secondary firepits 

 and one or more small basin-like or bell-shaped cache pits. 



Ceramics were preponderantly simple-stamped, with Talking Crow 

 and "Category B" rims most usual. Other artifacts were not distinc- 

 tive, and except for several copper pendants and an iron blade hafted 

 in a split bison rib from the Hawk site, there was no evidence of 

 European contact. 



On August 12 Jensen transferred his field party to the right bank of 

 the Missouri where he assisted Hoffman in the excavation of the La 

 Roche sites. The party completed work on August 23, after 72 days 

 in the field. 



At the beginning of the year, a third party of nine men, directed by 

 William J. Folan, was assisting John J. Hoffman in excavations at the 

 La Roche sites. On July 16 the Folan party moved to the left bank 

 of the Missouri to begin work at the Chapelle Creek or Grandle site 

 (39HU60), a large, fortified, multicomponent village in the central 

 Big Bend Reservoir. Extensive trenching, exposing sections of three 

 houses, a section of the defensive ditch, and a number of other features, 

 was completed. Evidence of the earliest occupation consists solely of 

 artifacts that are invariably found in the prehistoric, rectangular 

 house complexes of the Big Bend region. The second component 

 consists of the fortified settlement proper, which seems to be attribut- 

 able to the historic Stanley-Le Beau complexes usually regarded as 

 Arikara. The uppermost deposits contain additional European ma- 

 terials that are suspected to be the remains of a small (and poorly 

 documented) trading post. In view of our present knowledge of the 

 early history of the Big Bend region, it may be difficult, if not impos- 



