RIVER BASIN SURVEYS — F. H. H. ROBERTS 361 



people stayed there. The artifacts suggest that the same cultural 

 group was involved from first to last. A considerable amount of dry 

 material such as is rarely present in open sites was obtained. There 

 are pieces of cordage, sections from arrow and spear shafts, fragments 

 from bow staves, portions of textiles, matting, and bits of basketry. 

 Among the objects made from more durable materials are stone projec- 

 tile points, bone implements and beads, and shell beads. Study of the 

 specimens made from perishable materials should throw considerable 

 light on that phase of the arts and industries of the people in the area. 

 On the basis of the artifact collection as a whole, it would seem that 

 the shelter probably was occupied by either the Nespelim or their 

 eastern neighbors, the Sanpoil. There were objects of white manu- 

 facture on the surface in the cave attributable to picnic parties held 

 there by modern residents of the region. No such material was found 

 beneath the surface, yet it is difficult to assign the aboriginal objects 

 either to historic or pre-Columbian times. 



VILLAGE REMAINS 



A major part of the excavation program has been concerned with 

 village remains. In the Columbia Basin digging has been carried on 

 at village sites in the McNary, O'SuUivan, and Chief Joseph Reservoir 

 projects. At each of the locations the activities were directed toward 

 the clearing of house pits and the excavation of midden deposits. 

 The house remains show that many of the structures were circular to 

 oval in form, with diameters ranging from 25 to 40 feet, and were 

 grouped in clusters along the terraces above the river. Because of the 

 scarcity of timber, the Indians apparently took the main supports 

 from the structures with them when they moved from place to place. 

 As a consequence there is little to indicate the type of superstructure. 

 It probably consisted of a framework of poles to which branches or 

 mats were lashed. The poles were not embedded in the earth, but 

 some stability was obtained by heaping dirt against the outside of the 

 walls. The floors were 2 to 3 feet below the surface near the center 

 and sloped gradually upward to about the ground level at the rim. 

 The village patterns were very simple. At most sites the houses were 

 strung along the river without any particular attempt at formal place- 

 ment, although in several instances there apparently was a twofold 

 division, with one group of dwellings being located at one end of the 

 site and the second group at the other. Where such arrangements 

 were noted one cluster of house pits generally was smaller than the 

 other. Wliat significance there may be, if any, to that situation has 

 not been determined. 



At a number of the sites the remains of long, oval, or rectangular 

 "mat houses" were found. The latter apparently represent a later 



