19Q BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



extremeties. There was a descent of about 35 feet from the front 

 to the first floor and a distance of about 100 feet from the front to 

 the back wall. A passage led off to the right and another to the left, 

 both being rather precipitous, due to the many tons of talus which 

 had accumulated at the entrance. This latter condition thwarted any 

 possibility of excavation, except for an area 25 feet by 8 feet along 

 the back wall. 



Former habitation was evidenced by the presence of a few dozen 

 small potsherds which appeared to be similar to those recovered 

 from previously investigated caves of the area. A fragment of hu- 

 man parietal was also found. 



RECTANGULAR STRUCTURES 



One of the outstanding features revealed by the archaeological 

 survey of the Norris Basin was the rather general use of rectangular 

 house structures of various sizes. Of these structures, those of 

 smaller size were designed for use as dwellings. The larger struc- 

 tures probably served as public buildings or town houses. The pres- 

 ence of such structures was generally revealed by post-mold patterns 

 in the structure floor. In a few cases the remnants of posts were 

 found still in situ. On 18 sites there have been uncovered 54 post- 

 mold patterns, every one of which indicated a rectangular structure, 

 generally with rounded corners. 



Dwelling Houses 



Of the 54 structure patterns found, less than half were thought 

 to be dwelling-house sites. This conclusion was reached because of 

 their smaller size and because of their irregular placement on the 

 village sites. In general, these dwelling houses rarely exceeded 20 

 feet in greatest dimension. Their post-mold pattern indicated the 

 use of small posts in the walls, generally not exceeding 4 inches in 

 diameter. Such structures were erected above a prepared floor which 

 was usually well made and easily distinguished from the humus- 

 filled soil about it. Such floors and post-mold patterns lay in the 

 hardpan just under the humus layer and were usually not more than 

 10 inches below the surface. It is hardly possible that such buildings 

 could have had any earth on their roofs, since their presence is not 

 marked by any recognized mound. They probably were constructed 

 of cane and grass thatch on a system of poles or posts set in the 

 earth. Occasional post molds in the interior of the post-mold pat- 

 tern may indicate the presence of furniture in the dwelling, or pos- 

 sibly such molds may have been made by roof and wall supports. It 



