48 BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



together to form a complete rib of the building, like an inverted 

 basket. This process was carried on throughout the length of the 

 side walls, stopping about 3 feet short of the corners of the building. 

 End walls were set in the same way and as the small ends of the 

 posts of these end walls were bent down a 1-foot mesh of cross poles 

 was formed overhead. Each of the poles was 3 inches in diameter. 

 On top of this mesh of poles split cane was spread and grass thatch- 

 ing attached. The whole building was then covered with earth. The 

 weight of the roof was partly supported by large individual posts 

 set up vertically in the interior of the structure. Later the building 

 was burned, and, being covered with earth, incomplete combustion 

 resulted. 



The structure then collapsed on the floor, as shown in plate 25, b. 

 This structure suffered very uniform partial combustion. Because 

 of this fact, the falling excavated portions readily revealed the 

 method of construction. Plate 26, <z, was taken of a section which 

 showed grass thatch. After falling, the earth on the roof of the 

 building smothered the fire and left the charred remains preserved 

 without further damage. When this charred structure was removed 

 from the clay floor a number of plaited and twisted strings, some 

 with knots, were found in the charcoal, which definitely suggests the 

 method for attaching the grass and split cane to the post£ of the 

 structure (pi. 27, b). 



Plate 26, 5, shows the floor of Structure No. 2 cleared of the fallen 

 building. The pile of charcoal resulting from the burned structure 

 is shown in the trench at the right. This figure also shows, in the 

 right distance, the humus line, and clearly indicates that the top of 

 the yellow clay hardpan was above the floor of this structure. The 

 floor was, by actual measurement, 12 inches below the hardpan sur- 

 face. This was the only instance recorded in the investigation of 

 this region where a structure floor was dug below the hardpan. 

 Further evidence of this fact is shown in plate 27, a, which shows 

 a section of the east wall of Structure No. 2. The light-colored soil 

 to the left represents the filled-in subterranean primary floor. The 

 clay hardpan dug out by the builders shows yellow in comparison 

 with the village soil when photographed with a color-sensitive film. 

 In Structure No. 2 no horizontal molds were found in the trenches 

 and no rocks were used to chink up posts. It is believed that these 

 were not used because, due to the hardness of the clay subsoil, such 

 aids in construction were unnecessary. 



When the floor was cleared many scattered post molds of small 

 size were found within the structure. These were formed, it would 

 seem, by stakes driven into the floor. There was also a central line 

 of post molds, designated as Feature No. 28, figure 14. This central 



