1(34 BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



chunks of cedar in the layer of earth which covered the structure 

 floor to a depth of 30 inches. If they protruded through the earth 

 roof they must have been in the way of new construction and were 

 probably removed by the builders by the simple expedient of wiggling 

 them loose and pulling them up. In such a case the post hole would 

 have been filled with black soil and there would have been neither 

 cedar logs nor post molds remaining. This may account for the ab- 

 sence of any evidence of posts in the gaps on the east and north walls. 

 On the south wall the builders seemed to have desired to make the 

 new wall coincide with the old, for they set in a new post between 

 each of the old ones which protruded through the roof earth of the 

 old building. This placement of the new structure made the removal 

 of the old posts unnecessary and left the two structures interlocked. 

 Each post of the old structure was in juxtaposition to one of the new. 

 This was beautifully shown by some seven pairs of post remnants in 

 the south wall (pi. 108). 



It was conjectured in the preceding paragraph that the posts which 

 came up through the roof earth might have been pulled out by the 

 builders of the new structure by the process of wiggling them free 

 and lifting them out. Conversely, the difficulty of removing the posts 

 on the south wall, some of which had settled as much as 30 inches 

 farther into the soil of the old village, because of the weight on the 

 roof, might have decided the builders to make the south wall of the 

 new structure coincide with the old. 



Whatever the motives governing its construction, the secondary 

 structure was raised on the site of the old. The south walls were 

 practically coincident vertically, but the orientation of the two build- 

 ings seemed to have been slightly different. The north wall of the 

 secondary structure was somewhat difficult to locate, therefore its 

 dimensions are in doubt. Enough was found, however, to make it 

 reasonably safe to assume that it was not very different in size from 

 the primary structure. 



The primary posts were rather large — 14 inches in diameter. They 

 were in all probability mostly good specimens, and were embedded in 

 river-bottom soil which was often very wet and never thoroughly dry. 

 This condition of the soil resulted in the fair preservation of many 

 of these posts and in the extraordinarily good preservation of a few, 

 which were so little affected by their long burial that, when sawed in 

 order to get sections for the study of tree-growth rings, they gave off 

 the usual odor of red cedar. Most of these post fragments, however, 

 were so decayed that they were reduced to "punk", which would keep 

 its form while wet, but would split, check, and disintegrate when dry. 

 It was possible to preserve most of these remnants by treating them 

 with a saturated solution of paraffin in gasoline. After several ap- 



