70 BULLETIN 1S3, UNITED STATES NATIONAL MUSiEUM 



up >\'hen the central post holes began coming to light. It seems 

 probable now that after collapse of the lodge the resulting depression 

 was used as a refuse dump by the occupants of other nearby units of 

 the village. 



As to structural details of the house, it may be inferred that the 

 four central posts, 10 inches or less in diameter and perhaps 10 feet 

 in height, carried four beams to form a square above the hearth area. 

 Smaller posts, not over 4 inches in diameter and 5 or 6 feet high, were 

 set just within the edge of the house pit and were likewise coimected 

 with stringers at the top. Against these were leaned short poles 

 resting on the original ground surface just back from the edge of the 

 excavation. Efforts on our part to verify such a placement by care- 

 ful stripping of the rim of the pit brought only negative evidence. 

 This was perhaps due to modern cultivation, which, penetrating to 

 a depth of 7 inches or more, may have destroyed such traces of wall 

 base as once existed. The roof form is still less certain. It may 

 have consisted of converging raftei*s, none much exceeding 10 to 12 

 feet in length, with their lower ends resting on the outer row of 

 stringers and their tips ahnost meeting at the smoke hole directly 

 above the fireplace. This would have given a subhemispherical ap- 

 pearance to the finished structure when the grass and sod were in 

 place. Alternatively, shorter rafters could have covered only the 

 narrow zone between the exterior stringers and the central square, 

 the latter area being subsequently covered with a flat roof. A priori, 

 such a roof would seem to have been less well adapted to a region 

 where the annual precipitation averages 35 and sometimes reaches 

 50 inches. On the other hand, rafters running to the smoke hole 

 would have had only the lower third of their length between the two 

 rows of stringers, which might or might not have provided adequate 

 support for the eai-th covering their upper portions on the central 

 part of the roof. 



"While only one lodge site was completely worked out, there is evi- 

 dence to show that this was not unique. Kemnants of another about 

 20 feet across and 2 feet deep, with vertical sidewalls, were found in 

 the road cut in the northeast corner of the terrace. The remaining 

 strip of floor was less than 4 feet wide, and no postholes or firepit 

 could be found. Basins possibly representing shallow caches were 

 present and yielded a few artifacts. Lying 400 yards northeast of 

 house 1 on the 830-foot contour, this was tentatively designated house 

 2 (see fig. 7). 



The bits of burnt grass-impressed clay found just east and also south 

 of the Steed farm buildings suggest the presence of additional sites 

 there, as do the refuse-filled caches and midden deposits between the 

 two gullies in the west part of the site. More recently, since close of 

 our work, I have been informed by Mr. Shippee that fall plowing has 



