24 BUREAU OF AMERICAN ETHNOLOGY [bull. 118 



door of the primary structure was located at that point, rather than 

 on the southwest wall. 



The primary floor, which was covered by the fallen burned struc- 

 ture erected on it, showed the effect of fire, as did also the clay 

 above and adjacent to the charred remains. This fact, together 

 with the obvious fact that the charred remains did not suffer com- 

 plete destruction by fire, definitely shows that the fire was smothered 

 out. There was here no evidence of any plastering of clay on wattle- 

 work walls, so often reported elsewhere. 



The finding of burned clay resting on top of the fallen structure 

 very naturally raises the question of how the clay could be so burned 

 while serving to smother out the fire. This smothering must have 

 occurred since carbonaceous material so readily burned, as dead 

 grass, split cane, and small dry poles, was not completely destroyed. 

 One naturally wonders if, in case a wooden structure so constructed 

 had been fired, it would have been humanly possible to have covered 

 over such a burning structure with earth quickly enough to have 

 produced the rather uniform partial burning observable over the 

 whole primary floor, and at the same time to have allowed enough 

 time for the heat of the fire to have discolored the covering clay. 

 From the condition of the clay it is not believed that the building 

 was covered over with fresh earth while it was burning, but rather 

 that the structure, which was built of logs, split cane, and grass, 

 had been covered with a considerable thickness of earth while it 

 was standing. Under such conditions, when fired, it would have 

 had an opportunity to burn over the whole interior slowly in an 

 atmosphere deficient of oxygen, and finally, when it did collapse, 

 the outside covering earth would have soon smothered out the fire. 

 It would, itself, have been discolored by the burning which took 

 place while the building was standing. 



Notwithstanding the large amount of excavation and abundance 

 of evidence of occupation, this site yielded very few artifacts. In 

 the general digging, a small fragment of a steatite pipe and a frag- 

 ment of a large biconcave discoidal were found. Other than a few 

 crude hammerstones and flint chips, these were the only stone arti- 

 facts discovered. 



In the mounds there were scattered shells of snail and mussel. 

 Many of these mussel shells were perforated in the center and 

 showed evidence of use as hoe or scraper. It appears that they 

 were used to scrape up the hard clay from its native deposit at 

 the time of the construction of the mound, and by accident became 

 incorporated in the mound. 



Pottery from this mound was quite scarce. Even on areas which 

 were definitely house sites, sherds occurred infrequently. All of 



