12 BULLETIN OF THE NATURAL HISTORY SOCIETY. 
larger fragments would be flung carelessly out of the door and 
left to fester in the sun. Occasionally a fit of house-cleaning 
would seize upon the occupants of one of these huts and the 
fragments lying within on the floor would be scraped together 
and shovelled out at the door to mingle with the heaps of shells 
and broken bones without. The evidence of these actions on 
the part of the inhabitants of hut A is found in the frequently 
alternating layers of charcoal mingled with broken bones, 
of pottery clay and other refuse from the hut, which are ' 
found in the kitchen-midden before the door (H). These occur 
with considerable regularity and frequency, in alternation 
with layers of clam shells which form the bulk of the shell 
heap. Owing to these occasional house-cleanings, and the 
vast quantities of clam shells thrown down around these 
dwellings, the kjokken-modding increased much more rapidly 
than the deposits within the huts; and although fresh material 
was frequently brought in to level up the interior of the huts, 
the boughs and perishable matter within gradually decayed 
away and the floors sank down, so that now the depth of the 
deposit within the site of the huts is only about half as great 
as that of the shell heaps without. 
The fire-place (K) of this hut was found to have been kept 
in the same spot from the time the hut was first built, almost 
until the settlement was abandoned. Such, however, was not 
the case with an adjoining hut bottom (B), which was gradu- 
ally shifted to the east, so that at the close of the occupancy 
of this village site it was about two feet from its original 
position. A similar want of permanency in the position of the 
individual hut was found to exist in other hut bottoms. The 
huts, therefore, must have been re-built from time to time, 
and perhaps were deserted for a part of the year. 
There seems, however, to have been sufficient permanency 
to these hut bottoms to warrant the assumption made in a pre- 
ceding page, that each saucer-shaped depression in the kitchen 
middens of this village marks the site of a hut. 
From the present aspect of the surface of the kitchen - 
middens at this village site, a rough approximation to the 
population of Bocabec River during this latter part of the 
