246 BUREAU OF AMERICAN ETHNOLOGY [BuLu. 185 
uncovered. Prior to our excavations, a portion of this hearth had 
been destroyed by river action during some period of high water. 
The remaining section measured 4.3 feet across its longest diameter 
and 0.8 foot across its widest sector. 
Resting upon a bed of ash and charcoal on the semilunate hearth 
were 42 river cobbles, which were either fire-cracked and/or broken. 
This assemblage of rocks and ash lay in a slight saucer-shaped de- 
pression. It was at this same level and near the hearth that a number 
of animal bones, mostly bison but a few deer, were collected. None 
of them showed any evidence of having been worked or utilized as 
tools. 
Immediately adjacent to the animal bones were a number of small, 
light-gray-colored circular outlines, Feature 17, surrounded by a 
yellowish-colored clay which made each quite distinct. The light 
gray clay was derived from the stratum above. Each impression 
measured approximately 0.5 foot in diameter. All were carefully 
sectioned, both horizontally and vertically, to determine not only their 
origin but their purpose. It was decided that they were the imprints 
of ungulate hoofs of a fairly large size. Since a large number of 
bison bones were found on this same level, we have inferred that a 
number of bison, or perhaps only a single bison, wandered across the 
then surface during a wet spell when their hoofs sank into a lower 
stratum, intruding some of the clay from a higher level into a lower 
stratum. They were actual imprints of bison hoofs. This not only 
explains what Enger (n. d.) took to be random postholes, but also 
the circular outlines found during the present testing. 
Section 3—A third section located 133 feet south of Section 2, 
measuring 20 feet long and 5.5 feet wide, was excavated to a depth of 
9.5 feet. As found in the other excavation units the deeply buried 
dark clay stratum dipped at a greater angle as we went south, so that 
at this point it lay just slightly less than 7 feet from the present 
surface and was 0.8 foot in thickness. To insure that no earlier and 
deeper deposits containing human remains existed, we carried the 
digging well into the sterile sandy substrata. 
Kach of the many strata was carefully removed, but it was only 
upon reaching a light yellowish layer, at a depth of 6.2 feet below 
the present surface, that another series of bison imprints was noted, 
Feature 6. These, like the former, were carefully investigated and 
yielded the same evidence. 
The main reason for digging this particular section was the pres- 
ence of portions of two burned areas, identified as Features 18 and 2, 
which were exposed by caving action of the banks. Both of these 
burned areas turned out to be the remains of small hearths or places 
where fires had been built directly upon the existing ground surface. 
No prepared surface could be identified. There was greater change 
