486 ARCHEOLOGICAL INVESTIGATIONS—II [ETH. ANN. 44 
not at all worn on the edges, but all presenting the appearance of 
having been washed clean by running water or little waves. Lying 
on the bedrock, of which this minor wall was the lower margin, was 
a deposit of breccia similar to that below, but nowhere more than 
2 feet thick; it extended to the foot of the high cliff. It was covered 
with bluish-gray clayey earth containing many small stones, and 
merging into the soil above, there being no distinct line of demarca- 
tion; it was very tough and difficult to pick down. Scattered 
through it were large masses of débris from the cliff. No gravel 
appeared in this part of the trench. 
In the black deposit, and to a less extent in the clay and rocks 
above it, fragments of mastodon bones were abundant. All of 
them, with one exception, were so broken and scattered it seemed 
impossible that they could ever have been entire at or near the 
spot where they lay. The exception was a femur, in many pieces, 
which might at one time have been intact where it was uncovered. 
As this was near the present surface, it may have been broken by 
falling or rolling stones. 
No satisfactory explanation can be given for the presence of such 
a vast number of these remains in so small an area, or as to why 
any at all should occur under such conditions when they are rarely 
found elsewhere in the vicinity. 
A theory has been offered that, like the modern elephant. the 
animals came to this place when they felt death approaching. But 
the gravel and the conglomerate prove that the site was then under 
water. Another theory is that a great herd was driven by a savage 
people or by carnivorous animals to the edge of the cliff where, 
being hard pressed, they plunged over to destruction. But they 
could easily have made their way to the bottom on either side of 
the cliff within a short distance. Another hypothesis, plausible but 
not convincing, is that after they had died somewhere else the bodies 
floated to this spot and were stranded. There is abundant evidence 
that at the height of the glacial floods this valley was filled with 
water which covered to a depth of many feet all the space between 
the hills. Due to the surface drainage around this lake it would have 
a gentle outflow to the Mississippi. The point on which most of 
these bones are found, projecting into this lake, would cause an eddy 
to form on its lower side. Carcasses floating near the shore would 
lodge against the upper part of this point; those a little farther out 
would swing into the eddy and stay there. When they decayed, the 
bones would settle to the bottom. Sediment from the water and 
talus from the cliff would cover them. Drift, also, would wash 
ashore; as the water receded vegetation would follow its margin. 
These would decay and form a mold over and among the bones, 
accounting for the black deposit. 
