FOOD AND TREE CUTTING 109 
all around. This of course makes the work somewhat easier, 
for then not as wide a notch has to be cut as when the cut- 
ting is all done from one side. Figures illustrating this 
point have already been given, in the cases of the large aspen 
and pine. 
On a steep bank by Tower Creek I found two aspen stumps 
close together, about three feet high. Each had a notch 
cut on the lower side, but the cuts which felled the trees 
were made on the up-hill side several inches above. Evi- 
dently the beavers had decided it was unsafe to continue 
cutting below, and had gone to the upper side to finish the 
job (Fig. 60). 
A beaver does not always work continuously to cut down a 
tree. Thus the seventeen inch aspen was first seen July 
20, when it had a notch cut around it, not especially recent 
work though evidently done some time the previous spring 
orearly summer. The tree was seen quite frequently during 
the succeeding weeks, and was still standing August 28. 
On September 4 it was found to have been felled during the 
week preceding. In the Colorado Museum of Natural 
History, Denver, is a cottonwood stump 293 inches in di- 
ameter, concerning which Director J. D. Figgins wrote me 
that there was evidence of long intervals between cuttings. 
Shiras states that a colony, estimated to contain eight 
beavers, had from twenty-five to thirty trees in the process 
of felling at one time, showing that they did not always work 
at a tree until it was felled. ‘“‘On an average it took from 
ten to fifteen days before large trees were felled by reason of 
this intermittent cutting.”’ He also shows a picture of a 
beaver cutting on a black ash ninety-one inches in circum- 
ference. Beavers worked at this tree in three different 
years and never cut it down. 
There is much variation in the height of beaver-cut 
stumps. I have found a few two or three inches high, or 
