564 THE AMERICAN NATURALIST.  [VOL. XXXIIL 
and carriages, without having the effect of frightening the 
birds; and they showed no concern when I collected the food 
pellets from the ground beneath them. In the coldest part of 
the winter the owls regularly faced the east, and at one time, 
when four roosted together, three of them roosted near the 
center of the tree, one near the top, and each occupied 
regularly a particular branch. 
About thirty feet to the southeast of the house is a magnifi- 
cent Norway spruce, the roosting place of a short-eared owl. I 
first observed this bird on February 2, and kept it under regular 
observation from February 26 to March 26, except for a lapse 
of a few days between March 8 and 14. For a part of this 
time ong of the long-eared owls, after leaving the arbor vitæ 
tree, roosted in this Norway spruce also. 
These two trees were the only regular roosting places of owls 
in the neighborhood at this time, though there were a number 
of evergreen trees in the vicinity which were casual feeding 
perches, under which I found pellets. 
My observations were of the following nature: Each day the 
number of owls were counted on the two roosts, and at regular 
intervals (usually each Sunday) all pellets were collected from 
the ground beneath each roost. These pellets were carefully 
examined, the number and kind of animal remains in them 
noted, and so the daily food average per owl computed. Here 
I wish to thank my friend Mr. Witmer Stone, curator at the 
Philadelphia Academy of Natural Sciences, for his assistance 
in determining the species of mammals. 
First, a few notes as to the construction of the food pellets. 
It is well known that owls, as a rule, swallow their prey whole, 
that the food remains in the stomach until all digestible sub- 
stances are extracted and passed into the small intestine, and 
that finally the remaining mass of undigested substances (bones 
with hair or feathers, hard parts of insects, etc.) is disgorged 
as a pellet (bolus). The pellets are more or less ovoid in form, 
and when freshly disgorged covered with a slimy mucus. 
Each pellet, of the two species of owls studied, usually contains 
the remains of only one food individual (ż.e., one mouse or one 
bird); about one in twenty contains two individuals; rarely 
