58 USEFUL BIRDS. 
While the Catbird’s stomach might not be large enough to 
contain thirty full-grown locusts, it would easily contain more 
than thirty small ones. The statement that thirty grasshop- 
pers were found in the Catbird’s stomach might also need 
modification in another way. The least fragment of an in- 
sect found in a bird’s stomach is usually considered good 
proof that the bird has eaten that insect. There might be 
found in the stomach of a bird a mass of unrecognizable 
material, from which the expert would be able to sort out 
and recognize enough of the harder parts of different grass- 
hoppers to prove that thirty of. these insects, of consider- 
able size, had been eaten within a certain time, even though 
a greater part of those first swallowed had already disap- 
peared from the stomach. 
Prof. F. E. L. Beal writes me as follows regarding the 
methods used at the United States Department of Agri- 
culture in counting the insects found in the stomachs of 
birds : — 
In the case of grasshoppers and caterpillars it is the jaws (mandi- 
bles) that are counted. Birds when not sleeping appear to eat all the 
time when not occupied in other duties, such as nest-making or feeding 
their young. The process of digestion is continuous. The more easily 
digested parts pass out of the stomach very quickly, but the hard parts 
remain somewhat longer. In this way when a bird is feeding upon 
grasshoppers the jaws of those first eaten remain after the rest of 
the body has passed on. When the stomach is opened the jaws are 
counted, and for every two we estimate at least one grasshopper killed. 
In cases where only a few insects were involved I have taken the pains 
to pair the jaws, and in this way have often found that the number that 
had been eaten was more than half the number of jaws. In this work 
each head that appears to be whole is carefully examined, to see that it 
has not lost one or more of its jaws; were it not for this precaution, 
the insect might be counted twice. Caterpillars, like grasshoppers, 
are easily broken up, and so the heads are counted when whole; other- 
wise the jaws are counted. 
The variation in size of different species of insects should 
also be considered. While the caterpillars of some species 
of moths reach three or four inches in length, others never 
grow to be half an inch long. 
These and other similar considerations, well known to 
