788 Wisconsin Academy of Sciences, Arts, and Letters. 
tempted, so that, inj all, the constituents were grouped under 
twenty-two different heads. Chironomus and Simulium were 
noted separately from the other Diptera because they were rep¬ 
resented chiefly by their aquatic larvae. The term “insect 
fragments” includes all fragments that were too small to be 
identified positively. The other terms used in the table be¬ 
low are self-explanatory. 
In recording the contents of a stomach, the relative amount 
of food it contained was noted, that is, whether it was well 
filled or contained but little. Then the various constituents 
were sorted out, and the percentage of the entire quantity 
which each element constituted was carefully estimated. 
The specimens from which the stomachs were obtained were 
of various sizes. Twenty-five were so small, not exceeding 
three centimeters in length, that it was impossible to identify 
them, so they are given in a separate column in the tabula¬ 
tions. Likewise a number of small brook trout (Salvelinus 
fontinalis ), ranging in length from 2.5 to 5 centimeters were 
obtained, and their food differed so much from that of the 
larger specimens of brook trout that they have been given in 
a separate column also. The majority of the other specimens 
from which stomachs were obtained were from 15 to 40 centi¬ 
meters long. 
The following table shows the average per cent of the dif¬ 
ferent elements constituting the stomach contents of all the 
specimens of each species. The averages are based on the 
following numbers of food-containing stomachs:—land-locked 
salmon (Salmo sebago ), 19; greenback trout (S. stomias ), 
64; rainbow trout (S. shasta), 106; small brook trout (Sal- 
velinus fontinalis), 29; large brook trout, 126; fry, 25. 
