498 Illinois State Lahoratorij of Natural Histonj. 



same general conditions and at the same time. If perch and 

 catfishes caught in the same haul of the seine show more 

 marked differences in food between the two groups than those 

 exhibited by the individuals of each group among themselves, 

 the probability is considerable that the differences are specific 

 instead of accidental; and such probability becomes greater the 

 greater the number of species found to present corresponding 

 differences under corresponding circumstances. Although it 

 was rarely the case that examples enough of two or more spec- 

 ies comparable as to size and range had been taken at the same 

 time and place to afford a tolerable average of the food under 

 local conditions, yet a sufficient number of such cases was 

 found to give a considerable amount of evidence on this point. 



Thus three specimens of the marbled cat, Amiunis mar- 

 morafus, taken at Peoria, Nov. 1, 1878, had derived nine tenths 

 of their food from Hexagenia larvae, the remainder consisting 

 of leeches and a few spiders; while eight specimens of the 

 large-mouthed black bass, Micropterns salmoides^ taken at the 

 same time and place, had eaten nothing but the young gizzard 

 shad (Dorosoma). 



Comparing the food of four examples of the channel cat 

 (Ictalm^KS pundatus) with seven croppies (Pomoxys), both 

 taken at Peoria, Apr. 10, 1878, I found that aquatic insects 

 made ninety-eight per cent, of the food of the latter, seventy 

 per cent, being Hexagenia larvas, while only sixty-two percent, 

 of the food of the catfishes consisted of insects (ephemerid 

 larvse twenty-eight per cent.), the remainder consisting of 

 vegetation and scraps of dead fishes, 



A contrast equally decided is shown by three specimens of 

 the gizzard shad (Dorosoma) and four of the rock bass (Amhlo- 

 plites rupestris), all obtained at Ottawa, July 8, 1879. The 

 former had swallowed large quantities of fine mud containing 

 about twenty per cent, of minutely divided vegetable debris, 

 while the latter had fed wholly upon insects, fishes, and cray- 

 fishes, — the first chiefly aquatic larvte. 



Even in the shallow muddy pools left behind in the retreat- 

 ing overflow of the Mississippi in southern Illinois, fishes of the 

 same size but differing widely in alimentary structures exhibit 

 corresponding differences in the selections made from the 



