The Food of Fishes. 33 



Insects are also an important item, — amounting to twenty- 

 four per cent., nearly all being the larvae of Neuroi^tera, — 

 May-flies (Ephemeridae), dragon-flies and case-flies (Phry- 

 ganeidae). A single specimen from Peoria Lake had eat- 

 en one small fish — a "darter" of the genus Poecilichthys. 



The second group, twelve specimens from Lake Mich- 

 igan, presents a curious and instructive contrast in food to 

 the foregoing. Mollusks and insects wholly disapjjear, 

 and Crustacea are limited to the commonest crawfish of 

 the lakes ( Cambarus virilis, Hagen), which forms fourteen 

 per cent, of the food. The remaining eighty-six per cent, 

 consisted wholly of fishes, all minnows (Cyprinidas) so far 

 as recognized except one, and that was some undeter- 

 mined percoid, — probably itself a perch. 



It will thus be seen that the common perch has a food 

 history of three periods, — the periods of infancy, youth, 

 and mature age. In the first it lives wholly on Entomos- 

 traca and the minutest larvae of Diptera; in the second, 

 commencing when the fish is about an inch and a half in 

 length, it takes uj) first the smaller and then the larger 

 kinds of aquatic insects in gradually increasing ratio, the 

 entomostracan food at the same time diminishing in im- 

 portance ; and in the third it appropriates, in addition, 

 mollusks, crawfishes and fishes, — in the lake specimens de- 

 pending almost wholly on the last two elements. 



We have here the first instance of a fact which we shall 

 see again and again illustrated, — that the young, having 

 at first an alimentary apparatus too small and delicate to 

 dispose of any insects but the minutest larvae, live almost 

 wholly on minute crustaceans. 



It is proper to note that the lake and river perch are by 

 some good authorities regarded as separate species, — the 

 latter being much more highly colored than the former. I 

 have not found so strict a separation of the two forms as 

 that described by Mr. E. W. Nelson, but have frequently 

 taken both in the same haul of the seine in different parts 

 of Calumet R. and in Lake George, Ind. — a body of water 

 communicating with Lake Michigan by an outlet three or 

 four miles long. Occasional pale specimens are also taken 

 far from the lakes, in the Fox and Illinois rivers. The 



