AAr 



likewise took a large amount of vegetation, amounting to a 

 third or fourth of the whole. 



A group with small mouths, and blunt conical teeth in the 

 throat, illustrated by the common bream or pumpkin seed, was 

 distinguished especially by the number of small snail-like mol- 

 lusks eaten, these making, in my specimens, more than a third 

 of the food. The remainder was chiefly aquatic insect larvae, 

 the medium-sized Crustacea, and water plants. 



The fourth group, illustrated by the croppies, have the 

 mouth long but narrow, and the gill-rakers numerous and long. 

 By these a ie\M fishes are taken, but the food is chiefly insects 

 and the smallest crustaceans — those commonly referred to as 

 Entomostraca, a food resource which they are enabled to draw 

 upon by the straining apparatus in the gills. 



Passing to the pike or pickerel of our western rivers, I find 

 that the common large river pike, Esox Indus, is almost wholly 

 piscivorous, a single specimen only out of the thirty-seven ex- 

 amined, having taken a number of dragon flies. About a fifth 

 of the fishes were sunfishes (half of them croppies) and black 

 bass. Twenty of these thirty-seven pike had taken gizzard 

 shad, which made, in fact, nearly half of the food of the entire 

 group. Minnows were found in only two, and three had eaten 

 buffalo fish. 



The striking features of this record are the importance of 

 the gizzard shad, the abundance of the spiny-finned fishes, in- 

 cluding some of the most valuable kinds, and the insignificant 

 number of minnows and suckers taken. 



The ^^ grass pickerel^' a species which rarely reaches a foot 

 in length, had eaten tadpoles of frogs, and fishes, and insects, 

 the latter making more than a third of the food, and consisting 

 chiefly of larvae of dragon flies. 



T\\& gizzard shad, mentioned above as an especially valu- 

 able element of the food of the higher fishes, feeds itself almost 

 wholly upon mud, with which the long and coiled intestine of 

 every specimen was filled from end to end. This mud con- 

 tained, on an average, about twenty per cent, of minute vege- 

 table debris, with occasionally a little animal matter. 



The great ininnozv family I can scarcely pass by, since it 



