458 Illinois State Lahoratovy of Xafural Hisfori/. 



ments. Indeed in only a single instance had the posterior ad- 

 ductor been torn loose. The Unionidse were usually large and 

 thin — probably in most cases Anodonta. 



I have been repeatedly assured by fishermen that the cat- 

 fish seizes the foot of the nioUusk while the latter is extended 

 from the shell, and tears the animal loose by vigorously jerking 

 and rubbing it about. One intelligent fisherman informed me 

 that he was often first notified of the presence of catfishes in 

 his seine, in making a haul, by seeing the fragments of clams 

 floating on the surface, disgorged by the struggling captives. 



Still more interesting and curious was the fact that the 

 univalve Mollusca found in the stomachs of these fishes were 

 almost invariably naked, the more or less mutilated bodies 

 having only the opercles attached. How these fishes manage 

 to separate moUuskslike Melantho and Vivipara from the shell, 

 I am scarcely able to imagine, unless they have the power 

 to crack the shells in their jaws as a boy would nuts, and then 

 to pick out the body afterward. Certainly the shells are not 

 swallowed, either whole or broken. 



The number of mollusks sometimes taken by a single cat- 

 fish is surprising. As high as one hundred and twenty bodies 

 and opercles of Melantho and Vivipara were counted in a spot- 

 ted catfish taken at Havana in September of last year. 



Insects were, however, the principal food of the specimens 

 studied, making forty-four per cent, of all, eaten by twenty- 

 eight of the specimens; five, in fact, had eaten nothing else, 

 and nine others had taken ninety per cent, or more of insects. 

 These were mostly aquatic, although now and then a fish had 

 filled itself with terrestrial specimens. About half the in- 

 sects were Neuroptera, nearly equally dragon-fly larvae and 

 larvas of Ephemeridae; but Hexagenia larvae were rarely recog- 

 nized. Chironomus larvae made thirteen per cent, of the food, 

 and were so frequently taken with the sand tubes they inhabit 

 as to make it certain that they were commonly obtained from 

 the bottom. Leeches appeared in the food of three of the speci- 

 mens, and Gordius in one. Fragments of Plumatella were no- 

 ticed in two, and a fresh water sponge likewise in two. 



Four immature examples of this species, ranging from two 

 and a half to four inches in length, had fed almost wholly 



