DRAGONFUES AND DAMSELFIJES IN PONDFISH CULTURE. 203 



Of course, the snails crawl about over these sponges, as well as over the algae. The 

 nymphs probably took in the spicules while feeding on these insects and snails, for they 

 would hardly eat the sponge itself. The Libellula nymphs had eaten a much smaller 

 number of mayfly larvae than those of Anax and were content with the Amphipod 

 Hyalella, in place of crayfish. 



Twenty of these nymphs were taken from pond 7 D, June 23, 191 6, while the others 

 came from ponds 1,2, and 3 D, at different dates in July and August. Pond 7 had been 

 stocked in the spring with 75,000 small fry of the buffalofish; on July i these young 

 fish had reached a length of i inch, and on July 1 5 specimens were taken an inch and a 

 half long. Previous to June 23, on this basis of growth, the fish were small enough to 

 be caught and eaten by the nymphs, if the latter had made the attempt. The entire 

 absence of fish from the diet of these nymphs shows that they chose other food even 

 when fish were present. The remains of odonate nymphs in so many of their gizzards 

 is good evidence, on the other hand, that the luctuosa nymph is not inert and lethargic. 



The food of the Erythemis nymphs consists also of snails and entomostraca, with 

 a moderate amount of beetle and mayfly larvae, and an almost complete absence of the 

 larger Crustacea and of damselfly nymphs. A large amount of algae was present in 64 

 per cent of the gizzards, and in several individuals nothing else had been eaten; at least, 

 there were no remains in either gizzard or intestines. Algae, therefore, must constitute 

 a respectable portion of the food of these nymphs. 



It might seem strange to report Simulid larvae from a fishpond, but the screens at 

 the outlets of all of the ponds were covered with the larvae of Simulium vittatum, and 

 some of the nymphs evidently picked them off the screens. The ephippium surrounding 

 the Cladoceran egg is proof against the digestive juices of the nymphs, and eggs taken 

 from the posterior end of the intestines were as plump and uninjured as those just 

 swallowed. The Desmids and Diatoms, like those found in luctuosa nymphs, are not 

 numerous enough to make it certain that they were really sought for and eaten. They 

 might well have been taken in accidentally with some of the food. Ten of these nymphs 

 were taken from pond 7 D on June 26, 191 6, and 15 were taken from pond 9 D on July 7. 

 As already stated, the former pond contained an abundance of small fish, while the 

 channel cats in the latter pond had produced a brood of fry previous to the capture of 

 the nymphs. None of them had eaten fish, however, and their small size makes it 

 improbable'that they could overpower any but the smallest fry. 



Kntomostraca and copepods are the chief articles of diet for the Pachydiplax nymphs, 

 and there is a minimum amount of snails, beetle larvae, and mayfly larvae. The larger 

 Crustacea are practically absent, and damselfly nymphs are the only odonates repre- 

 sented. The food in the gizzards of these nymphs and that in the Erythemis nymphs 

 was particularly well ground up, so that only small fragments were left. Even the 

 shells of the small entomostraca were broken and resembled the debris obtained from 

 the posterior intestine of Anax and Libellula. 



The damselfly nymphs included species of Enallagma, Ischnura, Argia, and Lestes, 

 and were obtained from the various ponds indiscriminately. No attempt was made to 

 separate the different species, and they were treated as though all one kind. The 

 food in the gizzards of these numphs was broken up into smaller fragments than that 

 of any of the dragonflies, and in much of it the identification of species or even genera 

 was almost impossible. As will be seen the great bulk of the food consisted of ento- 



