DRAGONFUES AND DAMSElvFUES IN PONDFISH CUIvTURE. 235 



If there is more mud than water plants, the best implement is a sieve net like the 

 one recommended by Needham and shown in figure 37a. With this the mud and silt 

 can be drawn ashore and there sifted out, the mud escaping while the nymphs are 

 retained. Much of the material pulled in by the rake can also be advantageously put 

 through the sieve. Needham's sieve net had a framework of steel rods, sides of galva- 

 nized iron, and a bottom of galvanized-wire screen. If something of this sort can not be 

 readily obtained or manufactured, the following may be substituted: Bend a piece of 

 large and stiff wire 3 feet long into the shape of a stirrup, the ends coming together at the 

 center of the curved side. Weld the ends together and insert them with a ferrule into 

 a stout handle 6 or 7 feet long. Make a bag out of bobbinet and fasten its mouth securely 

 to the wire stirrup. The mud will sift through this net as well as through the wire screen, 

 but of course the net can be torn easily. Armed with such a net and a rake, a boy can 

 easily secure several hundred nymphs in a few hours. Needham records that he once 

 collected enough nymphs of Gomphus descriptus to fill a quart fruit jar from Six Mile Creek 

 near Ithaca, N. Y., in an hour's time (Needham and Betten, 1901, p. 453). The 

 only objection to this method of stocking a fishpond is the fact that no one but a scientific 

 expert can distinguish between the different nymphs obtained. They must all be put 

 into the pond together, good, bad, and indifferent; but there are likely to be plenty of the 

 desirable species among them, and the rest can be safely ignored. 



In the summer the pond can be stocked much more intelligently by obtaining the 

 eggs of desirable species and hatching them. In the case of the dragonflies this can not be 

 done by capturing the adults and breeding them artificially or in captivity, as is the unani- 

 mous testimony of all who have made the attempt. 



A. C. Weeks, at that time secretary of the Brooklyn Entomological Society, made in 

 1889 an extensive experiment with Libellula pulchella and Diplax (Sympetrum) ruhi- 

 cundulum by catching the full-grown adults and confining them in the upper story of his 

 dwelling house, which had been cleared of its furniture and arranged with a view to 

 attract the dragonflies. But they would neither feed nor mate nor oviposit. The same 

 experiment was tried later on Anax junius with equally negative results (Lamborn, 1890, 

 p. 78). This was in a crowded city, however, and it might well be supposed that the 

 insects were distracted by the surroundings. 



Fig. 37a. — Sieve net recommended by Needham. 



To test this , the present author experimented with a breeding cage large enough to allow 

 great freedom of motion under conditions that were ideally natural, except for the single 

 element of restraint. This one thing proved a fatal stumbling block, and, although both 

 sexes of Libellula luctuosa, L. pulchella, Erythemis simplicicollis , Leucorrhinia intacta, and 

 Anax junius were placed in the cage at different times, they Avere all immediately imbued 

 with an overwhelming desire to escape. The cage was open at the bottom and was put 

 down over a section of pond 3D containing cat-tails, reeds, algae, and other water plants, 



