238 BULLETIN OF THE BUREAU OF FISHERIES. 
of the remainder indicated at least 2,000 eggs in this one branch. It also was cut 
into short lengths, and these were kept in tumblers of water until they hatched. 
The damselflies will breed and deposit their eggs in captivity as already men- 
tioned. (See p. 236.) 
REARING THE NymMPHS.—The best method of caring for the nymphs is to place 
them in the fishpond as soon as they are hatched. Conditions might arise, however, 
which would render it desirable to rear the nymphs to a certain size before using them 
for fish food. And this can be easily accomplished by supplying them with requisite food. 
The first lot (about 350) of nymphs of L. ductuosa that were hatched in the 
laboratory were kept throughout the entire season in a small aquarium. They were 
fed every two or three days with ordinary tow obtained from the river or one of the 
ponds. They seemed particularly fond of the minute crustacea and devoured large 
numbers of copepods, daphnids, and cladocerans. 
Balfour-Browne fed his newly hatched nymphs upon Parameoecium, which he 
obtained by making an infusion of horse dung in water. One jar of this infusion pro- 
duced for five or six months enough Paramcecium to supply all his nymphs. As the 
nymphs grew in size the Paramcecia were replaced by Daphnids, and in this way he 
carried the nymphs of Agriom pulchellum and Ischnura elegans through from the egg 
to the imago stage. 
Warren placed individual nymphs in separate petri dishes as soon as they were 
hatched and fed them on newly hatched mosquito larve until the third or fourth molt. 
They were then transferred to larger dishes and fed on larger mosquito larve. In this 
way he carried four nymphs of Pantala flavescens successfully through their entire 
life history from the egg to the imago. Two of them were fed daily with large amounts 
of food, and they transformed into the imago stage in about two months. The other 
two were given much less food, and in consequence they required over three months for 
their development. 
EXPERIMENTS IN HATCHING ODONATE EGGS. 
There are but few records of the length of time spent by American dragonflies and 
damselflies in the egg state. Wm. Beutenmuller (Lamborn, 1890, p. 125) stated that 
Libellula auripennis and L. pulchella deposit 25 to 40 eggs each time the female dips 
her abdomen beneath the water. He added on the next page that eggs laid by L. 
pulchella July 23 hatched August 31, making the period of incubation 39 days. In 
L. auripennis the interval was only 8 days, in Plathemis lydia 10 days, and in Diplax 
(Erythrodiplax) berenice and Sympetrum rubicundulum 10 days. 
Among the damselflies Needham (1903, p. 229) said of Lestes in New York State: 
“The eggs, deposited well above the water, develop normally from the first, and in the 
course of two or three weeks attain a condition which is apparently almost that in 
which they will hatch. They then estivate through the remainder of the summer 
and early autumn till the pools are refilled, and the stem and leaves, now dead, fall into 
the water.’ Eggs laid in July and gathered in October hatched within a week in his 
laboratory. But of course such a record would not apply to other genera and species 
where there is no estivation. 
In view of the meagerness of American records, the following experiments at 
Fairport are recorded as a partial guide for the fish breeder, and there are also included 
descriptions and figures of the eggs and newly hatched nymphs to help in identification. 
