236 BULLETIN OF THE BUREAU OF FISHERIES. 
everything being left undisturbed. The upper portion of the cage was covered with 
chicken wire whose meshes were large enough to allow free access to the insects upon 
which the dragonflies usually fed. But as soon as the dragonflies found they were shut in, 
they would neither alight on the water plants, nor touch the insects that flew through the 
cage, nor mate, nor oviposit. Instead, they spent their time beating their wings against 
the sides and top of the cage until they became exhausted and fell into the water, or were 
picked out through the meshes of the wire by red-winged blackbirds, which were 
attracted by the beating of the insects’ wings. 
While the dragonflies are thus averse to breeding in confinement the damselflies 
take kindly to it, and Balfour-Browne (1909) succeeded in obtaining a large number of 
eggs from specimens of two English species, Agrion pulchellum and Ischnura elegans, 
which he kept for a week ina cage. He fed the captives on flies and other insects cap- 
tured in a sweeping net and placed alive in the cage. The eggs hatched into nymphs 
in about three weeks, so that he was able to determine definitely the period of incubation 
for these species. ‘This method was not tried on our American damselflies, but there is 
little doubt that it would work as well with them as with the English species. It was 
found absolutely necessary to keep the cage in the bright sunlight, because in the shade 
and on dull days the damselflies became torpid and simply clung to the sides of the cage. 
It is evident, therefore, that nothing can be accomplished by endeavoring to breed dragon- 
flies in this manner; their natural habits and instincts are against it. And although the 
damselflies are more susceptible to captivity, this artificial breeding is a rather laborious 
process for any except the scientific expert who wishes to absolutely isolate a given 
species. 
Fortunately there are other ways of accomplishing the desired results, and these 
prove to be highly successful. As already stated, many female dragonflies deposit 
their eggs loosely in the water or upon the floating algee by hovering close to the surface 
and touching the water at intervals with the tip of the abdomen. If such a female be 
caught as she comes down to the pond to lay her eggs or while she is ovipositing and one 
pair of wings be folded together over the back and held between the thumb and fore- 
finger, leaving the other pair free, she will continue to lay eggs in large numbers if the 
tip of her abdomen be dipped in water in a convenient tumbler, basin, or small jar. 
Tillyard (1917, p. 358) claimed that it was ‘‘necessary to have the water dirty, with 
mud, sand, or small pieces of débris for the eggs to fall upon; otherwise the eggs will 
simply all stick together and quickly go moldy.” While agreeing that the presence of 
dirt is a positive advantage in the way that Tillyard suggests, the present author can 
not agree that it is always ‘‘necessary.” All the experiments hereinafter recorded were 
made in clean water, and while the eggs usually did stick together, they did not mold 
except in a single instance, and even then practically all of them hatched. However, 
if the nymphs are to be used simply to stock the fishpond, some dirt and débris are 
desirable, since then the conditions are more nearly natural. Similarly, when the 
male accompanies the female during ovipositing, on being captured the female will 
deposit her eggs freely in the tumbler or basin; this applies to such species as Tramea and 
Celithemis. 
Needham recorded a female Gomphus graslinellus captured while ovipositing, 
from which he “obtained in a tumbler of water an immense number of eggs’’ (Need- 
ham and Hart, rgor, p. 69). Also a female Gomphus externus, “captured in the weeds 
