222 BULLETIN OF THE BUREAU OF FISHERIES. 
came to the laboratory building (100 yards from pond 1D) and picked house flies, May- 
flies, and the like off the window screens. In view of the present widespread movement 
against the house fly conducted by boards of health and hygienists everywhere in the 
United States, this fly-eating habit of the odonates ought to receive every encouragement. 
In this connection also it is worth noting that Dr. G. D. Carpenter, during his inves- 
tigation of the sleeping sickness in Africa, observed one species of damselfly and two 
species of dragonflies feeding upon the dreaded tsetse fly, the damselfly even picking 
them off the clothing of the collector. (Campion, 1914, pp. 498 and 500.) 
Another record in this same paper and one by Poulton (1906, p. 399) credit the odo- 
nates with eating horseflies (Tabanide). 
ENEMIES OF ODONATE IMAGOS. 
Most authors state that the imagos do not suffer much from natural enemies except 
during the teneral period, and this appears to be true. But during this teneral period, 
which lasts for a varying length of time after their emergence from the nymph skin, 
they are so weak and limp that they fall an easy prey to even the humblest enemies. 
1. AccIpENTS.—A small percentage always fails to emerge properly, and in 
collecting exuviz one will occasionally be found with the imago only partially emerged. 
Something prevented it from getting clear of the nymph skin, and it perished in the 
effort. Again, one or two of the wings may fail to expand properly after the imago has 
gotten safely out of the skin, and it is then unable to fly and soon perishes. Sometimes 
the teneral is forced to try its powers of flight too soon, and it falls into the water and 
drowns. The number of these accidents is probably larger than appears at first, for 
such drowned imagos easily disappear. 
Rain sometimes catches the tenerals before they have become sufficiently hardened 
to withstand it. Kennedy (1917, p. 530) makes a note of this: 
With many western species the most serious cause of premature death among imagos seemed to 
be the occasional cold rains which come even in desert regions. On Satus Creek (Yakima County, 
Wash.) I have seen Ophiogomphus severus practically wiped out for the first day or two after a rain and 
regaining its numbers only after more had emerged. 
In the Mississippi Valley a heavy thundershower will sometimes produce the same 
effect upon the tenerals, the rain fairly sweeping them off their perch and drowning 
them in the gutters. 
Usually those which perish in these different ways, however, form but a small 
percentage when compared with the innumerable hosts that pass through the meta- 
morphosis successfully. 
2. Brrps.—Libellula luctuosa emerges mostly in the early morning, and for a long 
time hundreds of teneral wings of this species, easily recognizable by their markings and 
varnished appearance, were found every forenoon lying on the ground and among the 
vegetation on the embankments of the ponds. At length the culprits were caught in 
the very act of seizing and devouring the imagos, and they proved to be English sparrows. 
They flocked to the embankments at daybreak and hunted through the herbage until 
they found a teneral; they then seized it, beat its wings off, and either swallowed it or 
carried it to their young. In this way they destroyed large numbers every day and kept 
it up as long as the species continued to emerge. When its wings are once hardened, 
the sparrow can no longer catch the imago, and it is thereafter free from this enemy. 
