16 
of those that have been killed by fungous attack. When its numbers 
are not so great it shows more discrimination and seems to prefer 
half-grown individuals for oviposition. 
This species of Lysiphlebus is parthenogenetic, as was first ob- 
served by Mr. Phillips at Richmond, Ind., and afterwards more fully 
elucidated by Messrs. E. O. G. Kelly and T. D. Urbahns at Welling- 
ton, Kans. The experimental breedings by Mr. Phillips in 1907 
indicated that the offspring of virgin female Lysiphlebus were nearly 
always exclusively males. In a series of upward of 80 breeding 
experiments carried on indoors, in 1908, by Kelly and Urbahns, only 
48 gave results of any kind. In only four of these were females pro- 
duced, the others giving exclusively males. In the 4 exceptional cases 
the females remained virgin, and all finally gave birth to males alone, 
2 with the first generation, 1 with the second, and 1 with the third. 
The mode of procedure was as follows: 
Starting with a mated female, the females from among her off- 
spring were isolated, even before emergence. On their appearance 
these were given Toxoptera reared under cover to preclude parasit- 
ism; the few females from among the second generation were again 
isolated in the same manner, the females in all cases being kept un- 
mated. It was thus found possible to breed a limited number of 
females parthenogenetically to and including the third generation. 
Beyond this all offspring were males, this seemingly being the limit. 
Just why such a large percentage of these experiments should have 
proved abortive is not clear. The conditions under which they were 
carried out were of course unnatural but much more protected from 
the adverse elements of the open field. 
The egg of the Lysiphlebus normally develops to the emerging 
adult in about ten days, during the first six of which the host insect 
remains alive, and at the end of which it commences to take on a 
yellowish hue, the larva of the parasite showing clearly through the 
skin of the abdomen. 
The celerity with which an invasion of Toxoptera is overcome by 
Lysiphlebus is frequently a matter of wonder, as it hardly seems 
possible that this host alone could be the source of the swarms of 
parasites tlfat make their appearance after a few warm days have 
clapsed in the midst of an unseasonably cool spring preceded by a 
winter abnormally mild. 
In order to determine the origin of these myriads of parasites 
Messrs. Kelly and Urbahns began a long series of experiments at 
Wellington, Kans., to determine whether there might not be a mul- 
tiplicity of host species from which great numbers of Lysiphlebus 
would emanate to fall upon and destroy Toxoptera whenever it be- 
comes excessively abundant. The entire failure to introduce these 
[Cir. 93] 
