298 PARASITES OF GIPSY AND BROWN-TAIL MOTHS. 
Accordingly the percentage of emergence of Parexorista chelonizx 
has aggregated nearly or quite 90 per cent each year as against the 
relatively small percentage of Blepharipa which has been carried 
through its transformations. 
A very few of the flies were liberated in 1907, but there were too 
few puparia of the species received the summer before to make any- 
thing like a satisfactory colony of the species possible. In 1908 in 
excess of 2,000 of the flies issued from the previous season’s importa- 
tions, and of these about 1,500 were liberated in one colony under 
circumstances which were the most favorable that could be 
imagined. The remaining ones were used by Mr. Townsend in a 
successful series of experiments which have already been summarized 
in an account of his first year’s work, published as Part VI of Tech- 
nical Series 12 of the Bureau of Entomology. Other equally satis- 
factory colonies were established later, but of these nothing more 
need be said at this time. 
The large colony liberated by Mr. Townsend in the spring of 1908 
consisted of flies of both sexes, very many of which had mated before 
they were given their freedom. This circumstance, which was not 
considered as particularly of interest at the time, has acquired 
significance more recently, as will be shown. 
Later in the spring and early in the summer caterpillars collected 
from the immediate colony site were found to contain the larve of 
the parasite, and a calculation involving the number of caterpillars 
within a limited area immediately surrounding the point of liberation, 
the number of flies liberated, and the percentage of parasitism pre- 
vailing in this area indicated a very satisfactory rate of increase. It 
will be remembered that the flies were in part ready or nearly ready 
to oviposit when they were given their freedom, so that dispersion did 
not have to be taken into consideration to the extent which is neces- 
sary when a long period elapses between the time of liberation and 
the time of recovery. 
In 1909 similar collections of caterpillars and cocoons were made 
in the same and in nearby localities, and the number of Parexorista 
which was secured from them was gratifyingly large. These coi- 
lections had not been made with the view of determining the rate of 
dispersion, but it was apparent that the increase had been accom- 
panied by a rate of dispersion that was, at the very least, satisfactory 
and which, for all evidence to the contrary, might be phenomenal. 
Accordingly in 1910 a series of collections was planned, some of 
which were to be made in exactly the same localities as those from 
which the flies were recovered the year before and which were 
designed to be indicative of the prevailing rate of increase, while 
others at varying distances and in different directions from the colony 
center were designed to show the rate of dispersion. No doubt what- 
