A STUDY IN HYPERPARASITISM 9g 
They have been allowed to oviposit in cocoons of Apanteles, and vari- 
ous points in their development and behavior have been determined. 
It was necessary to have available a considerable number of cocoons 
that were known to be parasite-free. These were obtained by subject- 
ing gipsy-moth larvae to attack by Apanteles and then rearing these 
caterpillars in covered trays until the Apanteles larvae emerged to 
spin their cocoons. 
- No difficulty was experienced in holding individuals of the various 
species of hyperparasites alive and in good condition for a long time. 
Many were kept for several months. Because many containers were 
required to care for the various isolated females or pairs, it was desir- 
able that these containers be small and easily handled. The ordinary 
4-inch shell vials were found most suitable. These vials were laid 
on the bottom of shallow trays, and were prevented from rolling 
about by the use of cardboard frames. A honey solution, consisting 
of about 40 parts of honey to 60 parts of water, proved to be thoroughly 
satisfactory as food for these parasites. This was supplied on small 
strips of white blotting paper. It is unsafe to use colored paper 
because of the solubility of the dyes. The blotting-paper strips were 
saturated with the honey solution and placed in the vials on pieces of 
towel paper, which had been introduced for the purpose of absorbing 
any excess honey water, thus preventing the vials from becoming 
sticky inside. ‘The parasites were fed once daily, and the same pieces 
of blotting paper were used for two feedings. Both the blotting-paper 
strips and the towel paper were changed every second day. When 
used longer than this during warm weather they molded, because the 
honey solution fermented quickly. The vials themselves were re- 
newed about every 10 days in order to keep the quarters in which the 
parasites were confined perfectly clean. When allowed to remain 
continually in the light the adult secondary parasites ran about in the 
vials almost constantly, and of course became weakened more quickly 
than when kept quiet for part of the time. Accordingly, they were 
held in darkness for a considerable part of each day. They were 
found to oviposit almost as readily in darkness as in daylight. 
Cocoons which were to be attacked were introduced into the vials 
at more or less regular intervals and were left for varying periods. 
When removed they were placed in individual small vials and held for 
the appearance of the adults or until it was desired to dissect the co- 
coons for the determination of one point or another. In order to learn 
how frequently the hyperparasitic larvae molt it is necessary to have 
particular larvae constantly under observation, since it is practically 
impossible, in the case of species so small as most of those studied, to 
be certain of finding all the molt skins by merely dissecting the 
cocoon after the hyperparasite has completed its development. Cer- 
tain individuals of species representing the various taxonomic groups 
concerned in the parasitism of A. melanoscelus were carried through 
their development from egg to adult in the cells on depressed glass 
slides. The larvae of these particular hyperparasites are external 
feeders, and so could be constantly observed when held in this man- 
ner. Hees of the secondaries were removed from the cocoons in which 
they had been deposited and placed upon a cocoon larva of Apanteles 
in one of the glass cells, which was then covered with a piece of thin 
celluloid sealed down with shellac. These eggs hatched, and the 
larvae fed to maturity in a perfectly normal manner. 
soul 2—2/———2 
