THE PARASITES OE POPILLIA JAPONICA 29 
and later develop into an imperfect beetle. Another succeeded in 
freeing itself from a second-stage parasite larva 3 mm. in length, 
but later died from mechanical injuries incident to the withdrawal 
of the larva through the aperture in the dorsum of the thorax. It 
is thus seen that the attachment of the parasite by means of its 
respiratory funnel to the derm of the host larva is very strong, since 
it is able to withstand the strain exerted by the drawing of the body 
through a relatively small opening in the newly formed derm of the 
pupa. 
First-stage larvae from the adults of the first generation were used 
in a series of experiments upon grubs of Popillia japonica. All of 
the latter which were available were in the advanced third stage and 
preparing to pupate. An excessive number of planidia effected 
entrance into each host, as was evidenced the following day by the 
many respiratory funnels visible underneath the derm. Within a 
very few days most of the grubs pupated and the parasite larvae 
were able to sever their connection with the respiratory funnel, and 
remained with the caudal end of the body protruding externally. A 
few maintained the attachment and held the cast skin in position 
upon the pupal body. 
Some of the parasitized larvae died from injuries incident to ex- 
cessive superparasitism, and in these instances the third-stage parasite 
larvae detached themselves from the funnel and forced the caudal 
end of the body out through the aperture in the derm. These 
host grubs then assumed a somewhat mummified condition, but this 
may be attributed to bacterial and fungus action rather than to any 
development in relation to the parasite itself. Feeding was con- 
tinued on the dead body and pupation took place normally and just 
outside the host remains, which had not been entirely consumed. 
The necessity, under Korean conditions, for these different hosts 
within which the successive generations may develop is due to the 
fact that the larvae can reach maturity only in grubs which them- 
selves are undergoing histolytic action preparatory to pupation. 
This physiological change seems to provide the necessary stimulus 
for development, a condition which is known to exist in respect to 
many other parasitic insects, and without which the parasite larva 
either remains in the early stages or dies. The writers' observations 
indicate the above state of affairs to exist in all the species of Dexiidae 
studied; and, although the time of pupation of Prosena would seem 
to contradict this assertion, the fact is that the presence of the 
parasite larva in the grub body tends to prevent actual pupation but 
permits the host to live several weeks or a month beyond the time of 
normal transformation. 
Habits of the adult. — Mating has not been observed in the field, 
but presumably takes place shortly after emergence. The males 
appear a few days prior to the females. In the laboratory mating 
was induced very readily in glass vials or test tubes of 1-inch diameter, 
and the duration of copulation was normally 10 minutes. 
Feeding also has never been observed under field conditions, in 
spite of the extensive observations made throughout the periods of 
adult abundance. Unlike Prosena, which has a long, lancetlike 
proboscis fitted for penetration to the deepest nectar glands of cer- 
tain blossoms, this species has a very short, fleshy proboscis and is 
therefore probably a feeder upon honeydew of aphids and other 
