195 
one of these Eulophines. He has kindly allowed me to use it in this 
connection. | 
The Chaleidid larve which feed externally on outside feeding larve 
(and we know only one genus—Euplectrus—in which this habit prevails), 
spin a coarse, rough silk, attaching the depleted skin of the host insect 
to the leaf on which it had been feeding, and transform to pup side by 
side in aregular transverse row in the silky mass. Frequently the host 
larva has supported so many parasitic larve that their web attaches the 
entire shriveled skin from end to end, but again they do not occur in suf- 
ficient number to accomplish this result, and only half of the larva skin 
FiG. 19.—Cratotechus sp. a, a, groups of pupz# on sycamore leaf, natural size; b, pupa from side; e, 
same, from ventral side; d, adult female; e, male antenna; f, female antenna, enlarged (original). 
is thus fastened (Schwarz states that with the Cotton Worm and Com- 
stock’s Euplectrus it is usually the anterior portion) and the remaining 
portion hangs down, is doubled back, or breaks off. 
The larve of the closely allied genus Elachistus pupate externally, but 
do not spin the loose silk characteristic of Euplectrus. I have seen the 
naked pupe of Hlachistus cacecie attached by their anal end to the silk 
Spun in its leaf-roll by the larva of Cacecia rosaceana, while the pup 
of H. spilosomatis MS. are found attached in a group among the long 
hairs on the dorsum of the abdomen of the larva of Spilosoma virginica 
(Fig. 20). In the allied genus Miotropis, M. platynote transforms with- 
out its host in the leaf-rolls of Platynota rostrana, as observed by Hub- 
bard (Orange Insects, p. 153). 
EKuplectrus, although it spins silk, can by no means be said to form a 
cocoon, and, therefore, does not form a true exception to the rule that 
the pupe are naked with the Chalcidide. 
The oft-repeated and hitherto accepted observation of Haliday to 
the effect that Coryna clavata does spin a true cocoon would, however, 
