112 
with wnich they come in contact. The body of the aphidid is 
cemented firmly to the object upon which it finally comes to rest. 
The inner abdominal walls of the plant-louse are also lined with silk, 
which firmly adheres to them, and it may be 
that the silk also acts as a tanning substance 
for the body of the aphidid, as the latter be- 
comes leathery and is apparently impervious 
to water ; the old leathery bodies of the plant- 
lice may often be found firmly attached to 
plants after a heavy rain. After the cocoon 
is completed the larva becomes quiet and in 
most cases assumes, according to the junior 
author, a position directly opposite to that 
which it assumed while feeding and develop- 
ing. Figure 22 shows a larva feeding, how- 
ever, in the reversed position; this seems to 
be unusual, the normal position being as 
shown in Figure 21, I. The larva oftentimes, 
on becoming fully developed, is in some way 
dislodged from the body of the aphidid. 
This is probably due to some interference 
while attaching the host to the leaf. These 
cases are quite numerous in badly infested fields and the larva? appar- 
ently never become adult. Figure 25 is a graphic illustration of 
one of these accidents. 
Mr. Kelly found that the pupal stage lasted from 3 to 4 da}^s. 
Fig. 24.— Larva oiAphidius tcsta- 
ceipes spinning its cocoon in the 
dead body of the spring grain- 
aphis, showing the slit or open- 
ing in walls of underside of 
host insect. Much enlarged. 
(Original.) 
Fig. 25.— Larva of A ph idius testaceipcs work- 
ing its way prematurely from the body of 
the spring grain-aphis. (From Webster.) 
Fig. 26.— Full-grown larva of Aphidius 
testaceipcs: a, Lateral view just prior to 
pupation; 6, front Anew of head. Greatly 
enlarged. (Original.) 
Figure 26 shows the larva just prior to pupation. These observa- 
tions were made indoors, during the winter, at the ordinary room 
temperature. It requires from 3 to 5 hours for the Aphidius to 
