37 
Although many host cocoons have been dissected, an instance has 
never been found where the host larva (and it is the larva apparently 
when full grown which is attacked) was not almost entirely, destroyed, 
the only remaining fragment being the strip of shriveled skin. The 
only exception to this rule which the writer has noticed was in the case 
of an Apanteles cocoon which was examined December 51, 1896. In 
this cocoon was found a single pupa of the Dibrachys, while the per- 
fectly developed Apanteles showed simply a large cavity at the base of 
the underside of the abdomen. The largest number observed to issue 
from a single cocoon of Pimpla annulipes has been twenty-three. Inthe 
instance in which this count was made the specimens emerging were 
counted, the cocoon was cut open and the number of pupal exuvia was 
found to correspond. Successful transformation had been effected in 
every case and the adults had all issued through two holes, one near 
one end of the Pimpla cocoon and the other near the other end. Pimpla 
cocoons were found, however, during the winter of 1896 in which there 
were aS many as twelve exit holes of this parasite, while in another 
instance twenty-two specimens of the Dibrachys issued from a single 
hole in a Pimpla cocoon. This time, however, the hole was near the 
middle of the cocoon. 
A number of cocoons of Apanteles were cut open December 13, 1896, 
and in one case contained two larve of this Dibrachys, which trans- 
formed to pup December 29. The pup were at first milk-white, and 
no darker color, even in the eyes, was observed. Thirty-six hours later 
the eyes had become red, and ten days from pupation (January 8, 1897) 
the adults emerged. From a single cocoon of Meteorus communis the 
highest number of this species reared was six. 
Judging from the large number of specimens which issued at the end 
of October and first of November, and from the fact that the dissections 
of the Apanteles cocoons just referred to in December showed larvie 
and pup, this insect seems to hibernate both as adult and as larva. 
The midwinter condition of this species in the cocoons of Pimpla is 
described in the closing paragraphs of the section on Pimpla inquisitor. 
It occurred both as larve and pup December 15 to 16, and adults 
issued in warm rooms December 27. 
The larva and pupa, as well as the adult, are shown at fig. 18. The 
simple nature of the mouth parts of the larva is especially brought out. 
So far as can be ascertained, the mavxillze are not differentiated, nor is 
there any differentiation of the parts of the labium. The only parts 
observed are mandibles, labrum, and labium. The mandibles are very 
minute and are excessively sharp, looking like minute slightly curved 
spicules. 
A singular and practically inexplicable instance was observed Decem- 
ber 16, 1896, where in a cocoon of the Orgyia a dead female chrysalis 
was found in anadvanced stage of development. The moth was appar- 
ently just ready to emerge at the time of death. All of the scales on 
