HYMENOPTERA. — ICHNEUMONID. 149 
membranous texture, and each of'them formed of a distinct elongated 
cap at each end of the cocoon. This formation was found to exist in 
several of the cocoons. Some cocoons are entirely white or yellow, and 
composed of glossy silk which can be wound off; but others are 
variegated and banded with black, of which Réaumur has figured va- 
rious species (Mém. tom. ii. pl. 35. and 37.), minutely describing 
the process of formation. Other cocoons are suspended by a fine 
thread to leaves or twigs, sometimes three or four inches long* ; 
and Réaumur, who described two kinds, has observed that one of 
them is evidently the parasite of the processionary caterpillars, 
as he found them by dozens in the neighbourhood of the nests of 
those caterpillars ; and that the other cocoon, when detached from the 
twig, sprang to a distance of several inches, the inclosed larve pro- 
bably contracting its body, or perhaps bringing the two extremities of 
the body together, and then suddenly letting them go, in a manner 
similar to the motions of the common cheese-hopper. De Geer ob- 
tained larvae of a species of Perilitus (Zele Curtis), which formed 
suspended cocoons from the caterpillar of Zygzena filipedula (Mém. 
tom. ii. tab. 44. f. 11. 13.); and Curtis reared a species of Perilitus 
(P. pendulator Zatr., Ephippium Curt.) from a cocoon suspended 
from the nut ( fig. 76. 18.); and see Latreille’s Memoir on Ichneumon 
pendulator, the last-named species in Bull. Soc. Philomat. 1799, tom. 
ii. p. 138. Some few species, however, especially amongst the Adsciti 
do not construct cocoons when their peculiar habits render this unne- 
cessary. Such, for instance, is the case with the Aphidii, which 
undergo their transformations within the indurated skin of the Aphis, 
of which they have devoured the interior. The spherical shape of the 
case thus formed accords with the curved attitude of the full-grown 
larva, and of the pupa developed from it (Haliday, Hint. Mag. vol. ii. 
p- 225.). In the majority of the true Ichneumonide, Braconidee Stephens, 
and others, —the pupa is not bent double, and the cocoon is more elon- 
gated. Ratzeburg figures the pupa of Hemiteles luteator as bent double ; 
and this is probably the case with all those species which oyiposit by 
extending the tip of the body beneath the fore-legs. In the Mag. 
Nat. Hist. (vol. viii. p. 171.) I have described one of the Adsciti 
(Chenon nigricans) produced from pupz found in the sheathing leaf 
of the flowering stalk of the common barley. The pupa (jig. 76. 19. 
* Réaumur has made no observation as to the mode of construction of this curious 
cocoon. It seems, however, most probable that the larva having first suspended itself 
by a thread, commences the formation of its cocoon whilst remaining suspended. 
t 3 
