OF CERTAIN CHALCIDIDjE AND ICHNEUMONID^E. 75 



and acute, beneath which is a curved fleshy lobe, formed by the union of the dilated 

 maxillae and labium." In this account he is perfectly correct. But while challenging 

 the description given by Reaumur of the head of the larva, he appears himself to have 

 fallen into the singular error of mistaking the rudiments of the future antennae for ocelli. 

 He says*, "The head is furnished with two distinct round points — resembling ocelli," 

 but he makes no allusion to the existence of antennae. No organs of vision exist in any 

 of the parasitic Hymenoptera at the "points" he has indicated in his text and figure of 

 the head of these larvae, the parts referred to being the apices of the antennae of the future 

 imago. I have traced the antennal nerves into these parts in the larva of Ichneumon 

 atropos, which closely resembles in this respect that of Paniscus. The small size and 

 deflection of the mandibles partly account for the difficulty which this larva has in re- 

 attaching itself to the caterpillar when detached from its shell, and the consequent neces- 

 sity for this attachment for its preservation, — premature removal resulting in starvation, 

 as with my first specimen. My second specimen, which left the shell only when matured, 

 I placed on some bight mould, in a covered glass vessel, to observe its changes. I ex- 

 pected to have seen it bury itself in the earth, knowing this to be the habit of the larva on 

 which it feeds ; but in this I was disappointed. Its instincts were more limited and im- 

 perfect than I had imagined. On the day after placing it on the mould, it was lying in a 

 slight hollow on the surface, made by contracting and turning its body, and was in the 

 act of spinning a delicate web of silk, under which it was lying, and where I hoped to 

 have seen it change to a nymph. It attached a few grains of earth to the inside of the 

 web, and between the threads scattered here and there, like the particles of earth inter- 

 woven with its threads by the caterpillar of Mamestra, but after remaining at rest for a 

 couple of days, I found it discoloured and dead. 



Thus checked in my observations, I had no hope of being able to complete this inquiry 

 by tracing the insect to its imago state, until, on the 6th of April, 1848, I found, on exa- 

 mining the earth in a breeding cage, in which I had kept many larvae of Mamestra pisi, 

 to obtain the pupae of that insect, — in several earthen cocoons formed by the caterpillar, 

 — from two to three cylindrical leather-bike cases of a black colour, applied so closely 

 together that their walls formed angular surfaces, precisely bike the cells of a honeycomb, 

 or wasp's nest. There were from two to three of these cylindrical cases in each earthen 

 cocoon of Mamestra. Each cylinder measured about six-tenths of an inch in length. In 

 some cocoons there were three cylinders, and in one only I found four. In the latter 

 instance they were smaller than when the cocoon contained but two, as if the inmates 

 had not been sufficiently fed. This was markedly the case with two of the four speci- 

 mens found together, which were scarcely more than one-half the size and length of the 

 others. Each of the cocoons of Mamestra was completed in the usual mode of this larva, 

 its earthen walls being smoothed on their interior, agglutinated together, and lined with 

 silk ; and each contained, besides the cylinders of the parasitic insects, the dried-up remains 

 of the caterpillar. 



Erom these facts it seems to follow, that the economy of the parasite is this : the parent 

 Ichneumon-fly deposits her eggs on the caterpillar when this is nearly full-grown, and 



* Introduct. vol. ii. p. 147, fig. 76. 14, p. 140. 



L2 



