OF CERTAIN CHALCIDIDjE AND ICHNEUMONID.E. 73 



membranes with its shell, although it is completely apodal and incapable of locomotion, 

 and has to derive nourishment, not by simple endosmosis through its foetal envelopes into 

 the tissues of its body by similar means, but by the direct abstraction of fluid from another 

 living body into its own. To ensure this, the larva requires to be attached by its terminal 

 segments to its shell during the whole period of its growth, and like the larvae already 

 described, it passes no faeces until it has attained its full growth, and becomes detached 

 from its shell, to prepare for its change to a nymph. 



After the second day I was accidentally prevented for some time from making any very 

 precise observations on these insects, further than noticing that they grew rapidly from 

 day to day, and that they retained their connexion with the shell. 



On the 4th of October, the ninth day of their existence, I was enabled to resume my 

 examinations of them, and then found that several of them had perished, and that three 

 only remained healthy and thriving. The caterpillar on which they had fed had become 

 shrunken and wasted, but still retained sufficient irritability and muscular power to 

 contract its body with a quick lateral motion when touched. The larvae that remained 

 healthy were now at least twelve times their original size (fig. 16). The head (16 a) of 

 the larva, which at first was the most ample region of its body, was now the smallest, 

 relatively to other parts which had increased more rapidly, and but little exceeded the 

 size of the ovum. The parasites were attached, one on the dorsal surface and one on each 

 side of the caterpillar, the latter one coiling round the inferior surface of the thoracic seg- 

 ments (fig. 17). Their bodies, enlarged and fattened, were of a dark pea-green colour, and 

 were formed each of fourteen segments, all of which, more especially the anterior ones, 

 were distinctly marked. The stomach in each was in a state of incessant to and fro ver- 

 micular motion, and had become enveloped in a thick tissue of little white follicles, which 

 have been regarded as the rudiments of fat-vesicles. No anal outlet had yet been formed, 

 nor was there, so far as I could discover, any perforated intestine. In the interval of 

 seven days, during which I had been prevented from following up my observations, each 

 larva had three times cast its tegument. This, I believe, is the first time that the soft- 

 bodied Hymenopterous larvae have been noticed to undergo this change. These larvae, 

 therefore, as I have already pointed out, are not exceptions, as they have been suspected 

 to be, to the general rule of development in hexapods, in so far as refers to the casting of 

 the skin ; although, as the body is still connected with the egg-shell, the cast teguments 

 are not entirely got rid of until the larva is detached to become a nymph. The skin is 

 fissured in Paniscus, as in other larvae, along the dorsal surface of the anterior thoracic 

 segments, and is gradually removed from the dorsal, lateral, and inferior surfaces of the 

 head and succeeding segments, by the growth and expansion of the new tegument beneath 

 it, the fissure being extended, and the covering slipped off from above downwards and back- 

 wards, as much by the rapidity of growth in the parts beneath, as by the occasional slight 

 muscular contractions of the segments. When I re-examined my specimens on the ninth 

 day, neither of the cast layers had been removed further than to the posterior segments of 

 the body, which they partially inclosed. The first shed skin (fig. 16 a) was attached to the 

 anal segment, and enveloped this part, surrounded by the shell, and covering the inferior 

 surface. The second (b), within the first, but larger and more corrugated, partially sur- 



VOL. XXI. l 



