AND ANATOMY OF CERTAIN HYMENOPTERA PARASITICA. 389 



and go on feeding till full-grown. If these spring larvse are opened up they 

 are often found to contain a nuu)ber of parasitic larvae which lie in the 

 hsemocoel. Their number varies from fifteen to sixty. If numbers of the 

 caterpillars be kept in a box and fed on their food-plant, it will be found 

 that some of them give rise to the cocoons drawn in PI. 24. fig. 5, at C. 



It is rare for any caterpillars to get as far as the spinning operation, for 

 the parasites inside their bodies choose this moment to bore their way outside 

 the body of their host and to begin preparations for pupation themselves. 

 In the closely allied genus Apavteies {glomeratus), a parasite on the 

 " cabbage-white " Pieris hrassicce, I have watched this process. Towards 

 the end of their larval life the parasitized caterpillars become ^'sleepy," and 

 can easily be distinguished from their fellows because of their behaviour. 

 I received a ^' sleepy^' pierine caterpillar at ten o^clock one morning; up to 

 two o^clock in the afternoon this caterpillar remained quiescent. Soon it 

 began to squirm and move about in a characteristic manner, ('linging on 

 by means of its hindmost prolegs, it slowly bent its body backwards and 

 forwards in the form of an arc ; gradually this movement became less 

 violent, and from the sides of its body were seen appearing numerous small 

 white points (text-fig. 4). These were the heads of the parasitic Apanteles 

 larvfe, which latter had grown at the expense of the caterpillar. The host 

 insect, clinging firmly to the branch, seemed soon incapable of anything but 

 the smallest undulating movement of its body. By the time an hour had 

 passed the parasites were mostly outside the body of their unfortunate 

 victim. The latter, when prodded with a pin, was found incapable of much 

 movement and would soon have died. The subsequent spinning of the 

 cocoon by each parasitic Apanteles larva was not watched by me, as the 

 caterpillar was killed with the parasites adhering to its body, as in text-fig. 5 

 at x. Groureau (6) remarks that a caterpillar of Noctiia oUracca had nourished 

 sixty Microgaster larvse, and after being pierced by sixty holes to allow 

 the exit of these larvoe, it still lived over a week, but it was incapable of 

 walking and only made the. smallest movements to show that it still lived*. 

 According to Goureau (6, p. 360) the parasitic larvse escape from the body 

 of their host by making holes on each side of the line along which lie the 

 stigmata. I found no such regularity, though most of the parasites did bore 

 out through the sides of the caterpillar (text-fig. 5, .i'). 



After the larvse of Microgaster connexus have made their way out of the 

 body of their host, the caterpillar of Porthesia simiUs, they spin their cocoons 

 quite near the shrunken skin of their victim, as in PI. 24. fig. 5. In other 

 forms (as in the Microgaster sp. described by Marshall (2)), the cocoons stick 

 separately over the body of the host. Such objects as the caterpillar skin 

 and cocoons of its parasites are common to all those who have bred cater- 

 pillars. The cocoon of Microgaster connexiis (PL 24. fig. 1) is of white or 

 dirty white silk and is very strong and tears like rough parchment ; 



* See Eutomol. Month. Mag. vof. v. p. 19. 



28* 



