PARARTTES TTTBERNATTNG IN BROWN-TAIL WEBS. 



265 



canal, its axis parallel to the axis of the body of the host caterpillar, 

 and its head in the ultimate or penultimate body segment, and 

 pointed toward the rear. 



The larva? of Zygobothria are similarly assigned to a definite 

 position, and in otherwise healthy caterpillars have invariably been 

 found embedded in the walls of the crop, as indicated by figure 63. 

 In appearance they are typical of the tachinid first-stage larvae gen- 

 erally, and with no extraordinary points of difference from most 

 others of the group to which they belong. Those of Compsilura 

 (fig. 64) may be found in similar positions, but they are easily dis- 

 tinguishable from Zygobothria by the presence of the three chitinous 

 anal hooks or spines, as indicated in the accompanying figure. 



Nothing is known of the hibernating stage of Mesochorus. It does 

 not seem probable that it should resemble the planidium of Perilam- 

 pus, which, like Mesochorus, is a secondary parasite which gains 

 access to its host before the latter has left the body of the caterpillar 

 which harbors both primary and secondary. It is 

 presumed that -it will be representative of a highly 

 specialized type of development which fits it for 

 the peculiar role which it plays, but that this de- 

 velopment will have been along wholly different 

 lines from that which has taken place in the case 

 of Perilampus. The whole genus, apparently, pos- 

 sesses habits similar to those of Mesochorus pattipes. 

 A very beautiful and, according to Mr. Viereck, an 

 undescribed species has been reared from the 

 cocoons of Apanteles fiskei, parasitic upon a species 

 of Parorgyia, under circumstances which indicate 

 positively that attack was made while the pri- 

 mary host was still alive. The same may be said 

 of another undetermined species which has simi- 

 larly been reared from Apanteles hyphantrise. 



Mesochorus pattipes is not an uncommon parasite of Apanteles 

 lacteicolor Vier., having been reared from only a few among the many 

 localities from which its host has been secured in numbers, but the 

 average proportion of parasitized individuals has been only about 

 2 per cent. 



The interrelations of these several parasites thus closely associated 

 with one stage of the same host, and consequently with each other, 

 are interesting and peculiar. Pteromalus, of course, cares little 

 whether the host caterpillar selected for attack is parasitized by 

 one or more of the endoparasites which hibernate as first-stage larvae. 

 The female will undoubtedly attack parasitized as freely as it will 



Fig. 64. — Compsilura con- 

 cinnata: First-stage lar- 

 va. Greatly enlarged. 

 (Original.) 



