RILEV NOTES ON NORTH AMERICAN MICROGASTERS. 3OI 



as the work proceeds ; and finally, to complete the enclosure of 

 the cocoon, again reversing its position so as to appear in the atti- 

 tude shown in Fig. 6, ". Mr. Wm. A. Buckhout (ibid. p. 752) 

 notes several points of difference in the formation of similar 

 cocoons by the species infesting Macrosila. This Microgaster, 

 after completing one-half of its cocoon and reversing its position, 

 flexes its body so as to bring its head back to the top of the cocoon, 

 "from which it spins downwards till the back of its host is reached 

 and it is entirely enclosed," the spinning of the two halves of the 

 cocoon being upon practically the same plan, the first half being 

 spun from below upwards, the second from above downwards. 

 In lining the cocoon the larva repeatedly changes its position. 



In the case of Apanteles aletice^ which spins its cocoon not in a 

 perpendicular position as do those on the Sphinges, but recumbent 

 upon the cotton-leaf, the outer covering being somewhat more 

 ^ , loose and "fuzzy," Mr. II. G. Hubbard, in 



~^ - t";:^^» ^ a recent report to me, describes the process 



T:: /^^^fl as follows. "In quitting its host the para- 



W^ ^ ^f "i site maintains its connection therewith by 



jl^,'^' ^ means of a single thread. After crawling 



y~^~ \ -Y~M to a distance of about half an inch, it fas- 

 4 (/-XP\/^V~'\^^ ^^"^ '■^"^ *^'^ thread to the surface of the leaf 

 ^ C-X^^Xv/wC/^ begins its cocoon. The larva forms 



— ^^^^^^^^^3^*^=-- the exterior by throwing out loops of ropy 

 Fig. 7.— Commencement of CO. flu i(j which undcr the lens are seen to 



coon by lar\a of Microgaster. 



(AfterMarshaii.) bccome rigid as they fall, and to harden 



rapidly, forming a rather coarse strand of white silk, which is 

 often beautifully furred. These loops are piled one upon another, 

 and the walls of the cocoon rise rapidly until they meet overhead. 

 The inside is then lined in the manner usual with Lepidopterous 

 Iarv£e, until the whole has become opake. The process of spin- 

 ning occupies about two hours' time." 



It is in this as in almost every other genus or group, verv diffi- 

 cult to separate species, so called, when a large number of indi- 

 viduals or abundant material is studied. Some of the forms 

 indicated as good species by entomologists, especially if de- 

 scriptions were based upon one or two individuals without refer- 

 ence to the variation that occurs, are sure to prove to be only 

 ■varieties. It is quite natural to infer that differences in the host. 



