OF WASHINGTON. 147 



the very numerous pteromalid parasites of larger Lepidoptera, 

 also with many of the forms of the family Eulophidae as denned 

 by Ashmead, and especially perhaps in the subfamilies Ente- 

 doninae and Tetrastichinae, and some at least of the Elachis- 

 tinae, we may hope to find this mode of development. We 

 have only to remember the manner in which a lepidopterous 

 chrysalis is sometimes packed with parasites of the genus Cir- 

 rospilus, or a large ichneumonid larva with Dibrachys, to see 

 the force of this suggestion. 



Apanteles among the Braconidae also naturally occurs to 

 one, and the great number of larvae issuing from a large sphin- 

 gid larva is readily explicable by polyembryony, and this offers 

 an easy and important field of investigation. For many years, 

 the speaker said, he had had parasitized sphingid larvae in dif 

 ferent stages upon his desk awaiting the opportunity for dis 

 section, but the chance had never come. 



With the proctotrypids, Marchal has already shown us what 

 occurs with Polygnotus and also what is to be expected with 

 other forms of the subfamily Platygasterinae (and here the 

 speaker referred to a published announcement in the Bulletin 

 of the Entomological Society of France for October, 1906, of 

 the publication of Marchal's paper upon this group) ; and other 

 subfamilies will bear investigation. 



That polyembryony is a highly specialized function, and with 

 insects one of the concomitants of parasitism, seems likely. It 

 is therefore not to be expected in groups in which the parasitic 

 mode of life is not old and firmly established for a very long 

 period of time. With the parasitic Diptera, therefore, where 

 from the non-specific character of the host relation it is to be 

 assumed that the parasitic mode of life is of comparatively 

 recent acquirement, polyembryony is not to be expected to 

 exist. The large number of Tachina eggs with which a lepi 

 dopterous larva is frequently plastered is an indication of this. 



In the discussion of the paper Mr. Reeves stated that he 

 had found larvae of the Hessian fly filled with the parasitic 

 larvae and pupae of Polygnotus, upon which the transparent 

 skin of the host larva had shrunken so as to lose its own form 

 and reveal nothing but the contained mass of parasites. The 

 largest number of parasites found in a single host larva was 

 41, the usual number ranging from 6 to 12. Doctor Stiles 

 remarked that the figures of Marchal were very suggestive of 



