Dipterous Genus Conops. 289 



nation of its career, exposed to atmospherical vicissitiules, l;ecomes 

 decomposed and disrupted in its connecting ligaments, therfehy 

 affording a ready means of escape to tlie Cunops at the fitting 

 season. 



The question which now suggests itself to inquiry is that of the 

 mode whereby the parasite obtains access to that position within 

 the body of the Ponip'ilus, upon which result the parent Conops is 

 exclusively dependent for the location and sustenance of its future 

 progeny. 



That the eggs or viviparous offspring of the Conops are not de- 

 posed in the larva cells of the Pompilus would seem necessarily 

 to follow from the fact, that at the period when the Pompilus is 

 provisioning its cells in July, — or at any rate between the first 

 appearance of the Conops in the imago state towards the end of 

 June, and the capture of the Pompilus having the larva of the 

 parasite feeding within its abdomen early in August, — the perfect 

 Conops so produced in June from a perfect Pompilus of the pre- 

 ceding year (having hybernated in the pupa state within the 

 inanimate body of its victim) has already found means to introduce 

 its progeny in the body of an imago Pompilus of the succeeding 

 generation, which had hybernated in its larva cell, or as a perfect 

 insect in some other recess, during the same intervening period of 

 pupa-coma in the Conops. 



Under these circumstances it would seem scarcely doubtful that 

 the Conops must find means to inoculate the imago Pompilus with 

 its ovum or viviparous larva, by depositing these in the interstices 

 between the abdominal segments ; which is also in accordance 

 with the observations of M. Robineau Desvoidy and M. Leon 

 Dufour, already referred to, as to the manner in which the Conops 

 rujipes and auripes had been seen pursuing the Bombi for the 

 supposed purpose of depositing its ova. 



It would moreover be contrary to analogy, that any insect in the 

 larva state, feeding upon another larva, should continue so to feed 

 upon its victim after the latter has assumed the imago condition. 

 Any such attacks in fact must involve the destruction of the larvae 

 subjected thereto ; whereas in this case it is essential that the Pom- 

 pilus should attain the perfect state, in order to fulfil the conditions 

 associated with the development of the perfect Conops. 



The entire term allotted to the growth of the parasite in its 

 larva state would appear to be of brief duration, probably not 

 exceeding the limit of from ten to fifteen days ; diminutive speci- 

 mens being found simultaneously with those of the largest dimen- 



