142 



THE AMEBIC AX XArURALIST 



[Vol. LV 



fly, than some iclnieumoiis from others.- Equally varied 

 habits exist in at least a few cases even in a single species 

 of ichneumon, for certain Itoplectis may be either para- 

 sites of caterpillars, hyperparasites, or inhabitants of the 

 egg-cocoons of spiders where they devour the contained 

 eggs. From the entomophilous wasp has been developed 

 the parasitic one and we have alluded to its origin as 

 traced by Wheeler ('19). 



From the foregoing, it is seen that we might derive the 

 habits of the wasp from those of the parasite, or vice 

 versa, with but little difficulty, although the more elabo- 

 rate instincts of the wasp appear more naturally as the 

 latter development. 



If now we return to the free-living phytophagous 

 Chalastogastra, it appears for morphological reasons 

 especially that the entomophagous ichneumon flies have 

 been derived directly from them and I think that the 

 transition from phytophagy to parasitism is quite clear. 

 Whether it involves the interpolation of predatism or 

 sarcophagy is perhaps more a matter of conjecture. 



The Siricoidea of the Chalastogastra, on account of 

 their legless, cruciform, lignivorous larvae and reduced 

 wing venation appear to have been derived from some 

 sort of ancestor with a caterpillar-like larva having the 

 more complex wing-venation seen in the saw^-flies or Ten- 

 thredinoidea. So far as is known, no member of either 

 group is parasitic. Until recently the family Oryssidse 

 has been regarded as a degenerate group quite closely 

 allied to the Siricidae. Eohwer has, how^ever, shown that 

 they are really very different and finally (Rohwer, '17) 

 regarded them as a distinct suborder of Hymenoptera. 

 It seems reasonable to suppose that they have Siricid- 

 like ancestors, and as they are now kno^\^l definitely to be 

 parasitic on the larva) of wood-boring Coleoptera, it ap- 



