1920] Cocker ell: Hymenoptera 191 



of development largely to the abundance of insect life. Thus 

 the insects, at first comparatively immune from attack, were 

 hard pressed by enemies on the ground, in the trees and even 

 in the air. The larger, more conspicuous types disappeared,.' 

 giving place to a great variety of rather small forms, which 

 could easily be concealed, or were perhaps in some cases, 

 extremely prolific. In the meantime the increasing diversifica- 

 tion of the flora, with the development of many kinds of woody 

 plants, afforded new opportunities for specialization. The 

 advantages of metamorphosis under these conditions are 

 obvious and no one can doubt that several distinct types of 

 larva began to diverge from the primitive orthopteroid con- 

 dition, in the soil, in the water, and within the tissues of plants. 

 It thus seems to me probable that the primitive Hymenopteron 

 did not possess a caterpillar-like larva, but a boring one, in the 

 manner of the existing Siricoidea. Such a larva would be in a 

 position to take advantage of the newer types of vegetation, 

 and at the same time avoid to a large extent the attacks of verte- 

 brate enemies. The birds of those days had teeth, and the 

 bill as a boring organ, such as we find in the woodpecker, had 

 not developed. This view is supported by the only well-known 

 Mesozoic Hymenoptera, the Pseudosiricidse. These large insects,, 

 as I have lately shown, survived into the Eocene, but are now 

 extinct. They possessed what is to be regarded as a stout 

 ovipositor, probably derived from an orthopteroid ancestor. 

 Their venation does not lend any support to the idea that the 

 basal nervure is a cross-vein; it should rather be a branch of 

 the cubitus fusing distally with the media. From such a type, 

 it is not difficult to derive on the one hand the Ichneumonoid 

 parasites, and on the other the ants. The resemblance of the 

 venation of the Pseudosiricidae to that of the ants is so striking 

 that they were once regarded as gigantic ants. Just as the 

 New Zealand parrot, from feeding on vegetation, has taken to 

 attacking sheep ; so the early Hymenoptera, successful as borers 

 in wood, may have taken to boring in various larvae, securing 

 thereby more elaborated, more concentrated nourishment for 

 their young. The ants, on the other hand, met the problems 

 of life in another way, by developing community organization. 

 The ant larva is a highly specialized being, which has to be fed 

 by the adult. Is there not reason to suppose that it was originally 

 self-supporting, and was therefore in pre-ant times a borer in 



