BY K. J. TILLYARD. 



247 



that the forking of R, could have been secondarily acquired, in 

 those genera that possess it, either by splitting of the tip into 

 two (as in Planipennia) or by a veinlet taking on the character 

 of a branch vein during evolution of the pterostigmatic region. 

 The fact that the hind wing has a simple R, in all genera of the 

 family, might be held to support this; but I think the question 

 is not of sufficient importance to warrant an}' prolonged argu- 

 ment. If we agree to overlook it, then we may say that Bel- 

 montia is the first known representative of a new Order from 

 which the 31icropterygida?, and likewise all other existing Lepi- 

 doptera, are undoubtedly derived. Bearing in mind the present 

 distribution of the Lepidoptera Homoneura, it seems quite likely 

 that they did originate somewhere in the Southern Hemisphere; 

 and the fact that Australia is the headquarters for the Hepialidce, 

 while New Zealand is the head-quarters of the Micropterygidce, 

 might suggest that this place of origin was not far removed from 

 Australia. That being so, there is, perhaps, a greater chance of 

 Belmontia itself having been actually in the ancestral line of the 

 Lepidoptera than in that of the Trichoptera. The point of im- 

 portance, however, is that the Lepidoptera must in any case have 

 been derived from some type intermediate between Belmontia 

 and the Homoneura. 



We may, then, legitimately claim that both the Trichoptera 

 and the Lepidoptera are directly descended from the Paramecoptera. 

 The archetypes of the Trichoptera and the Lepidoptera, as at 

 present constituted, contain certain archaic characters not shared 

 by both. For instance, in the Trichoptera, there is the retention 

 of M 4 in the forewing, and the non-development of scales: in the 

 Lepidoptera, the abdominal prolegs of the larva, the retention 

 of the complete pupal tracheation of the wing, the presence of a 

 frenulum in the hindwing, the small but functional mandibles, 

 and the normal maxilke (31icropte7-ygid<K). It is quite clear, 

 therefore, that neither of these Orders can be derived from the 

 other, as Handlirsch plainly showed some years ago (2, p. 1253). 

 Handlirsch, however, derived them both from the older Meco- 

 ptera. We are now able to correct that statement, and to sav 

 that the Trichoptera, and Lepidoptera were undotibtedly derived 



