276 STUDIES IN AUSTRALIAN NEUROPTERA, iv., 



like, however, to point out that, in my opinion, Ithone stands 

 not very far from the point at which the Megaloptera may be 

 supposed to have diverged from tlie main Neuropteroid stem, 

 and that it is quite possible that the larva may be of a general- 

 ised type, and not possessing the sucking mandibles of the true 

 Neuroptera. Tf this be so, we must perhaps consider the separa- 

 tion of the Order Megaloptera from the true Neuroptera to be 

 unwarranted. 



The character which seems to me to be of the greatest import- 

 ance in this family is the peculiar generalised condition of the 

 radius and its branches in the forewing. Naming the three 

 radial sectors Rg, B3, and R4, respectively, from the most distal 

 backwards towards the base, there can be little doubt that it is 

 Rj, with its course laid parallel to and beneath Rj (the main 

 stem of the radius), and with its numerous subparallel branches, 

 which is the true homologue of the single lis, found in all Neu- 

 roptera except the HeTnerohiidce, DiJaridce, and lihowdcu. The 

 two sectors, R3 and R4, arising closer to the base of R, are 

 strongly suggestive of an archaic formation, which we know 

 occurred in the forewing of the great Protodonate Meganeura 

 (Upper Carboniferous), and which was once probably of frequent 

 occurrence in archaic unreduced types with dense venation. 

 Not a trace of these two sectors exists in recent Odonata, in 

 which the specialisation of the wing-venation appears to have 

 set in earlier, and to have been of a far more drastic character, 

 than we find in the Neuroptera. In the Odonata, the wing is 

 essentially utilitarian, the last word in the development of a 

 magnificent flying type. Tn the Neuroptera, on the other hand, 

 the wing is, if I may say so, purely artistic, a beautiful expres- 

 sion of the development of a symmetrical plan, which conserAcs 

 almost all the archaic features of the insect-wing, and, as a 

 result, is of little value for strong flight. How the two sectors, 

 R3 and R4, have been eliminated in the newer forms, we are not 

 in a position to determine. Either they have been simply sup- 

 pressed during progressive simplification of the venation, or they 

 have passed distad on to the base of Rg. where they would take 

 on the character of branches of the radial sector. In either case, 



