BY R. J. TILLYARD. 5(3 



Trichoptera and Megaloptera. Examining the preceding pupal 

 tracheation, we find that this also agrees exactly with the 

 Megalopteran scheme. Where all the evidence from fossils, 

 tracheation and venation alike, agrees exactly, where is the need 

 to invent an explanation such as that given by Comstock, the 

 only merit of which, as far as I can see, is apparently, but not 

 actually, to bring the venation into line with Comstock's hypo- 

 thetical type. For it should be noted that, in making the cubitus 

 two-branched, to conform with this type, Comstock transfers 

 the primary cubital fork to the position of the secondary fork, 

 far distad along the cubital vein ; yet he makes no effort to 

 explain how it comes about that such a high specialisation 

 should have taken place, within the limits of a considerable 

 number of Orders. 



Further proof of the correctness of the above evidence may 

 be obtained outside the limits of the Panorpoid Complex, by 

 examination of the nymphal wings in the Hemimetabolous 

 Orders Psocoptera and Homoptera, in which the cubitus has 

 been interpreted as being two-branched by Comstock. I have 

 examined the last three consecutive instars in Psocus, and the 

 last two in a Cicada; and in all these instars the supposed 1A 

 is truly a branch of Cu. If it were really 1A, we can scarcely 

 suppose that so early a stage as the antepenultimate nymphal 

 instar. in an ancient Order like the Psocoptera, would show the 

 specialised fusing already complete. If it did, then our faith in 

 the value of the Comstock-Needham System must be completely 

 shattered, seeing that there would be no telling whether, for 

 instance, Rs might not be really part of M, or M two separate 

 veins fused together. It would only be necessary to explain 

 that a fusion had taken place anywhere, and the venational 

 scheme could be altered ad infinitum! 



The evidence, then, seems to admit of no doubt whatever. 

 Within the Panorpoid Complex, the cubitus is three-branched in 

 the Orders Paramecoptera, Protomecoptera, Megaloptera, Tri- 

 choptera and Lepidoptera; while in the Mecoptera, Paratrichop- 

 tera and Diptera it is only two-branched. In all the above 

 Orders, the true primary cubital fork is the more basal forking 

 of that vein. The extra fork, when present, belongs to Cui 

 and may or may not have been caused by incomplete distal 

 fusion between M 5 and Cu x when the Y-vein of the arculus was 



