BY R. J. TILLYAKD. 565 



Reviewing the above evidence, we are bound to assume that, at 

 the very beginning of the evolution of the Order Lepidoptera, 

 M 5 had already begun to undergo some amount of degradation 

 from its original condition of a normal main vein preceded by 

 a trachea. That this degradation had not, perhaps, proceeded 

 very far. is indicated by the frequent redevelopment of the 

 trachea M 5 at metamorphosis. The condition seen in many 

 imagines of the Hepialidae, in which two tracheae traverse the 

 vein Cu l5 must be held to be more archaic than the commoner 

 condition, in which only a single trachea, viz. that arising from 

 M via M 5 , is to be found in this vein. The whole line of evo- 

 lution seems to me to be exactly paralleled by the similar 

 development in the evolution of the distal Y-vein, which is dealt 

 with in Section vi. In fact, the normal method of reduction 

 of the number of main veins by fusion, in the Panorpoid Com- 

 plex, is that of the formation of Y-veins of the type discussed 

 in this section. Other examples of such formations are, of 

 course, the fusion of M 3+4 with Cuj in forewings of the Myr- 

 meleontoid groups of the Order Planipennia (in which case the 

 upper arm of the Y becomes the oblique vein), and the fusion 

 of 1A with Cu 2 in the hindwings of Mecoptera, Trichoptera, 

 Lepidoptera, and some Megaloptera. This latter fusion is dealt 

 with under Section v. 



(4) The Diptera. 



In this Order, as in the Mecoptera and Trichoptera, our 

 study is handicapped by the incompleteness of the pupal 

 tracheation. It is not at all easy to obtain pupae of archaic 

 Diptera in the right condition for studying the tracheation of 

 the wings. In the few which I have so far examined, there 

 is no sign whatever of the median trachea; though I have found 

 no less than four distinct trachea (C, Sc, R and Cu) all well 

 developed in the wing of the Bombvliid Comptosia (Text-fig. 

 50). 



In most Diptera, it would appear clear enough that the stage 

 of fusion between M x _ 4 and Cu lt already described as the 

 third stage in the evolution of the Y-vein in the Rhyacophilidae, 

 had already been reached. But some of the more archaic 

 families show definite evidence of an earlier stage of evolution, 

 the upper arm of the Y (M 5 ) being present in the form of a 

 shortened transverse Vein, corresponding fairly closely with 



