BY R. J. TILLYARD. 



507 



which, therefore, we might expect some light on the problem, 

 we are surprised to find that there does not appear to be any 

 basal connection at all between M and Cu in the forewing, 



h 



Text-Fig.44. 

 Cubito-median Y-veiu in hindwings of Megaloptera. a, from Stenosialis 

 sp. (tarn. Sialidae); b, from Archichauliodes guttiferus (Walk.), 

 (fam. Corydalidae). Lettering as on p. 535. 



except by means of an apparent cross-vein situated usually just 

 before the cubital fork, and therefore not- in the position in 

 which one would expect to find M 5 , even if it were reduced to 

 a cross-vein in form. In the hindwing, on the other hand, the 

 Y-vein, though small, is quite completely formed in Archichau- 

 liodes (Text-fig. 44, b) ; and this condition is exactly that from 

 which the somewhat more advanced stage already figured in 

 Stenosialis can easily be derived. 



We have to ask ourselves, then, whether the stage found in 

 the forewing of the Corydalidae is due to the shifting basad of 

 vein M'5, after reduction to a transverse vein, or whether, per- 

 haps. M 5 never existed as a main vein in the forewings of this 

 Order, or whether there may be some further explanation of 

 the case. Remembering how closely allied the Megaloptera are 

 to the Planipennia, we may well turn to these latter for en- 

 lightenment, before we make a final judgment. 



(6) The Planipennia. 



Nearly all of these show no sign of Ms, either in the fore- 

 or in the hindwing. It would be easy to conclude that M 5 

 had never existed in this Order. However, in the course of 

 my study of a fine Triassic Prohemerobiid (28), I came across 

 what appeared to be a distinct vein in the position of M.5. 

 As this fossil, Archepsychops, is clearly allied to the recent 









\ 





CN. 





