504 THE PANORPOID COMPLEX, iii., 



which extends towards the margin of the wing in the longi- 

 tudinal part of the cubitus, and parallel with the cubital 

 traebea." In other words, both trachea M 5 and tracbea Ciij 

 are present in this family, — a condition which may be taken as 

 more primitive than that mentioned above for Wingia, and 

 found also in many other Lepidoptera. But an examination 

 which I made of the freshly formed imaginal wings of seven 

 Australian genera of Hepialidae (Leto, Charagia, Pielus, Pbrina, 

 Perissectis, Oncopera and Fraus), shows that this condition, 

 though frequently present, is not by any means a constant. 

 Trachea M 5 was found to be frequently of weaker calibre than 

 trachea Cuj ; and there were a few extreme cases in which it 

 had failed to develop much beyond the confines of the upper 

 arm of the Y. In one extreme case (in a specimen of Charagia 

 splendens Scott) it was altogether absent, so that the tracheation 

 of this region of the imaginal wing was the same as that of the 

 pupa . 



Comstock (15, p. 328) suggests that the trachea which de- 

 velops along Cuj in the imaginal wing, via M 5 (his "posterior 

 arculus"), may be the missing trachea M 4 , split back in the 

 same manner that the other tracheaB frequently are in the Lepi- 

 doptera. This suggestion cannot be entertained for one mo- 

 ment, for the simple reason that, in many imaginal wings which 

 I have examined, trachea M 4 is present in its natural position, 

 leaving M 3+4 much further distad along the wing, and des- 

 cending the upper branch of the distal Y-vein, formed by fusion 

 of M 4 with Cu la , which is characteristic of the Order Lepi- 

 doptera. This formation is fully dealt with in Section vi. 



In the great majority of the Lepidoptera, the evolution of 

 the cubito-median Y-vein follows a single definite line, viz. the 

 tendency to strengthen the upper arm, M 5j at the expense of 

 the lower, Ciq. The result of this is that, in all the higher, 

 families, the connection of the main part of vein Ciij with the 

 true cubital fork is quite lost, and this vein is continued back 

 more or less directly, via M' 5 , to the main stem of M. A 

 reference to Text-fig. 91 shows that, in such forms as Wingia, 

 M 5 and Cuj are not yet completely aligned. In the highest 

 types, the venation at the base of the wing becomes greatly 

 thickened, and the serial vein formed from M 5 and Ci^ becomes 

 of great strength, and runs perfectly straight from its base to 

 its secondary or distal forking, as in Euschemon (Text-fig. 100). 



