718 THE PANORPOID COMPLEX, iii. 



Plate xxxv. 

 (Order LEPIDOPTERA). 

 Fig. 23. Xyleutes eucalypti (Scott) (fam. Cossidae). Lower distal 

 portion of pupal forewing, to show the branches of M (in 

 middle) and Cui. (x6). A small part of Cu2 is see in the lower 

 right-hand corner. 



Fig. 24. Small part of hindwing of same pupa, much enlarged (x 10), 

 to show the fusion of Sc (the large upper trachea) with Ri_ 

 Note also the simple Rs. Base of wing lies to the right, and a 

 part of M is visible in the lower portion of the photograph. 



Postscript, added 30th September, 1919 : — During a recent visit 

 to Jervis Bay, 1 carried out some researches which have strength- 

 ened certain conclusions come to in this paper. These may be 

 briefly summarised as follows : — 



(1) A native Cockroach of the genus Escala was found very 

 commonly under bark, and a number of specimens were taken 

 which were still white and soft after undergoing an ecdysis. The 

 wing-sheaths of these were dissected off, and the tracheation care- 

 fully studied. The result is that there can be no doubt whatever 

 that the vena dividens, determined by Comstock as vein 1A fused 

 with Cu, is in reality vein Cu2. Thus the Cockroaches fall into 

 line with the insects of the Panorpoid Complex, as well as with 

 the Cicada and the Psocids, already mentioned in Section iv. of 

 this paper. 



(2) Further research upon the wing-tracheation of the Hes- 

 periidae was carried out by obtaining pupa? of the well-known 

 Skipper, Hesperilla picta Leach, in fine condition, and dissecting 

 out their wings. This insect belongs to the Subfamily Trapez- 

 itinae. The results show that its pupal tracheation agrees in 

 every imjjortant particular with that of Euschemon rafflesia 

 already dealt with fully in this paper ( Section xiv. ) . The only 

 important difference is to be found in the imaginal venation, the 

 vein JV^ failing to chitinise in the hindwing of H. picta, wmereas 

 it does so in the hindwing of E. rafflesia. We may therefore 

 conclude from this, (a) that Euschemon is a true Hesperiid, and 

 (b) that the Butterflies as a whole are a monophyletic group, 

 the first dichotomy of which resulted in two lines of evolution, one 

 leading to the Hesperiidae, with Euschemon as an archaic off- 

 shoot from near its base, and the other leading to all the rest 

 of the Butterflies. It is clear that no existing type of Hesperiid 

 represents the ancestral type of the Butterflies; but, on the other 

 hand, Euschemon itself has not departed far from that lost an- 

 cestral type. R. J. T. 



