BY R. J. TILLYARD. 



675 



Doratifera longerans. There can be no doubt here that trachea 

 M] has become permanently attached to Rs. The dotted lines 

 indicate the position of the veins closing the basal cell in the 

 imago. The only changes at metamorphosis are the much more 

 complete fusion of Sc and R p — a fusion which also involves a 

 part of the base of Rs, — and the drawing together of the two 

 main branches of M, so that a single vein M traverses the cell 

 from one end to the other. 



The basal Y-vein formed by 1A and 2A in the hindwing is 

 absent, in Doratifera. A reference to the pupal tracheation 

 shows that this must have been brought about by abortion of 



.Rs 



Text-Fig. 96. 

 Tracheation of pupal hindwing of Doratifera longerans (Walk.), (fam. 

 Limacodidae). (x 9|). Lettering as on p. 535. 



trachea and vein 1A, since the only trachea present does not 

 arch upwards towards Cn 2 and must therefore be 2A. 



The tracheation of the pupal wings of Apoda agrees closely 

 with the above, but trachea Sc in the hindwing is shorter. 



Of the Zygaenidae I have, unfortunately, been unable to ob- 

 tain any pupa? in a suitable state for dissection. The Megalo- 

 pggidae do not occur in Australia. But it has been shown by 

 many authors, without any doubt, that both these families are 

 closely related to the Limacodidae; so that we may reasonably 

 assume that the pupal tracheation will not differ very greatly 

 from that of this latter family. 



The Psychidae appear to be the most highly specialised family, 

 because of their lack of scales, their extraordinary life-history, 

 and the great differentiation between the small, swift-flying 

 males and the huge, larviform, wingless females. The venation 

 is correspondingly specialised, M 1 having apparently been com- 



