﻿March 1895] DyAR. CLASSIFICATION OF LePIDOPTERA. 25 



As regard the larval characters : 



i') The arrangement of crotchets on the feet. The hooks on the 

 abdominal feet are not very reliable characters, as their arrange- 

 ment is quickly subject to modifications whenever the necessity 

 arises for an exposed mode of life by the larva. The presence of 

 the unmodified condition is of more value than its absence, inas- 

 much as no larva of the higher families possesses this typical 

 structure even when returned to a concealed mode of life, whereas 

 some of the lower ones may possess a considerably specialized 

 foot structure. 



2) The tubercles. The special modifications of the tubercles 

 (warts, processes etc.) are not of more than family rank, often of 

 only generic rank. 



There remain, tlien, for primary divisions in the wing veins 

 the presence of vein ic ; in the larvoe the approximation in position 

 of tubercles iv and v. In the higher moths of the first division 

 (Pyralidje, the Saturnians above the Bombycid^e and a few others) 

 there is a tendency, often complete, to the disappearance of vein 

 IC. In the larvai of several genera of the lower families (Tineidas) 

 tubercles iv and v are scarcely approximated or even remote and 

 out of line.* But taken together, the exceptions tend to eliminate 

 themselves. As I am only combining two classifications, some gaps 

 may occur. 



In the second primary division vein ic is absent from the 

 wings and larval tubercles iv and v are remote. In some Dioptidai 

 there is a partial development of vein ic on hind wings, but the 

 larval characters are normal. 



First division. I have divided this into three super-families on 

 larval characters but, as there appears to be no corresponding 

 character in the venation, my Cossina and Anthrocerina must be 

 united. Further, the Pterophoridre cannot be placed positively in 

 the Cossina or Anthrocerina as different genera exhibit the charac- 

 ters of both divisions. 



We have left, then, first, the Cossina (Microlepidoptera) with 

 generalized wing veins and usually a frenulum, the larvi^ with 

 tubercle ii either present or disappearing by fusion with i ; and, 



* This is probably due to the generalized condition of these larvre, which 

 have retained the piimitive arrangement of the seta- found also in stage I of the 

 butterflies. 



