March, 1898. J GROTE : CLASSIFICATION OF LEPIDOPTERA. 19 



There remains, then, but the absorption of IIIi 4- 2 on primaries, and 

 the loss of VIII on secondaries to evolve out of Hemileuca the type of 

 Saturnia ; and this without violence and following the lines of evolu- 

 tion which we have shown to be followed by the lepidopterous wing. 

 Now to form the Hemileucid wing out of the Citheronian or Automerid 

 type we must have recourse to violence, and this violence is apparently 

 not considered but committed by Professor Comstock and Dr. Dyar. 



The neurational type of Aglia and Automeris is practically identical, 

 so that their position is parallel to that of Attacus and Saturnia. We 

 may consider them together. They differ exactly by characters on a 

 line with the evolutionary advancement we have everywhere pointed 

 out. In the first direction a hesitating and half-expressed step has 

 been taken by Aglia. The cross-vein, still uneven, still distinctly rem- 

 iniscent of its true character as a crossvein becomes oblique between IV2 

 and IVi. In all the Automerinse from South America I have yet been 

 able to study, the cross vein is transverse as in Automeris io. The point 

 of issuance of IIIi +2 varies somewhat, but little. In this, the second 

 direction, as we have above seen, Aglia is again more specialized. But 

 otherwise the wings are identical. Neither express any of the distin- 

 guishing features of theSaturnian type. Inasmuch as the first direction, 

 the suppression of the Media, is everywhere less progressed, both Aglia 

 and Automeris are more generalized than the Hemileucid and Saturnian 

 type. In their progression they have lost vein VIII of secondaries, here 

 passing Hemileuca by, while the absorption of the radial veins would have 

 rested at the Hemileucid stage. These are all secondary lines of ad- 

 vancement, unequally entered upon. We conclude that Aglia repre- 

 sents Automeris in the Old World and that it is the more specialized 

 type. Both have sprung from the same near ancestors, the same stem, 

 whether independently, or together, or whether Aglia may be looked 

 upon as the outcome of an Automerid form, we can only surmise. But 

 there they are and they belong together, their sundering, by any sys- 

 tem of classification, from their common stem, is an act of violence and 

 equivalent to a denial of any lessons to be derived from the neuration, 

 at least so long as their common characters cannot be explained away. 

 We are confident that it is impossible and that the classification we pro- 

 pose is natural and in accordance with the facts. 



It does not diminish the difficulty to multiply the families ; if we, 

 out of the six subfamily groups originally proposed by me, make, in- 

 stead of two, the whole six figures as families in our books. Always 

 will Hemileuca, Saturnia and Attacus come together, always will athe- 

 roma, Automeris and Aglia coalesce upon the type of wing. That 



