March, 1898.] GROTE : CLASSIFICATION OF LEPIDOPTERA. 
23 
be. Vein VIII on secondaries has been retained by the two “ lowest” 
groups on the respective branches, Hemileuca and Citheronia , exactly 
as appears most natural, in my tree, wheras in Dyar’s Hemileuca goes 
to the top. The association of Hemileuca and Automeris as equivalent 
groups by Dyar seems, from this point of view, impossible. The whole 
wing pattern of the Agliid branch on my tree holds together, with 
Citheronia as its slightly dissenting feature, while the whole wing pat¬ 
tern of my Saturnian branch holds together without any discordant ele¬ 
ment whatever, unless the presence of A III in Hemileuca is one, but this 
does not prevent Dyar placing it with Automeris . So that it is possible, 
from the neuration, to admit of three “families:” Saturniadae, Agli- 
adx, Citheroniadse. Further than this we cannot go, and the matter 
must be left for more light. If Aglia belongs to the Saturnian branch 
and Hemileuca to the Automerid, then Dyar is correct, if not, then I 
am justified. 
The strength of Dyar’s argument and his system in general lies in' 
the indifferent nature of the position of the tubercles. Where such 
ornaments or their details can be proven to be useful to the organism, 
adaptive, they are clearly secondary and their importance fails. I 
cannot judge of the value of the tubercle on the anal plate, but 
must take Dyar’s word for it that it is primary. So we are at a 
deadlock. The pattern of the wing venation, not the position of 
the movable veins, is for me primary. In this case Hemileuca dis¬ 
plays the Saturnian pattern. The presence of vein VIII on second¬ 
aries is subordinate in value to this. Hemileuca y from the pattern 
of neuration, can not, by any reasonable process, have either been 
derived from Automeris , or alongside of it, or represent its ancestor_ 
the role Dyar expects to fill, since it is less specialized. Its capabilities 
are exceeded by one and all of these demands. Automeris , on the other 
hand, may very well have thrown off Aglia, indeed I believe that Aglia 
sprang from Automerid-like forms. I can also clearly see, that Saturnia 
must have sprung from Hemileucid-like forms, j So different are Saturnia 
and Aglia they are with difficulty compared. Citheronia , while at the 
bottom, showing the Castnia-like pattern of Aglia and Automeris , pre¬ 
sents a modification in the movement of vein IVi, analogous to the 
Sphingidae, Pierids and Nemeobius. Attacus and Saturnia show the 
Nymphalid movement of the meridian branches, but add to it the Pierid 
and Lycsenid specialization of the radial branches. Rothschildia iaco- 
b(z<z has the most specialized neuration of any lepidopteron known to 
me. On another line, the common White butterfly competes with it. 
