1060 The American Naturalist. [December, 
THE CLASSIFICATION OF THE LEPIDOPTERA ON 
LARVAL CHARACTERS. 
By Harrison G. Dyar. 
Several articles' have appeared in the AMERICAN NATURAL- 
IsT, presenting different views of the classification of the Lepi- 
doptera. Certain studies on the larve have tended to show 
that there are characters of classificatory importance in this 
immature stage, and it may be interesting to compare the evi- 
dence furnished by them with that deduced from the mature 
structures. 
Prof. V. L. Kellogg, accepting the division of the Lepidop- 
tera into the suborders Jugate and Frenate, finds in the fam- 
ilies of the former certain generalized characters in the mouth 
parts; but the Hepialide exhibit an atrophied condition. In 
the larve these conditions are reversed. The Hepialid larvee 
present distinctly the characters of classificatory importance, 
while the leaf-mining Micropterygide are considerably atro- 
phied. In the view advocated by Dr. A. S. Packard, the 
Hepialide are placed, not in a separate suborder, but low in 
the scale, near the Tortricide. Therefore, these larve will 
serve as something of a test between the two views advanced. 
Dr. Packard has discussed the larve of the Hepialide and 
quotes their characters as supporting his views, saying that 
the hairs are arranged in the same way as in normal Tortricid 
and Tineid larve “ the four dorsal hairs arising from minute 
warts arranged in a low or short trapezoid.” He has also 
given figures of several species (Journ. N. Y. Ent. Soc., iii, 70, 
pls. iii and iv). This article is, however, open to criticism in 
two essential points. In the first place, the differential char- 
acters of the families of Lepidopterous larvæ do not reside in 
the dorsal warts. By this argument, Hepialus could equally 
well be proved to be a Noctuid or a Butterfly. In the second 
place, the figures of Hepialus larve do not show all of the 
1 Am. Nat., March, June and August, 1895. 
