﻿24 Journal New York Ent. Soc. [vol. hi. 



Clisiocampa neustria. 



Mature larva. Several coarse black hairs arise from an area representing i, 

 and a single hair frpm ii; other warts obsolete. Secondary hairs fine and soft, 

 quite abundant sub-ventrally. Body scarcely flattened, a slight noctuiform dorsal 

 enlargement on joint 12. 



Lasiocampa quercus. 



Secondary hairs abundant, thick, foiming broad bands of dense hair absent 

 only at the edges of the segmental incisures but becoming short and thin laterally 

 and sub-ventrally. Primary hairs absent, the position of i and ii indicated by a 

 disturbance in the uniformity of the dorsal hairy coat. Just below the spiracle and 

 behind it an area can be distinguished which represents iv, and before and below 

 this, another smaller one, rather a colored spot than a structural area, represents v. 

 The area corresponding to vi, is well clothed with hairs and projects a little. 



Gastropacha quercifolia. 



Greatly flattened with well developed sub-ventral " lappets " and an unpaired 

 dorsal prominence on joint 12. Hair rather abundant, but not long, all secondary. 

 Patches of vivid white, short, dense hairs occur between the legs sub-ventrally. 

 Seen at a distance a series of blackish patches seem to represent tubercles i and ii, 

 but under the lens nothing appears but patches of dark pigment; a black pigment 

 spot behind the spiracle. vSub-ventral area in the region of tubercle vi produced in 

 two rounded confluent and hairy processes, forming the lappet. 



CONCLUSION. 



As regards the characters of wing structure, I believe the 

 following are be eliminated as unreliable. 



i) The frenulum. This structure seems very flexible and 

 readily subject to adaption. It is not reliable even for family 

 definition in many cases. I have before me the European i5'/;^/;v;;///i' 

 versicolor and the Australian Chelepteryx collesi. In all characters 

 they approach each other very closely, even the peculiar pattern 

 of maculation is identical, and they must belong to the same family 

 (Endromid?e, Plotz 1885); yet the former is without frenulum with 

 an expanded numeral lobe while the more primitive Australian 

 form has a well developed frenulum. Other examples could easily 

 be cited. 



2) Tlie brancJies of cubital vein. The apparent number of 

 branches of the cubitus (three or four), which may be otherwise 

 expressed as the direction of migration of 'vein 5 ', is insisted upon 

 by Mr. G. F. Hampson as a character of super-family value. In 

 comparison with the larval classification it is seen to be of family 

 value only. 



