30 MEMOIRS OF THE NATIONAL xVCADEMY OF SCIEXCES. 



The remarkably woolly ami penciled larva? of Apatelodes are congruous with the very distinct 

 imagines of the subfamilj- Apatelodimv, which are so well defined by their structural characters. 



The hairy and brightly banded larvie of the Pygccrincc, so unlike those of other Notodontians, 

 are i)aralleled by the general ajipearance and structure of the moths, so much so that the group was 

 regarded as a distinct family (Pyga-rida^) by Duponchel. The larvie of the European Pygnerinte 

 are hairy and gaily striped, and related in much the same way to our larval Datana as the imago 

 is to our iraaginal Datana. 



The larvae of the subfamily Ichtlnjurinxv, rei^reseuted by only a single genus, need not be 

 confounded with those of any other division of the family, though there is a great deal of plasticity 

 within the linuts of the group. The most generalized species is the larva of I. apicaliH (ran) and 

 its allies hrucci (m ultnvma), iiiuce it has no large specialized tubercles like those of inch(sa and 

 albosigma, and the latter species differs, both as regards larva and imago, from /. iucliiint. The 

 incongruence iu this group is not greatly emphasized. 



But in the two next subfamilies there is a striking lack of congruity between the larva and 

 moth, both in the genera and species. 



Among the Xotodoutinffi we have Uyparpax, whose imago is so different, in the shape of the 

 wings and in the color of the body and wings, compared with any other genus of the group or 

 even of the family; yet the larva is very nearly allied to those of XyHiwilcs and of iSchi;:itra. 



A remarkable case of incongruence is the larva of ^chizura concinna. This well-known 

 caterpillar, with its formidable armature of long hobnail like spines and its gay head aud swollen 

 coral-red dorsal Inxmp, would seem to be the type of a distinct genus, and yet from a study of its 

 adult character it is not separable from the other species of Schizura, and we have dropped the 

 genus (Edemasia we originally proposed for it from the lack of stable differential characters. 

 The freshly hatched larva', however, is undistinguishable from that of other .Schizura- jet known, 

 and perhai)s we have done violence to the principles of classification iu not allowing it to remaiu 

 in the genus we originally proposed for it. At all events, it with other Schizurte evidently had a 

 common parentage, and it has diverged since it first molt farther away from the stem than others 

 of its cospecies and may be regarded as an incipient genus. It is also plain that the causes which 

 have acted upon this organism have from the first been of a quite different nature from those which 

 have been efficient in causing fixed variations in other directions, resulting in the fixation of the 

 other species of the genus. As the change takes place after the first molt, this may have been 

 produced in the Tertiarj' period. Its larval stages are discussed at sojue length under the head of 

 the species in the systematic portion of this work. 



On the other hand, iu the genus Seirodonta we have a remarkable case of congruence in its 

 lai'va as compared with that of Heterocampa manteo. It is almost impossible until after repeated 

 aud careful comparisons to distingush the caterpillars of Seirodonta billnc<(ta and H. manteo, though 

 the imagines differ somewhat, perhaps generically. At times I have' united Seirodanta with 

 Heterocampa, but for the present conclude to keep them apart, as others have done, but really the 

 genus is not so "good" a one as CEdemasia.' 



In the genus Heterocampa, as the name imiilies, there is a remarkable degree of diversity 

 between the caterpillars of the different species, and our knowledge of them, especially of their 

 early stages, has greatly extended since the days of Doubleday. 



If we take account of the fully grown caterpillars, it seems quite evident that there .are several, 

 perhaps three, "larval" genera in the group. In H. manteo, (juttlvitta, hiundata, ohliijna, and astarte, 

 the body in the fully grown larva is smooth aud unarmed, but in j>«/rfrea, which has a pair of 

 small tubercles on the jirothoracic segment, we have a notable persistance of early larval features. 

 Unfortuuately we are not yet familiar with the early stages of this caterpillar. Possibly this 

 species is the stem form of the group. 



Iu H. unicolor we have a transfer of the differential generic characters from the prothoracic 

 region to the anal legs. Though the high prothoracic tubercle appears iu the first stages and 

 perhaps, as iu Macrurocampa, iu all except the last stage, when' the larva is on a level with the fully 



' I have some sketches made by Mr. Bridgham of a larva in its first three stages which is Schizjira-like, and as 

 it feeds on the elm it is probably Seirodonta. Should it prove to be sucli, tliis genus is a Schizura in the early stages 

 and a Heterocampa in the last. 



