8 LEPIDOPTERA HETEROCERA 
in quite a different way, being furnished with an elaborate arrangement of special spine-bearing (or 
hook-bearing) processes whereby they clothe themselves with debris of the food. plant. Pupa more or 
less rugose, greenish or light-coloured, not subterranean nor enclosed in a strong cocoon, but merely 
spun by a few threads among leaves. See, however, ProfopAyta and Sterictobsis, to which will probably 
have to be added other archaic forms. 
Geographical distribution of species. 
Palcarctic and Nearctic (excepting the far north); 
Indo-Australian (strongly represented); /Ethiopian; Neotropical. Wanting in New Zealand, Hawaii and 
and probably numerous Pacific Islands and in the extreme south of South America; almost wanting in 
Chili. Distribution doubtless in part governed by the general attachment of the species to trees and 
advanced types of vegetation; perhaps on account of the absence of circumpolar representation, no 
forms common to the Palearctic and Nearctic Regions (a sharp contrast to the Larentiinae). 
Subdivision of the Subfamily, — As shown in our introductory remarks, the scheme of 
classification does not lend itself readily to the differentiation of clearly-defined Tribes. It would not, 
indeed, be difficult to isolate a few of the groups as such, notably the genera Dysbhania and. Cusuma 
(Dysphantinae, Warren) or the genera Uliocnemis, Comibaena, Euchloris, etc. (vide supra), but it seems 
inexpedient to depart from the general unity of plan for the sake of these few, inasmuch as the sequence 
of the vast residue would not in any way be assisted thereby. We merely indicate, for convenience of 
analysis, six groups, the last two of which are marked off by the specializations of the frenulum-develop- 
ment, while the first four are subdivisions of the group with frenulum complete, or nearly so, and are 
founded on (a) the point of origin of SC? of the forewing and () the nature of the response to environ- 
mental conditions and needs : i. e., the protective Aabitus. The last-named, which is mainly responsible 
for Warren's threefold division, DyspAaniinae, Pseudoterpninae, Geometrinae, may, or may not, prove natural. 
Dr. Turner thinks not, and believes Dysf/ania and one of the leaf-green genera (Gznochlora) to be quite 
closely related to the Pseudoterpninae; but as the great majority have certainly taken independent paths 
we consider that a convenient, even though possibly arbitrary, working arrangement may be obtained 
by recognizing them. The New World genera, which nearly all fall into our fourth and fifth groups, 
have certainly followed the same general course of evolution as these of the Old World, but the Q seems 
to have been on the whole more conservative of the frenulum, thus offering some examples of genera 
(Racheospila, Synchlora, etc.,) which are awkwardly on the border-line between the two groups named. 
For this reason, as well as on grounds of practical convenience, we provide a separate key to the New 
World genera, some of which, although differing sufficiently from their Old World relatives to render 
their fusion undesirable, are yet much more sharply-defined when compared with the other genera of 
their own fauna. The only two genera which are common to both hemispheres, of which one presents 
some curious problems of geographical distribution (viz. Eucrostes), are of course inserted in both keys. 
Unfortunately a few genera which seem likely to be valid are only known in the Q sex, or are 
otherwise quite imperfectly known. Where it seems reasonably safe to do so, these are introduced into 
the keys as accurately as circumstances will allow, though in one or two cases only by some super- 
ficial distinction such as that of coloration, which would not necessarily hold in the event of the discovery 
of species with similar coloration but differing structure. No key, however, can be made to cover 
possible future contingencies. The following are the only genera known to us which we have felt com- 
pelled to omit from the keys 1), and which must therefore be consulted separately, being here relegated 
to footnotes : CAloroparda, Lathochlora, Mesurodes, Rhomtochlora. 
1) The entirely unknown genera ZesfograéZza, Hübner, and .XenocAt/aena, Lower, are not placed in tlie body of our work at all, but merely 
referred to at the end. 
