2 LEPIDOPTERA HETEROCERA 
which it was founded, is an absolute synonym of Henmithea. Hemitheinae, ex Bruand (1846), is therefore 
certainly its correct appellation. 
The subfamily is evidently on the whole a very-natural one, although we have allowed it to 
include a few doubtful forms (ProtopAyta, Turner, Cacochloris, Prout, etc.), its least specialized (ProtopAyta, 
in any case) being almost as susceptible of being placed among the heterogeneous QEmochrominae. 
In working out the classification we have been incalculably assisted by the work of Dr. A. Jefferis 
Turner, with whom we were in constant correspondence during the progress of his revision of the 
Australian genera, since published in the Proceedings of the Linnean Society of New South Wales, Vol. 35, 
pp- 555-653 (1910). Previous classifications had been based chiefly on the antenna, tibial armature, dorsal 
crests and venation, all of which furnish useful characters, but none, so far as can be ascertained, a 
consistent scheme. The gradual, progressive obsolescence of the frenulum, though of course it had not 
been entirely overlooked, had not been systematically consulted; but it is this which forms the foundation 
of Dr Turner's revision. The Australian fauna, comprising all the most archaic forms and at the same 
time some of the most highly specialized (e. g. Cemochlora, Warren), furnishes excellent material for 
obtaining an insight into the general lines of evolution, and we have made use of the basis with which 
Turner has furnished us. At the same time, its application is not without difficulties. The fact that the 
organ differs in the sexes, and that its atrophy is not always concurrent in both, would probably bring 
its employment into disfavour with extremists who admit no sexual character as even generic. But still 
more, the fact that it results in (roughly) 4orizontal sections from the genealogical tree places it somewhat 
out of harmony with the ideal principle of natural classification (which would take vertical sections, or, 
rather, would seek to cut off separate branches for « tribes ») and often necessitates our placing rather 
widely apart genera which are pretty clearly in an almost direct line of descent, such as the interesting 
group of genera with characteristic palpal and genitalic structure and kindred larval specializations 
typified by Comibaena, Hübner (with frenulum present) and EucAloris, Hübner (ífrenulum absent). 
Dr. Turner, with his usual acumen, has fully recognized these limitations of the system, and has in part 
compensated for the second by supplying a carefully thought-out genealogy of the Australian genera. 
This would be much complicated by the inclusion of the entire world's fauna, with its multiple ramifica- 
tions and interlacings and its many imperfectly known genera (especially African); and we have neces- 
sarily been content with the broadest general outlines, supplemented by occasional comments on obvious 
or pretty apparent relationships under the individual genera. 
Apart from the frenulum, the principal lines of specialization have been the loss of tongue, loss 
of metathoracic und abdominal crests, loss of median spurs of hindtibia, pectination of the Q antenna, 
shortening of the cells, with increased tendency to stalkings in the venation. The palpi are also very 
interesting and important, but specialize in divergent directions, and will probably require an exhaustive 
microscopical study before their taxonomic significance is fully understood ; in the African fauna, in 
particular, apparently near allies frequently show extreme differences in palpal length. None, however, 
of these characters has shown nearly such regular progress as the frenulum ; and unfortunately the few 
which proceed uniformly in the two sexes (cresting, venation, etc.,) seem among the least adapted to a 
stable classification. Even excluding ProtopAyía, which must have escaped developing them, the crests 
begin to disappear in some obviously rather primitive forms, such as Eftfristis, Meyrick, yet are retained 
in some genera which show in other respects rather high specialization (e. g. Lasiochlora, Warren, 
Lophochorista, Warren, etc.), perhaps even in one where the frenulum has disappeared (see Lof/hostola, 
Prout). The venation is extremely inconstant in details, although its general course of evolution is as 
here indicated. SC? of the forewing only arises from the cell in a few very primitive genera, while the 
stalking of SC! is indicative of at least a moderate degree of specialization; with the exception of 
Ornithosbila (which we regard as high up in its group) it occurs in no genus with the Q frenulum 
