Revision of the Genus Hrebia 171 
ing to decide many difficult points of this character. We 
have found that in most cases when one knows them well 
enough they confirm the ideas we had arrived at from 
other considerations, and only exceptionally do difficulties 
arise as to the relationship of the species. 
With regard to the question of nomenclature I adhere 
to the views tbat I have often expressed, which are: that 
as 1t Is in many cases impossible to be absolutely certain 
about the species described by old authors, who often, 
indeed I may say usually, were most imperfectly ac- 
quainted with the species they attempted to describe, it 
is far better to use a name which has been generally 
accepted and adopted by modern naturalists and writers, 
than to try to apply too strictly the laws of priority. Since 
I last wrote, the trinomial system, which has been adopted 
by many ornithologists, has been introduced by Mr. 
Rothschild in his revision of the Eastern Papilionide. It 
is no doubt necessary to have some lower grade of defini- 
tion than what I consider specific, and the number of 
pamed varieties which have been described, but which are 
seldom capable of exact definition, has among the Palearctic 
Lepidoptera assumed alarming proportions; but if in such 
‘a case, as for instance, /. alecto, Hiibn., or #. cecilia, 
Hiibn., we adopt the trinomial system, we lose sight of the 
facts which have been so clearly pointed out by Staud- 
inger on p. xxiil. of the introduction to his Catalogue, that 
there are several kinds of varieties, namely, local varieties, 
more or less constant, which he defines by the prefix of 
“var.” ; accidental varieties or aberrations, which he defines 
by the prefix of “ab.”; and seasonal varieties or genera- 
tions which he defines by the prefix of “gen. i.” or 
“gen, 1.” as the case may be. Now if I write #. glacialis 
alecto, Hiibn., I obscure the true facts, because Hiibner 
was not the author of glacialis, and his alecto is in some 
districts of the Alps a local variety, and in others accord- 
ing to Calberla an aberration only. £. manto var. cecilia, 
Hiibn., which in the Alps is a rare aberration, in the 
Pyrenees is a constant variety, and if I could be sure that 
the typical manto did not also occur in the Pyrenees, I 
would treat it as a good species and give it a new name, 
as I have never seen an Alpine specimen which is exactly 
like it. 
Therefore however suitable and justifiable the trinomial 
plan may be in ornithology, I do not think it is applicable 
