Review of the Genus Hrebia. 211 
which there is no approach in any other species. The 
other parts I have not studied, so I can only say that, in 
some species at least, they present marked characters. 
Since the sickle is so constant in form in Hrebia and in 
neighbouring genera presents many considerable varia- 
tions, one might say from an Erebian stand-point that the 
sickle presented generic characters, whilst those of the 
clasp were specific ; for in the allied genera we find forms 
of clasps very like in general outline to those of some 
Erebias. Were we to adopt clasp forms as generic char- 
acters, there would be extreme confusion; thus Hrebia 
ame, Enodia hyperanthus, Chionobas brucei and C. aello 
would be in one genus, Hipparchia semele, Hrebia glacialis 
and Chionobas chryxus in another, and so on. It is, per- 
haps, not quite correct to say this without noting that, 
though there is this great similarity of form, there is a 
recognisable something distinguishing the clasps of the 
Erebias, chiefly perhaps, that throughout the genus, they 
exhibit a vigour and strength about the spines or styles 
which is rare in the neighbouring genera. 
This circumstance emphasises the necessity of always 
interpreting the evidence of the appendages with close 
regard to other characters, and though most valuable for 
distinguishing separate species, otherwise much alike, it 
must only be used to unite those as to which such a 
presumption can be otherwise supported. 
Nevertheless it is by no means futile to attempt some 
sort of classification of the forms of the appendages as an. 
indication of the relationship of the species. Indeed, I 
think, in most cases the appendages give a more certain 
indication of alliances than is derivable from wing forms 
and patterns; because, in H7ebia the variety in the ap- 
pendages is of a much less erratic character than it is 
said to be in many other genera, where close alliance 
otherwise is often associated with the greatest diversity 
in the appendages ; whilst it also happens that Hrebia is 
especially a genus in which colour and marking are very 
misleading. The case of H. melas, in which varieties of 
two very distinct species (H. nerine and #. glacialis) were 
associated as one with a third species, LZ. lefebvrei, itself 
possibly really a variety of another species (ZL. pronoe), is 
by no means a solitary instance of confusion that an 
examination of the clasps corrects. 
Some such form of clasp as that of #. manto or 
