458 Messrs. Elwes and Edwards on 
wings; but they are equalled in this respect by some 
species of Satyrus, S. hermione, for example. 
The following types of clasp-form are found in this 
genus; in norna and jutta there is an unusual amount 
of individual variation, but the differences are merely 
those of degree, and are not, in the most extreme cases, 
sufficient to obscure the relationship of the species :— 
1. Clasp oblong, more or less pointed at the apex, its upper edge 
bearing a single tooth or projection somewhere between the 
middle and the base. 
ex. chryxus, bore, taygete (figs. 3, 10). 
. Clasp rather broadly triangular, and wanting the projection on 
the upper edge . - : . ex. wrda, whleri (fig. 1). 
. Clasp oblong, gradually narrowed; apex bluntly rounded or 
subtruncate; serrulation of the upper edge inconspicuous or 
absent . . : : . ex. hora, walkyria (fig. 7). 
4. Clasp not curving inwardly at the apex, which is bluntly rounded, 
and, together with the upper edge nearly or quite to the 
middle, bears unequal and comparatively large teeth. 
semidea, jutta, fulla (figs. 5, 9). 
5. Clasp rather suddenly narrowed about the middle, strongly 
curved inwards at the apex; the teeth decumbent and 
almost confined to the upper apical angle, which is some- 
what produced. 
ex. subhyalina, brucei, norna (figs. 6, 18). 
bo 
[Sy) 
(Eineis appears to be the most cold-enduring of all known 
genera of butterflies. With the exception of the aberrant 
nevadensis group there is not one species inhabiting a 
region in which the winters are not very long and severe, 
and usually of an Arctic character. About half the 
species are confined to high mountain regions, where 
they frequent stony and rocky ground above timber-line; 
another section, like whlert and alberta, inhabit grassy 
steppes and prairies; a third, like bore and subhyalina, 
occur on the coasts of Arctic seas; and a few, such as 
chryzus and nevadensis, are inhabitants of mountain 
forests. ‘Their larve, so far as known, are all grass- 
feeders, but most of them are unknown. Their geo- 
graphical distribution is, if we except the almost un- 
known Ci. antarcticus from Patagonia, strictly confined 
to the Palearctic region, in which I include, so far 
as Rhopalocera are concerned, the Nearctic region of 
Sclater. 
As will be seen from the references below, the litera- 
ture of the group is somewhat extensive ; but it is for the 
