152 THE ENTOMOLOGIST’S RECORD. 
degenerative cause, for in other Lycaenidi it is found elevated to the 
rank of a specific character (pheretes, etc.) ; it would rather seem an 
indication of a state of organic equilibrium adapted to live in Arctic 
‘or very damp climates. 
Race sarmatis, Gr., is from the mountains of Southern Russia; it 
has no orange lunules above and the underside is light gray. 
Staudinger and Seitz make what would be an infracandida, Vrty., of 
it, but Tutt has examined the “ types” in the Brit. Mus. and cleared 
up what Grum meant. 
Anyone can easily see that pallidefulva and montana stand outside 
the gradation which leads from medon to calida; in this continuous 
series there exists a tolerably constant proportion, in the medium of 
individual variation, between the grades of the different ¢haracters, 
both when they progress in a parallel fashion (extent of lunules, 
brightness of fulvous, etc.) and when there is an inverse proportion 
(characters just mentioned and intensity of gray below), although 
individual variation is broad. Form pallidefulva may be found in 
certain localities mixed with individuals belonging to the main series, 
but in others it constitutes a distinct race, not very variable; it stands 
out sharply on account of a very accentuated difference in the 
proportion between the grades of the different characters: by the extent 
of the lunules it is equivalent to the higher grades calida and 
infracandida ; by the entire, or nearly entire, absence of grey mixed 
with the fulvous it is also equivalent to the latter or to infralbens, as 
also to the individuals of calida with a very bright, but very pure, 
fulvous underside ; on the contrary by the tinge of the lunules on both 
surfaces and by the tone of the fulvous of underside it is equivalent to 
the lower grades gallica and aestiva and to this it adds an exceptionally 
light and reddish upperside on account of the strong reduction of the 
black. In montana=nevadensis these characters are still more 
accentuated, except the extent of the lunules, which is extremely small 
on the upperside and also on the underside, so that a form transitional 
to A. ramburi, Vrty.=idas, Ramb., is the result, although this is 
considered a species distinct from medon. 
This example of medon seems to show that individual varia- 
tion may include differences in the proportions between the 
erades of the different characters, but that in such cases ‘‘ specific 
elasticity’ gradually comes into play and tends to produce 
stable phylogenetic races, which by a further displacement of the 
centre of balance may originate subspecies and perhaps even species. 
Although this last conclusion would have such an enormous import- 
ance in the vewata questio of the origin of species that one scarcely dares 
admit its possibility, I must say that the more my acquaintance 
with variation in the Lepidoptera increases the more examples I find 
of extraordinary resemblance between species. I find that related species, 
perfectly distinct at one end of their series of geographical variations 
or in one of their annual generations, often are so similar in the 
opposite variation or in another generation as to be quite difficult to 
separate from each other. As, for the present, we are only able to 
establish specific distinction on the ground of morphological characters, 
such resemblances at one end of divergent series may, I think, be worthy 
of careful consideration. 
(To be continued.) 
