89 



VARIATION IN THE GENUS EEEBIA. 

 By Geoffrey Smith. 

 (Concluded from p. 71.) 



E. tyndarus (measurement of fore wings from apex to inser- 

 tion in thorax ; male and female variable. Proportion of females, 

 25 per cent.) — 



Eange in males . . 15 mm. to 19 mm. 

 Eange in females . . 16 mm. to 19 mm. 



These measm'emeuts are from fifty individuals collected in a 

 confined area in Haute-Savoie. The males fell into a normal 

 scheme, thus : — 



3 per cent, measured under 16 mm. 

 j-y >> >> )> -L I }> 



4o ,, ,, ,, io ,, 



OU ,, ,, ,) iy y) 



o ,, ,, ,, ^U ,, 



This is opposed to the theory, but the following considerations 

 afford a possible explanation. From the specimens I possess I 

 am strongly inclined to believe that the larger individuals of 

 both sexes occur earlier in the year, and graduate down to the 

 smaller as the season advances. By this means the variations 

 are kept separate and constant in proportion, despite the dis- 

 parity in number of the sexes, just as the numerical relations 

 between so many distinct unvarying species would be kept un- 

 changed from generation to generation, unless some outside 

 influence, e.g. natural selection, disturbed those relations. From 

 the variations being thus kept separate, the character of length 

 of fore wing may be considered as a constant one from our 

 point of view. We possibly see here a case of incipient seasonal 

 dimorphism. 



E. melampus (spot-power : male variable, female more or less 

 constant at 7. Proportion of female about 10 per cent.). 



In order to economise space, I will condense the remarks on 

 this species. 



The scheme derived from males of very different areas was 

 normal, and gave a M of 6 J. 



The scheme derived from males collected by myself in a con- 

 fined area was abnormal ; Mwas5. The M of var. sitfZeiica male 

 is about 7. 



E. ligea-euryale (spot-power : male and female variable. Pro- 

 portion of female about 20 per cent.). The questions relating to 

 the specific distinctness of these forms are very complex. In his 

 list of the genus Erebia (Trans. Ent. Soc. 1898), Mr. Elwes 

 separates the two as good species, but names such varieties as 

 E. ligea vars. adyte and livonica as transitional to E. euryale. 



