VARIATION IN THE GENUS EREBIA. 277 



does not mix at all with the type-species, as the latter is entirely 

 absent. The males are extremely variable in the character of 

 their facies; the females, on the other hand, are practically con- 

 stant, and in the proportion of about five to one hundred males. 

 (This question of numerical disproportion of sexes will be treated 

 of hereafter.) Although I have selected this variety as a concrete 

 example on which to fix theoretical considerations, the general 

 conclusions will apply to other species and varieties of this and 

 other genera where the facts are similar. 



The males of E. epiphroii var. cassiope are widely variable in 

 their wing-facies, but I want to fix on some variable character 

 that can be treated numerically. Obviously the black spots on 

 the rusty-brown patches will serve ; they are extremely variable 

 in number in the males, and, I believe, constant in the females. 

 The constant number for the females will be taken as nine (it 

 is really about seven), counting those on fore and hind wing. 

 In Table I. is given a list showing an imaginary percentage of 

 distribution of the spots among the males. An imaginary per- 

 centage is chosen for the sake of simplifying the numbers ; it 

 does not interfere with general conclusions. 



Table I. 



Sums from beginning. Cases observed. No. of spots. 



10 10 per cent, had 



20 10 „ „ 1 



25 5 „ „ 2 



30 5 ,, ,, 3 



40 10 „ „ 4 



55 15 ,, ,, 5 



65 10 „ „ 6 



75 10 „ „ 7 



80 5 „ „ 8 



100 20 „ „ 9 



The contention is that if there were equal numbers of males 

 and females — i.e. if all the individual males of Table I., with 

 their various numbers of spots, could pair with a corresponding 

 number of females with their constant number of spots (nine) — 

 then the whole species would tend to return to the original stable 

 condition of nine spots for males and females alike ; but since 

 only a small minority of the males can pair, the species is kept 

 in a state of fluctuation. 



It is necessary to make one preliminary hypothesis— that 

 male and female in inheritance each transmit half their charac- 

 teristics, and that this applies to the number of black spots. 

 This ignores heritages from remote ancestors, but I have chosen 

 J as preserving the proportion between the two sexes, and not 

 complicating the figures. Now, it might be argued from the 



ENTOM. — OCTOBER, 1901. ^ 



