VARIATION IN THE GENUS EREBIA. 279 



Table!., and in column 6, Table II., is that the former represent 

 numbers observed, and the latter represent averages of numbers 

 calculated. Now, halving the values for each sex, we derive the 

 following numbers from Table I. : — 



Table II. 



Spots calculated 

 Sums from Spots $ $ for next genera- 



beginning. Cases observed. observed. heritage, heritage, tion of males, fe- 

 males ignored.* 



10 10 per cent + 4i 4^ 



20 10 „ 1 * + 41 5 



25 5 „ 2 1 + 41 5A 



30 5 ,, 3 1* + 41 6 



40 10 „ 4 2 + 4* 61 



55 15 „ 5 21 + 41 7 



65 10 ,, 6 3 + 41 7i 



75 10 „ 7 31 + 41 



80 5 „ 8 4 + 41 



100 20 „ 9 41 + 41 



The mean must now be calculated from columns 1 and 6 in 

 Table 11. The curve is shown in fig. I., curve B. The mean is 

 about 7. 



If we compare the two means 4^ and 7, we see that on the 

 condition of every individual pairing an approach is made to the 

 typical number of the species (namely, 9), by raising the mean 

 every generation. It need hardly be remarked that a generation 

 with a mean " spot-power "of 7 is likely to produce more indi- 

 viduals with the typical 9 spots than a generation with a mean 

 "spot-power" of 4^, and so on in ascending order. In this 

 way, then, everything else being equal, the stability of the 

 species would be regained to a very great degree after a few 

 generations. 



Now it is obvious that, if not only every individual in the 

 tables does not pair, but, on the contrary, only a very small 

 minority can ever possibly hope to, the whole equilibrium of the 

 system is thrown out of order. A few individuals with various 

 numbers of spots will succeed by chance in pairing with the few 

 females that exist ; these will then establish their "spot-powers" 

 for the next generation, which will be overthrown in the next for 

 one also established by chance, and not in any regularly ascend- 

 ing scale. 



To show that this state of things would not lead to an 

 equilibrium, we must follow the subject rather more closely. 

 Let us suppose that the mean power 7 has been acquired by a 

 generation in the way described, and that then the majority of 

 the females has been destroyed before the possibility of pairing. 



* Reasons for ignoring females will be given later. At present, regu- 

 larity of increase is to be noted. 



z2 



