74. THE THEORIES OF EVOLUTION 



the 2 variations, after procreating 2Xn similar indi- 

 viduals, which will remain in the varied class, will 

 only produce 2X2 dissimilar individuals. From 

 these 4 variations the majority will remain different 

 from the initial type and will increase the previous 

 number of variations; only a small minority will re- 

 vert to the original type and increase the number of 

 constant types. Even if they should all revert to the 

 original type, the initial ratio of the variations to con- 

 stant types, 2/n, would increase, for we would then 

 have: 2 „ 2+g * w . We know that when we add the 



W + 2X2 



same figure to the numerator and to the denomi- 

 nator of an irreducible fraction, the fraction in- 

 creases; a fortiori it increases when we add a larger 

 figure to the numerator than to the denominator, as is 

 j)resently the case, since n is greater than 2. 



This is an unavoidable corollary of the formula 

 proposed by Delboeuf, but it is only a mathematical 

 corollary. 



The reality is quite different. It is not true that 

 if, in the first generation, there are 2 variations and n 

 constant types every one of them must in the second 

 generation, procreate n similar types. Individual 

 variations are not by any means transmitted so com- 

 pletely to such a large progeny. If they were, new 

 forms would appear much more frequently than they 

 really do and all the petty anomalies such as extra 



