224 THE VARIATION OF ANIMALS IN NATURE 



theory of inheritance has been supposed to have an enormous 

 advantage over the blending theory held by Darwin. For 

 with blending, a new variant, unless isolated, is always liable 

 to be swamped by the excess of normal individuals in the popu- 

 lation. Hagedoorn and Hagedoorn (1921) have emphasised 

 that, even with particulate inheritance, the establishment of a 

 variant from a few individuals almost equally demands the aid 

 of isolation. In almost all animals the number of individuals 

 which breed in any one year is only a small fraction of those 

 which existed at the end of the previous breeding season. 

 This seasonal fluctuation in numbers means that on the 

 average only very common types can survive and the chance 

 of any particular rare variant surviving is very small. The total 

 variance of the population is being repeatedly reduced, and 

 the additional chance of survival conferred on a variant slightly 

 better adapted to some one feature in the environment is very 

 small — much smaller than would be the case in more stable 

 conditions. With isolation, though the same factors would 

 be at work, a new variant might form a far more significant 

 proportion of the population. 



It may be argued that though the chance of survival is 

 small, yet, if the mutation occurs often enough, it may still 

 become established ; and that though the mutation-rate be 

 low, yet, in a species including thousands of millions of indi- 

 viduals, each type will occur relatively frequently in each 

 generation. It may be held that, even when the population 

 is reduced to a minimum, the numbers may still be very large 

 compared with those in which a mutant might be expected 

 to occur. In other words, as long as a mutant has a positive 

 survival value and the species is not a rare one, the actual 

 value of the mutation-rate is relatively unimportant, at least 

 within wide limits. 



In a species with a wide range, extending over a con- 

 siderable variety of environments, in each of which conditions 

 are subject to fluctuations of daily, yearly or of longer periods, 

 it is somewhat difficult to assign a definite survival value to 

 a particular mutant. The genetic make-up of the species is 

 itself unlikely to be homogeneous over large areas. The 

 idea of an average survival value is necessarily an unreal and 

 artificial simplification. What is useful in one place or in 

 one year will be harmful or neutral in another. Survival 



