THE DISTRIBUTION OF VARIANTS IN NATURE 91 



We now proceed to consider the actual mode of occurrence 

 of variants in nature, in so far as they form recognisable parts 

 or assemblages within natural populations. 



As we have stated (p. 8), all stages can be traced from a 

 variant which occurs sporadically in a population or occurs 

 in a small local enclave to a well-marked local or geographical 

 assemblage. Any attempt to isolate and classify particular 

 types of occurrence . must necessarily be arbitrary ; but it 

 seems to us that the following scheme illustrates the chief 

 stages in the process : 



I. Sporadic individual variation usually involving a 



single character. 

 II. The local combinations formed from a stock of variable 



characters. 

 III. The emergence of qualitatively distinct groups involving 

 large sections of a population. This embraces all 

 the divergences usually alluded to under the terms 

 polymorphism and geographical variation. 



Of these three stages the phenomenon usually known as 

 polymorphism includes both II and III, while geographical 

 variation illustrates III. 



These differences are seen in physiological as well as 

 structural characters, and the former will be discussed at the 

 end of this chapter. 



It will be understood that precise knowledge as to the local 

 distribution of variants (either in single characters or in 

 several) ought to be based on a very large array of specimens 

 collected at all points over the range of the species. Such 

 intensive studies are unfortunately uncommon. Population 

 analyses have been conducted on a large scale upon commercial 

 fishes, though it is at present uncertain to what precise extent 

 the characters studied (size, number of vertebrae and fin-rays) 

 are influenced by the environment. The population analyses 

 of Sumner (Peromyscus) are not sufficiently intensive and are 

 more concerned with the causes of local divergence. By 

 far the most valuable data are those based on the population 

 of land snails (Crampton, etc.), to which allusion is made 

 under II. 



It should finally be noticed that practical experience as 

 well as a more refined study of natural populations has revealed 



