go THE VARIATION OF ANIMALS IN NATURE 



contrast between Peromyscus polionotus and albifrons and the 

 gradual transition between P. leucopus and noveboracensis de- 

 scribed by Osgood (1909). 



It is also worth while, from the genetical point of view, 

 to summarise briefly at this point some of the data with 

 regard to intergradation in specific characters. 



(a) If two species meet but do not interbreed, then there 

 is no tendency for their character-complexes to break down 

 more frequently in the area where they meet than elsewhere. 



(b) When the intervening area is inhabited by a more or 

 less definite intermediate form, there is considerable break- 

 down in correlation. But the breakdown is of a predictable 

 sort and not altogether at random, some of the more strongly 

 correlated characters remaining in association. 



(c) When there is complete intergradation, correlation 

 between specific or racial characters is completely broken 

 down over an area of varying size. Specimens can be given 

 only a conventional taxonomic name on the basis of the 

 majority of the characters exhibited. Numerous instances of 

 such intergradation are noted in our examples (pp. 102-1 19). 



Grinnell and Swarth (191 3) also recognise these three 

 types of intergradation and see in them, probably correctly, 

 three stages in the fixation of specific type. 



7. Darwin (1884, p. 42) was the first to point out that 

 there is a relationship between the extent of the range of a 

 species and its variation. Most zoologists probably believe 

 that 'widely ranging species vary the most' (Darwin, I.e.). 

 By ' widely ranging ' Darwin clearly meant ' having a wide 

 distribution ' (as species) and not ' having a wide individual 

 range,' a distinction of some importance. Obviously, if we 

 take ' variable ' to involve merely the number of mutations 

 Darwin was at least theoretically correct, because there will 

 be a larger chance of mutation in a large population than in 

 a small one. If he meant that such forms tend to throw more 

 numerous varieties or regional forms, the statement is only 

 true in a very general way. We shall see later on (p. 105) 

 that the amount of regional variation is determined by a 

 variety of factors, among which habits play a very large part, 

 and that there are many cases of widely ranging species (e.g. 

 the Common Heron) which show very little or no regional 

 differentiation. 



