THE ORIGIN OF VARIATION 45 



taken from the desert to a new environment remained un- 

 changed for two to ten generations. Sumner, in fact (1932), 

 abandons the theory of the direct environmental origin of 

 ' desert ' pigmentation. On the other hand, the proof that 

 the racial characters are now germinally fixed does not show 

 that they were always so. Still, for the present at least, the 

 Peromyscus experiments cannot be held to favour Rensch's 

 views. We shall mention later other instances of characters 

 which are racially diagnostic in some species and known to be 

 due to external causes in others (cf. Robson, 1928, p. 166). 



The alternative explanation to Rensch's hypothesis would be 

 that all the races whose differences are correlated with trends 

 are merely ' somatic ' forms or produced by selection. Rensch's 

 theory is, we hold, no more than suggestive, and in the light of 

 Sumner's conclusions as to the intensive study of gradients, 

 perhaps less impressive than is at first sight apparent. It is, 

 however, more fully examined later on (p. 46). In the same 

 category is Ekman's theory (191 3) of the origin of the lacustrine 

 crustacean Limnocalanus macrurus, which, it is claimed, has 

 arisen in many places from the brackish-water L. gnmaldii 

 (cf. Gurney, 1923) owing to the progressive freshening of the 

 lakes which it has occupied since the Glacial Period. This is 

 a case for which we would require experimental evidence, 

 especially as the various lacustrine forms are not all similar, 

 though, as Gurney admits in his critical review {I.e. p. 428), 

 the tendency has been in the same direction. 



Rensch's data (with some supplementary evidence) may 

 now be considered in detail. Some of his most important 

 conclusions in the present connection are summarised in the 

 following three rules : 



(1) Bergmanrfs Law. — In nearly related warm-blooded 

 animals the larger live in the north, the smaller in the south. 

 This is also true to some extent of invertebrates, provided they 

 are compared within their optimum range, outside which 

 dwarfing may appear. 



(2) Allen's Law. — The feet, ears, and tails of mammals 

 tend to be shorter in colder climates, when closely allied forms 

 are compared. 



(3) Gloger's Law. — Southern races tend to be black, brown, 

 grey and especially rust-red ; northern forms are paler and 

 greyer. Humidity here has an important modifying influence. 



