Nos. 618-619] INHERITANCE IN PEROMYSCUS 



293 



races, indicated by the consistent reduction of all these 

 values, has not affected them to an equal degree. In re- 

 spect to all four of the characters (body length not here 

 considered) the local (La Jolla) race has been least modi- 

 fied, while in respect to three of them (tail length, tail- 

 stripe and ear) the Eureka race has been most modified. 

 Thus there has been a mean reduction of 11 per cent, in 

 the tail length of rubidus, 1\ per cent, in that of sonorien- 

 sis, and only 5 per cent, in that of gambeli. There is, 

 therefore, a convergence between the P]ureka and the La 

 Jolla races, and if I had only these two under comparison, 

 I might have been disposed to conclude that local condi- 

 tions had brought about a modification of rubidus in the 

 direction of gambeli. But the case of sonori<'n*is-, which 

 actually diverges farther from the local race in the C 2 

 than in the parent generation, shows that this explanation 

 is not the correct one. These differences in the degree 

 of modification are probably indices of the susceptibility 

 of these three races to the malign influences of captivity, 

 which have already been discussed. In harmony with 

 this view is the fact that the Eureka mice are likewise far 

 less fertile, under local conditions, than either of the 

 other races. 



V. Heredity of Individual Differences within Each 

 Race 



It has been shown that a wide range of individual varia- 

 bility occurs within each race in respect to just those 

 characters by which one of these groups is distinguished 

 from another. These major differences which distin- 

 guish one race from another have been shown to be 

 hereditary. Is this likewise true of those minor differ- 

 ences which distinguish one individual from another of 

 the same race ? 



This question can be answered by the well-known 

 method of computing coefficients of parental-filial corre- 

 lation—the "coefficients of heredity" of Pearson. I re- 

 alize that the validity of this measure of the force of 

 heredity has been called in question, 17 on the ground that 



