EYE COLOR AND HAIR COLOR IN MAN. 



59 



equal proportions in both cases. In fact many of the results ob- 

 tained by the Davenports and ourselves may be interpreted quite 

 as readily as blended inheritance as anything else. 



It is a rule laid down by the Davenports that " in the midst of 

 varying degrees of melanic pigmentation the intensity of melanic 

 pigmentation never exceeds that of the more intense parent." 

 This is a general rule borne out also by our data ; but it is by no 

 means universal. While the data compiled by the Davenports 

 show few or no cases in which this rule is plainly violated, I find 

 upon examining their data on the relation of parents and grand- 

 parents, a considerable number of exceptions. Apparently no 

 account of these relations was taken, the data concerning the 

 grandparents being used only on account of the light it might 

 throw on the probable constitution of the gametes. In one 

 respect these data are more valuable than what was used, which 



Table IV. 



Showing Inheritance of the Correlations of Eye Color and Hair Color. 



was derived from the members of the third generation, as the 

 effects of immaturity are mainly eliminated. In their table 7a, 

 the rule is violated in 5 out of 13 families, black hair in the 

 father of the Sim family coming from parents both of whom had 

 light brown hair. In the other tables given, less frequent devia- 

 tions are found but they are sufficient to justify a doubt that the 

 non-transgressibility of the upper limit represents anything other 

 than a more or less predominant tendency. In Table III. it may 



