662 



THE AMERICAN NATURALIST [ Vol. XLIV 



third class those expected from DR X DR matings. 

 Shall it therefore be concluded that all brunets as well 

 as intermediates are simplex in skin pigment? This 

 does not follow ; but it does seem to be a fact that duplex 

 "brunet" and " intermediate " are not common; 1 indeed 

 so uncommon as not to alter materially the proportions 

 that would be given on the hypothesis that they are al- 

 ways simplex. 



The principle of the non-transgressibility of the upper 

 limit to which we have called attention elsewhere 2 seems 

 to hold for skin pigment also although the result is less 

 clear-cut, probably because the terms were less accu- 

 rately assigned. To see how closely the law holds Table 

 VIII has been constructed. The three cases: (a) darker 

 parent blond; (b) darker parent intermediate; (c) 

 darker x>arent brunet — are chosen and the distribution of 

 offspring in each case indicated at the right. 



TABLE VIII 



Showing the Distribution op the Different Classes of Skin Color in 



Table VIII shows that exceptions to the rule of de- 

 limitation are, considering the vagueness of terms, 3 rela- 

 tively rare, only about 8 per cent., and, consequently, the 

 rule seems verified. 



The significance of the intermediate grades of skin 

 color is a question of prime importance. There is some 

 evidence, for example in Table IV, families Har. A, Keh. 



