THE AMERICAN NATURALIST [Vol. XLIV 



only 54 children out of a total of 96 that are either 

 brunet or intermediate, or about 56 per cent, of all. With 

 simple segregation and dominance 75 per cent, of dark 

 offspring are to be expected ; the deficiency is, I suspect, 

 partly due to the exclusion from the left side of the 

 table of some families with both parents simplex merely 

 because they failed (in their small families) to produce 

 blonds, although they were potential blond producers. 



(d) Intermediate X Intermediate (Table IV).— The 

 intermediate class serves to include those whose skin 

 has not the clear, transparent, pink quality of the typical 

 blond, on the one hand, nor the rich dark shade of the 

 brunet. It was intended to include a considerable range 

 of color from 10 per cent, to 18 per cent, of black in the 

 color wheel. As already stated, however, collaborators 

 assigned less than a third of the offspring to this class. 



The distribution of skin color in the offspring of two 

 intermediates offers, it must be freely admitted, great 

 difficulties. There are several possibilities. It might 

 be that the "brunet" type of skin color is typical for 

 skin pigment. Accordingly, the intermediate condition 

 may be conceived as having been stopped part way in 

 color development. This stoppage may be due to the 

 fact that the units essential to the later phases of color 

 development are lacking, or to the fact that the stimu- 

 lus to full pigmentation is weak. The first alternative 

 assumes many units for pigmentation; the second, one 

 unit that fluctuates widely. Again the intermediate con- 

 dition might be the consequence merely of its simplex or 

 heterozygous nature. If the latter were the case two 

 intermediates should produce light and dark offspring 

 again in nearly equal proportions as well as intermedi- 

 ates. But if either of the two first-named hypotheses 

 is correct, in accordance with results found by us in 

 hair color, the offspring should not exceed in pigmenta- 

 tion the more pigmented parent ; in the same way that in 

 the offspring of two blond parents the parental color is 

 not exceeded. 



Table IV gives the data precisely as reported. As in 



