12 HEREDITY OF SKIN COLOR IN NEGRO-WHITE CROSSES. 



for this study has been small. Such cases as have been collected are 

 given, with full details, in table 6. 



Table 6. — Proportion of black in skin color of offspring of two mulattoes, together with color 

 of parents and grandparents, so far as asctrtained. 



Comparing table 4 with table 6, it appears at once that whereas 

 the total range of skin color in the Fi generation is, for 52 individuals, 

 26 points; in the F2 generation, for only 32 individuals, it is 47 points. 

 Since the range increases, other things being equal, with the logarithm 

 of the number of individuals, we see .that the variability in the Fj 

 generation is more than twice as great as the variability in the Fi 

 generation. Similarly, the average range of variability inside of "a 

 fraternity of Fj offspring is nearly ri the average range of variability 

 inside of a fraternity of Fi olTspring. Also, whereas the lightest mulatto 

 in our records has more than double the percentage of black of a 

 medium-skinned Caucasian, 2 out of 32 of the Pj generation have 10 

 or n per cent black or close to the skin color of the ordinary brunet 

 Caucasian. Compare, for example, the following formula;: 



L. 2, I. Fi generation hybrid N lo, R 30, Y 12, W 48 



C.tucasian N 7, R 30, Y 15, W 48 



In these cases the proportion of white and of red is the same; the 

 difference is in the relative amount of black and yellow. 



Similarly, the darkest of the Fo generation may exceed the color 

 of the mulatto parents; though, owing to the great range of color of 

 pure-bred blacks, they do not closely approach the color of the darkest 

 blacks. The follo'W'ing case from Bermuda is instructive: 



F, father N 32, R 41, Y 14, W 13 



F, mother N 39, R 37. Y 10, W 14 



Child N56, R31, Y 6, W7 



Cliild N 54, R28, Y 6, W12 



Here the skin color of the darker child is almost one-half darker than 

 the darker parent, is darker than any mulatto in my records, and is 

 darker than many of the "pure-bred" Africans, 



V. Hypothesis. 

 The increased variability of the Fj as compared with the Fi genera 

 ■* tion indicates the probability of segregation. It now remains to decide, 

 if possible, how many factors for black are involved in the pigmenta- 



