RESULTS. 



13 



50 



40 



30 



^ 



20 



10 



tion of the negro skin. If there were only one (duplex) factor involved, 

 we should expect one-fourth of the children in the F2 generation to be 

 white, one-fourth black, and about one-half of all to be of the mulatto 

 grade. But, as a matter of fact, in 32 F2 individuals there are not 8, 

 but only 2 that are white, or i in 16; consequently it is certain that 

 more than one (double) factor is involved in black skin pigmentation. 

 Consideration of all the tables has led 

 me to the following hypothesis, which 

 may be stated now in order that it may 

 be tested by the results of other matings 

 to be considered directly. There are two 

 (double) factors (A and B) for black pig- 

 mentation in the full-blooded negro of the 

 west coast of Africa, and these are sepa- 

 rately inheritable. 



On the foregoing hypothesis we 

 may look for five conditions of skin 

 color, as follows: (i) no factor for 

 black — the Caucasian condition ; (2) no 

 B factor, the A factor simplex — the 

 light-colored ; (3) either no B factor and 

 the A factor duplex or both A and B 

 factors simplex — the medium-colored, 

 or mulatto; (4) one factor duplex and 

 the other simplex — the dark-colored 

 skin; (5) both factors duplex — the black 

 skin. It is evident, moreover, that these 

 five grades do not correspond to sharp 

 percentages of black, and, indeed, it 

 was not to be expected that they would. 

 Every character is subject to fluctua- 

 tions due to variations in conditions 

 during development, effects of sunlight, 

 etc. If, however, these five points are 

 real ones they should show themselves 

 when the grades of the skin color of the 

 entire population are thrown into one 

 frequency polygon. This has been done 

 both for the determinations made at 

 Bermuda and at Jamaica. The polygon 

 of percentages of skin black in Bermuda 

 shows the greater variability and con- 

 sequently lends itself the better to our 



purposes. The polygon of the Jamaica determinations has such an 

 overwhelming proportion of the light-colored individuals as to obscure 



2 



10 20 30 -JO • 60 60 1 70 

 Per cent of N in skin color 



80 



Fig. 3.- 



-PolyRon of frequency of c.ich grade 

 of N in skin color. Bcrmud.^. 



