Wright : Color Inheritance in Mammals 



233 



brown by very light red produced the 

 extremes of hair color — jet black and 

 albino. Holmes and Loomis noted 

 several cases in their records similar 

 to those above, and many such 

 cases can be found in the data of 

 Pearson, Nettleship. and Usher. From 

 all of these sources we get abun- 

 dant evidence of segregation of some 

 sort. It is clear, however, that if there 

 is one main factor by which red and 

 light brown differ from black, it must 

 be imperfectly dominant, and that there 

 must be other factors which raise or 

 lower the pigmentation of the hetero- 

 zygotes from one extreme to the other. 



SKIN COLOR 



With skin color the difficulties in 

 study are naturally greater than with 

 either eyes or hair color. As in the 

 other cases. Davenport^" concludes that 

 the darker shades tend to be dominant 

 over the lighter ones. Where wide 

 crosses are made, as in the union of 

 white with negro, it is well known that 

 there is not unit Mendelian inheritance. 

 The mulattoes are intermediate in color 

 and their children are still mulattoes. 

 Davenport has found, however, that 

 there is much variability in the second 

 generation, and that the results can be 

 explained on the basis of two pairs of 

 Mendelian factors, both imperfectly 

 dominant. Here, again, it is impossible 

 to speak of particular Mendelian fac- 

 tors as demonstrated. We can merely 

 say that there is evidence for Mendelian 

 segregation of some sort, and that the 

 number of factors which need be hy- 

 pothecated is not very great. 



CORRELATED VARIATION 



So far we have dealt with hair, skin, 

 and eye color as wholly separate char- 

 acters. In each case, we have been 

 forced to conclude that the evidence 

 has not demonstrated satisfactorily the 

 existence of particular unit factors wnth 

 clearly defined effects, but merely the 

 presence of Mendelian segregation of 

 a complex kind, with dominance tend- 

 ing toward the darker types, but prob- 



'^Davenport. C. B., 1913; Carn. Inst. Wash. 

 'Tearson, K., 1904. Biom., 3:459. 



ably imperfect as a rule. This is 

 practically the conclusion reached by 

 Holmes and Loomis in their study of 

 hair and eye color, and, indeed, has been 

 recognized, more or less, by all workers 

 in this difficult field. But even if we 

 were able to grant that the main effects 

 are due to single pairs of unit factors 

 in each case (considering only dark 

 granular pigment), w^e would find great 

 difficulty in accounting for the relations 

 of these characters to each other. The 

 three factors for dark hair, dark eyes, 

 and dark skin, respectively, must either 

 be identical or independent of each 

 other. If identical, correlation in hair, 

 skin, and eye color should be nearly 

 perfect. If independent there should 

 be no correlation, unless the population 

 is one in \vhich there is assortative mat- 

 ing. With partial linkage of factors, 

 the correlations would still be zero. It 

 would merely require a longer time for 

 a heterogeneous population to reach 

 equilibrium. 



It is common experience that blue 

 eyes tend to go with light hair and a 

 fair skin. The degree of correlation, 

 however, is far too low to harmonize 

 with the hypothesis that there is merely 

 one series of factors for all melanic 

 pigmentation. Pearson^^ has calculated 

 the correlations between hair and eye 

 color in data from several nations with 

 the following residts : 



Sweden (conscripts) 250 



Prussian (school children) 271 



Italy (conscripts) 309 



German Jew ( school children ) 338 



Baden (conscripts) 354 



Great Britain (school children) 420 



The correlations are so low that they 

 might be suspected to be the result 

 merely of racial heterogeneity. A cor- 

 relation table involving all Europe 

 \\ould thus show a very high correla- 

 tion simply because dark-haired, dark- 

 eyed Spaniards and fair-haired, blue- 

 eyed Scandinavians do not intermarry. 

 It might be an accident from a physio- 

 logical standpoint that both dilution of 

 hair color and of eye color have become 

 associated in the Scandinavian people. 



Pub. No. 188, 1914, Tour. Her.. 5 :555-560. 



