Wright: Color Inheritance in Mammals 



235 



wherever there is pigment of tlie proper 

 kind. Thus in such mammals as the 

 guinea-pig in which at least three in- 

 dependent sets of factors cHhite hair, 

 skin, and eye color, it has been foiuid 

 necessary to consider simultaneously 

 all of the effects of any factor in order 

 to analyze the heredity. It would seem 

 desirable to apply this method to 

 human data. The data of Holmes and 

 Toomis is admirably adapted to this 

 puipose. 



COMBINATIONS OF HAIR AND EVE COLOR 

 CONSIDERED AS UNITS 



Let us then consider as distinct color 

 varieties, the six combinations of hair 

 and eye color — dark, light, and red 

 hair each with dark or light eyes, and 

 investigate their relations. Tables in 

 which grandparents and parents are 

 compared together, parents and off- 

 spring, and oft'spring with each other 

 give fairly consistent results. By com- 

 bining all of the data into one table, 

 however, the results are brought out 

 more clearly, freed from some of the 

 irregularities due to small numbers. 

 The fact that both parental and frater- 

 nal correlations are usually about the 

 same, viz., .50 for characters which are 

 wholly genetic, will serve as a justifica- 

 tion for this procedure. In making u]) 

 the table, parents and oft'spring were 

 entered in both subject and relative 

 classes. In order that the fraternal 

 correlation should not be given undue 

 importance, each entry was weighted 

 inversely as the size of the sibship. 

 The usual method of making fraternal 

 correlations gives a familv of 15 a 



weight of 210, where a family of 2 has 

 a weight of only 2. It seems fairer to 

 weight a family according to the num- 

 ber of brothers of a random individual 

 and so give these families weights of 

 14 and 1, or 28 and 2 after doubling 

 to make comparable with the double 

 entries in the ]3arent-oft'spring correla- 

 tions. It may be said in passing that 

 the inimodified fraternal correlation 

 l)rings out the same relations as those 

 presented below, even more clearly. 

 The relations are most clearly brought 

 out by calculating the per cent of cases 

 in which each color combination occurs 

 in relatives of each kind. (Table 3.) 



11uis by reading down a vertical col- 

 umn we find that light hair with dark 

 eyes was the color combination of 

 17.0% of the relatives of brunettes, of 

 40.9% of the relatives of others with 

 light hear and dark eyes, etc. In the 

 tables D(L) means dark hair with 

 light eyes, R(D) red hair with dark 

 eyes, etc. 



The high degree of heredity of each 

 combination is the first thing which is 

 revealed by this table. This point, 

 however, shows that before going far- 

 ther it will be well to examine the data 

 for assortative mating. If light eyed 

 red haired persons have a special pref- 

 erence for another color some evidence 

 for which we have already noted, this 

 color should be in excess among rela- 

 tives, as well as red hair with light 

 eyes. Table 4 gives the per cent of 

 cases in which those of each color com- 

 bination married persons with the same 

 and the other combinations. This table 

 gives some idea of marriage prefer- 



T.-^BLE 3. — Per cent of the different color combinations among relatives of those irith each color com- 

 bination (reading horizontally). For measurement of degree of association of one color combi- 

 nation -with the others, read vertically. Based on 233 pairs of grandparent's, with parent, 608 

 of parent uith offspring and 242 random offspring ivith offspring. Each pair entered twice, 

 making table symmetrical before calculation of per cents. 



D(D) 



L(D) 



R(D) 



R(L) 



L(L) 



D(L) 



D(D) 41.7 



L(D) 16.8 



R(D) 11.7 



R(L) 13.9 



L(L) 11.3 



D(L) 17.1 



17.0 

 40.9 

 23.3 

 20.5 

 13.6 

 6.7 



1.6 



3.1 



10.7 



4.3 

 1.6 

 3.7 



4.7 

 6.8 

 10.7 

 21.9 

 5.2 

 3.4 



23.5 

 27.9 

 24.9 

 32.3 

 57.3 

 33.6 



11.7 



4.5 

 18.6 



7.0 

 11.0 

 35.5 



100.2 

 100.0 

 99.9 

 99.9 

 100.0 

 100.0 



