322 



SCIENCE 



[N. S. Vol. XL. No. 1027 



results have been used with plausibility by- 

 Davenport and the American workers to 

 elucidate the curious case of the mulatto. 

 If the descent of color in the cross between 

 the negro and the white man followed the 

 simplest rule, the offspring of two first- 

 cross mulattoes would be, on an average, one 

 black : two mulattoes : one white, but this is 

 notoriously not so. Evidence of some se- 

 gregation is fairly clear, and the deficiency 

 of real whites may perhaps be accounted 

 for on the hypothesis of cumulative factors, 

 though by the nature of the ease strict proof 

 is not to be had. But at present I own to 

 a preference for regarding such examples 

 as instances of imperfect segregation. The 

 series of germ-cells produced by the cross- 

 bred consists of some with no black, some 

 with full black, and others with interme- 

 diate quantities of black. No statistical 

 tests of the condition of the gametes in such 

 cases exist, and it is likely that by choosing 

 suitable crosses all sorts of conditions may 

 be found, ranging from the simplest case 

 of total segregation, in which there are only 

 two forms of gametes, up to those in which 

 there are all intermediates in various pro- 

 portions. This at least is what general 

 experience of hybrid products leads me to 

 anticipate. Segregation is somehow effected 

 by the rhythms of cell-division, if such an 

 expression may be permitted. In some eases 

 the whole factor is so easily separated that 

 it is swept out at once; in others it is so 

 intermixed that gametes of aU degrees of 

 purity may result. That is admittedly a 

 crude metaphor, but as yet we can not sub- 

 stitute a better. Be all this as it may, there 

 are many signs that in human heredity 

 phenomena of this kind are common, 

 whether they indicate a multiplicity of 

 cumulative factors or imperfections in seg- 

 regation. Such phenomena, however, in 

 no way detract from the essential truth 

 that segregation occurs, and that the or- 



ganism can not pass on a factor which it 

 has not itself received. 



In human heredity we have found some 

 examples, and I believe that we shall find 

 many more, in which the descent of factors 

 is limited by sex. The classical instances 

 are those of color-blindness and heemophilia. 

 Both these conditions occur with much 

 greater frequency in males than in females. 

 Of color-blindness at least we know that 

 the sons of the color-blind man do not in- 

 herit it (unless the mother is a transmitter) 

 and do not transmit it to their children of 

 either sex. Some, probably all, of the 

 daughters of the color-blind father inherit 

 the character, and though not themselves 

 color-blind, they transmit it to some 

 (probably, on an average, half) of their 

 offspring of both sexes. For since these 

 normal-sighted women have only received 

 the color-blindness from one side of their 

 parentage, only half their offspring, on 

 an average, can inherit it. The sons 

 who inherit the color-blindness will be 

 color-blind, and the inheriting daughters 

 become themselves again transmitters. 

 Males with normal color-vision, whatever 

 their own parentage, do not have color- 

 blind descendants, unless they marry 

 transmitting women. There are points stiU 

 doubtful in the interpretation, but the 

 critical fact is clear, that the germ-cells of 

 the color-blind man are of two kinds: (i) 

 those which do not carry on the affection 

 and are destined to take part in the forma- 

 tion of sons; and (ii) those which do carry 

 on the color-blindness and are destined to 

 form daughters. There is evidence that 

 the ova also are similarly predestined to 

 form one or other of the sexes, but to dis- 

 cuss the whole question of sex-determina- 

 tion is beyond my present scope. The de- 

 scent of these sex-limited affections, never- 

 theless, calls for mention here, because it is 

 an admirable illustration of factorial pre- 



