SKIN COLOUR 171 



which occurs in a family is but a repetition in a small way 

 of a much larger blending which antecedently has occurred 

 in the community as a whole if the blending hypothesis 

 has any meaning at all. Professor Pearson has himself 

 endeavoured to convince us by his biometrical instruments 

 that a man's ancestry ten generations back is merely a 

 sample of his race or community.* If this biometrical 

 abstraction is true, then Professor Pearson's general 

 evidence as to the community of mulattoes, quadroons, 

 octoroons, and samboes holds equally cogently for single 

 families of such classes or vice versa. In a few alternative 

 words, we may state the problem thus : On the hypothesis 

 of blending we should expect not only the individuals 

 of separate famihes to manifest considerable uniformity 

 in skin colour, but under certain conditions we should 

 not expect any large variation in skin colour in the 

 community of mulattoes and the other hybrids as a 

 whole. The chief of these conditions would be, of course, 

 that the Europeans who originally married to the negroes 

 in a given island should have been members of the same 

 community. For even on the blending hypothesis, it is 

 conceivable that the factors determining French skin 

 colour may be different to those determining EngHsh or 

 Danish, or Spanish or German colour. But even that 

 condition does not carry us far. For the breeding together 

 of the mulattoes within a circumscribed islandic area 

 would soon produce an uniform result if blending be true. 

 Professor Pearson does not tell us which particular island 

 of the West Indies his evidence relates to. But we are 

 probably not far wrong in assuming that so far as the 

 nature of the population is concerned, it has remained 

 approximately stationary for the past 250 years. That 

 will give us something like ten to twelve generations of 

 inter-breeding of mulattoes. This is as long a time as is 

 required to give us an uniform race, on Professor Pearson's 

 own calculation. And, since that racial uniformity has 

 not resulted, it is, as far as it can be legitimately carried, 

 adverse to the hypothesis of blending. 



But apart from any considerations of this sort, there 

 remains the clear and definite family of samboes which 

 we have already described (supra page 169). Such large 



* Grammar of Science. Second edition, p. 456. 



