﻿Manchester Memoirs, Vol. xlix. (1905), No. 6. 9 



heterozygotes is also beyond doubt, and that a quarter 

 of the population produced by breeding them together 

 is composed of albinos remains to be demonstrated. 

 The Table on p. 6 therefore is in no wise a refutation of 

 Mendelian theory ; at any rate the suggested Mendelian 

 interpretation of the facts cannot be regarded as disproved. 

 Nevertheless the facts are no less in accord with 

 Galtonian theory, though in a different way than I 

 at first held : the manner in which I believed that 

 the truth of that theory would be borne out was as I 

 have said before, and as I wish to emphasize again, by 

 the gradual diminution of the albino-producing capacity 

 of each hybrid of successive generations. It now appears 

 that the manner in which that theory is borne out 

 by these facts is by the gradual invasion of the 

 4 hybrid ' ranks in successive generations by dominant 

 individuals bearing the external hybrid characteristics r 

 whether there will appear among the ' dominants ' or 

 even recessives a compensating number of hybrids, or even 

 whether this is demanded by Mendelian theory is a ques- 

 tion of fact which does not affect my argument. What I 

 want to point out is that I fell into the error of believing 

 that that which was true of the whole population was also 

 true of the individual — a mistake which, I believe, was 

 due to an attempt to discover whether certain phenomena 

 were evidence in favour of the one or the other of two 

 theories without appreciating the essential character of 

 either theory and much less their mutual relation ; to a 

 failure, in short, to realise that a biometric formula of 

 heredity is true only of large masses, the component units 

 of which in most cases unite at random, while the Men- 

 delian theory is an attempt to account for the hereditary 

 phenomena exhibited by the union of individuals care- 

 fully selected, by a theory of the constitution of their germ , 



