OTHER EXPERIMENTS ON HEREDITY 381 



It seems to us that the results depend in part on whether 

 there are or are not sufficiently well-marked contrasted unit 

 characters in the two parents. When the differences between 

 the two original parent-types are not crisply definable in terms 

 of contrasted unit characters, the conditions of Mendelian in- 

 heritance are not afforded, and we have to fall back upon the 

 old-fashioned description of the inheritance as " blended " or 

 "particulate" or ''reversionary," and so forth. 



It must be clearly noted that Mendelian phenomena are not 

 known except in certain cases of hybridisation. They chiefly 

 occur in the inbreeding of the hybrid progeny of two well- 

 marked varieties or " elementary species." We do not know 

 how far they may be found to apply in the breeding of pure 

 strains. 



Karl Pearson's studies on the inheritance of coat-colour in 

 horses and dogs and of ej^e-colour in man are as far as possible 

 from suggesting that Mendelian inheritance is illustrated in 

 these cases. It appears that the colour of coat or of eyes 

 in the offspring is a function of the ancestral rather than 

 of the parental characters. Yet, it is quite conceivable that 

 Mendelian inheritance may be demonstrated in horses, dogs, 

 and man — in cases where the parents do not contain a medley 

 of latent strains, but are sharply contrasted with one another in 

 respect to one or more unit characters. The danger is of trying 

 to universaHs3 the Mendelian formula, and some of the attempts 

 that have been made to give a Mendelian interpretation to 

 discrepant facts seem to us very far-fetched. 



There is, we think, much reason to believe that in some cases 

 the unit characters ; re represented in the germ-plasm by deter- 

 minants which are very stable in themselves, which must be 

 everything or nothing in the hypothetical struggle antecedent 

 to and associated with development, whose expression will not 

 blend with, or even allow of the expression of contrasted analogous 

 determinants. There is, we think, equal reason to believe that 



