234 DEVELOPMENT AND HEREDITY. 



or which have sHghtly varied, give vigour and fer- 

 tility to the offspring." Mr. Havelock Ellis, after 

 investigating in detail the ancestry of men of gen- 

 ius, comes to the following conclusion : " Wherever 

 we find a land where two unlike races, each of fine 

 quality, have become intermingled and are in pro- 

 cess of fusion, there we find a breed of men who 

 have left their mark on the world, and have given 

 birth to great poets and artists." 



We have therefore to regard the consequences of 

 the sexual method of reproduction as of two kinds — 

 first as a profound stimulus to the general develop- 

 ment of the germ ; and second as a combining of 

 two sets of potentialities of development. The full 

 bearings of this latter consequence seem to have 

 been hardly recognised. The most certain and most 

 striking conclusions deduced from experiments in 

 breeding are — first, that uniting dissimilar forms 

 tends to produce an intermediate form ; and second, 

 that uniting similar forms tends to intensify the 

 peculiar characteristics of the parents. If new vari- 

 ations in form were really so scattered in time and 

 space, and affected only exceptional individuals here 

 and there, as has been sometimes supposed, it would 

 be difficult to imagine why " natural selection " 

 should maintain a method of reproduction which 

 would be constantly tending to obliterate all varia- 



