94 A NEW THEORY OF EVOLUTION. 



the descendants of the judges appear to be 

 incapable of any eminence." 



And his statistics of the families of great 

 wrestlers and oarsmen show, that fathers 

 may have distinguished brothers or sons, but 

 no grandsons above the average. 



We do not think that Galton's statistics 

 prove " a regular average increase of ability 

 in the generations that precede its culmina- 

 tion." Sometimes pre-eminent ability sud- 

 denly appears in a single individual or gen- 

 eration (amongst brothers and sisters), and 

 disappears with the individual or generation. 

 Pre-eminence seems to be attained within two 

 generations, and declines in the third ; but 

 if either the son or grandson was not pre- 

 eminent, the fourth generation may still be 

 above the average. 



There is 110 clue to the origin or cause of 

 pre-eminence ; it springs up in wholly unex- 

 pected quarters, and if it most frequently 

 appears in gifted families, it also sometimes 

 arises in families otherwise unknown to 

 fame. 



The case is similar among cattle and horses. 

 Neither selection nor protection from the 

 struggle for existence will ensure pre-eminent 

 excellence, although protection is necessary 

 for its full development. 



