EVOLUTION 487 



quota, the whole heritage being a blend of such contribu- 

 tions from a long series proportioned in a perfectly definite 

 manner. There is good reason to suppose that our results 

 substantially represent experience and observation for a 

 considerable number of characters in divers forms of life. 

 But if this be the true conception of all heredity, what 

 becomes of the phenomena included by biologists under 

 the heading of reversion and atavism ? Now I am 

 inclined to think that many observations recorded as cases 

 of reversion or atavism may be due to rather loose classi- 

 fication, because I so rarely find in the record any precise 

 description of a definite ancestor to whom the reversion 

 has taken place, nor any reasonable proof that the atavistic 

 feature belongs to a type from which the given type has 

 with high probability been evolved. Further allowance 

 must be made for singularities arising, as : — 



(i.) Extreme variations in blended inheritance. A 

 man 6' 2" may be born of quite short parents ; he would 

 be an extreme case of normal variation, the improbability 

 of which can be quite easily calculated from our stature 

 data. But if he had chanced to have had a very tall 

 great-grandfather, we may be fairly certain that he would 

 have been loosely termed a case of reversion. 



(ii.) Abnormal variations due to congenital malforma- 

 tion. These may be a result of imperfect nutrition of 

 the zygote, or be due to other sources, but if the result 

 bears any resemblance to a lower type of life it will be 

 classed as a case of atavism. Thus a man with a 

 remarkably long radius — possibly a malformation — or a 

 remarkably hairy skin will be said to exhibit atavism, and 

 probably compared with the anthropoidal apes. 



(iii.) Results of " skipping a generation." Certain 

 diseases and possibly other characters, although they 

 cannot be directly described as sexual characters, can be 

 transmitted as latent characters through one sex to the 

 other. Thus colour-blindness and tendency to gout may 

 be transmitted by a woman from her father to her son, 

 and such cases are very often spoken of as reversion.^ 



1 I have dealt more at length with such cases on pp. 292-298 of a Memoir 



