132 REVERSION AND ALLIED PHENOMENA 



§ 5. " Skipping a Generation " 



There is every reason to believe that an individual inheritance 

 is like a mosaic, built up of many contributions, through the 

 two parents, from the grandparents and great-grandparents. 

 It is, therefore, a normal and frequent fact of inheritance that an 

 offspring should re-exhibit the peculiarity of a grandfather, 

 though neither of the parents showed it. There seems little 

 utility in calling a very frequent occurrence like this a reversion, 

 though it is of the same general nature. 



It is obviously difficult to decide where it is convenient to draw 

 the line. For how long must a character have been absent — 

 i.e. latent — before its re-assertion or re-awakening is to be called 

 a reversion ? It is a question of convenience, and our view is 

 that it rather obscures the issue to use a special term when the 

 " throw-back " is merely to a grandfather or even great-grand- 

 father. If the term had not been abused it would have been 

 convenient to call these minor family throw-backs " atavisms." 



A drone-bee arises from an unfertilised e^g ; it has a mother 

 and two grandparents, but no father. But it seems rather 

 absurd to call its resemblance to its grandfather either atavistic 

 or reversionary. This is a reductio ad absurdum, for the drone- 

 bee would resemble its father if it had one ! But the case may 

 serve to show that it is undesirable to use the term unless the 

 throw-back is to an ancestor more than two generations 

 antecedent. 



§ 6. Splitting of Hybrids 



We have seen that it is a confusion of thought to associate 

 with reversion what is known as filial regression, which is an 

 every-day occurrence in blended inheritance. Similarly, to 

 associate with reversion what is demonstrably due to arrest of 

 development or to inhibitory modificational effects of nurture 

 is a misimderstanding. We have also seen that, for a different 



