FALSE REVERSION 133 



reason, there is no advantage in associating with reversion the 

 common phenomenon known as " skipping a generation " or 

 resembHng a grandfather. There is no confusion of thought in 

 the association just mentioned, but it is unnecessary. Reversion 

 is the reassertion of latent characters inherited from a more or 

 less remote ancestor. 



In the same way it is unnecessary to complicate the concept 

 of reversion by including the common phenomenon known as 

 the splitting of hybrids. It is now well known that in certain 

 cases the pairing of similar hybrids — e.g. by self-polUnation in 

 flowers — is followed by the so-called Mendelian phenomenon, 

 that a definite proportion of the offspring return to the two 

 original parental types. As this is a constant phenomenon in 

 certain well-investigated cases, it seems unprofitable to speak 

 of these as reverting to the parental types. Reversion is an 

 unusual and rare phenomenon, a harking back to a more or less 

 remote ancestor. 



A tabby-coloured cat of the Persian breed, crossed with a 

 white male of the same breed, produced four kittens, two tabby- 

 coloured and two white. The two white offspring were interbred 

 and produced four kittens, two pure-white like the sire and 

 grandsire, two tabby-coloured like their grandmother (Ewart). 

 There appear to be many similar phenomena, and it seems in no 

 way advantageous to apply the term "reversion" to what is 

 merely a reassertion of grandparental characters. 



§ 7. Reversion in Crosses 



False Reversion or Yicinism. — In his criticism of cases which 

 have been labelled " reversions," De Vries draws a sharp dis- 

 tinction between " true reversion," due to unknown internal causes 

 which induce long-lost latent ancestral characters to assert 

 themselves, and " false atavism or vicinism," which is due to 

 crossing. His investigation of a large number of cases led him 



