130 REVERSION AND ALLIED PHENOMENA 



ancestor with more than five digits ; but this is illegitimate, for the 

 so-called " heptadactylous ancestor " is a pure myth. Polydacty- 

 lism in man can only be called a reversion when there is in the 

 family history a previous occurrence of the same abnormality some 

 generations back. 



It occasionally happens that a particular part of the skin in man 

 exhibits a mouse-like covering of close-set hair. To interpret this 

 a mere random variation as a reversion is credulous in the ex- 

 treme. It may also be noted, incidentally, that to call the wool-like 

 covering of small hairs (the " lanugo ") on the human foetus a re- 

 version to a hairy ancestor is quite absurd ; it is a normal stage in 

 development quite outside the rubric of reversion. It may be an 

 inheritance from a distant past, but it is no more a reversion than 

 the occurrence of a notochord as a constant antecedent to the 

 development of its substitute, the backbone. Similarly the dog's 

 habit of turning round and round before it settles down to sleep may 

 be interpretable in the light of past history, but it has nothing 

 to do with reversion. 



" When horses are occasionally born at the present day in which 

 one or two accessory toes are present on two or even all four feet, 

 we are perfectly right in considering the development of these toes 

 to be due to reversion to an ancestor of the Miocene period." That 

 the modern horse which steps daintily on the tip of a single (third) 

 toe for each limb, and has merely hidden rudiments of the second 

 and fourth, has been evolved from a many-toed ancestor, is one of 

 the most certain of evolutionist inferences, but are we " perfectly 

 right " in interpreting the occasional development of supernumerary 

 toes, as on Julius Caesar's horse, to the reassertion of latent ancestral 

 items in the inheritance ? Is it not simpler to regard this as an 

 independent variation, comparable to multiplications of other parts 

 to which reversionary interpretations are inapplicable ? We must 

 remember, also, that vestigial organs are in many cases peculiarly 

 liable to vary. 



It ought not to be necessary to remark that the ancestor to 

 whom the organism is supposed to revert must be real, not hypo- 

 thetical. 



Some enthusiastic exponents of the reversion theory have not 

 scrupled to name or even invent the ancestor to whom the 



