MUTATIONS THEORY OP DE VRIES 133 



and spontaneous variation arising among a number of indi- 

 viduals of the same species at a given moment, and breaking up 

 the parent species into several differentiated daughter species. 

 The difEerence between such mutations and ordinary variations, 

 according to De Vries, is that the former alone are hereditary. 

 Mutations may occur in any direction : some are useful, some are 

 harmful ; natural selection secures the survival of those that are 

 useful, and the elimination of those that are not. Thus, accord- 

 ing to De Vries, natural selection, although stripped of any 

 positive and " formative " value, has a negative value, inasmuch 

 as it ehminates the useless mutations. On the other hand, 

 De Vries agrees with NageU in referrrag the initiation of new 

 species wholly and solely to ratragerminal action, which manifests 

 itself periodically in the emergence of mutations. For selection, 

 as we have said, does not, according to De Vries, have any share 

 in forming mutations, but only in eliminating those that are 

 disadvantageous. 



Although the view of De Vries, that mutations play an im- 

 portant role in the origin of new species, is undoubtedly correct, 

 it is probable that such mutations are much more frequent in 

 the plant world than in the animal world ; and De Vries, whose 

 great talent aU must admire, has perhaps built too broad a 

 generalisation on his botanical basis. 



This is not the place to enter into the^detaUs of the theory ; 

 we only wished to point out how the mutation theory of De Vries 

 lends support to the doctrine of an internal phyletic force 

 advanced by Nageli. 



Against the view which sees in species nothing but a complex 

 of adaptations, it has been urged that numerous variations 

 occur in parts which are only of secondary importance and possess 

 no biological value. It may be remarked, in the first place, that 

 biological value has often been denied to parts which may very 

 possibly possess such a value, which in some cases, iadeed, 

 almost certainly possess it. Sarasin has maintained that the 



