FLUCTUATIONS AND MUTATIONS 151 



to the tendency to reversion, or rather regression, these fluctuating 

 variations cannot be accentuated or piled on one another by 

 selection continued through any number of generations. Accord- 

 ing to them evolution is founded solely on discontinuous variations 

 (' mutations '), which, though they may not always differ from con- 

 tinuous variations in magnitude, yet differ totally in kind and in 

 that they are absolutely stable. 1 It is thought that, if an individual 

 in whom one of these discontinuous variations appears has mated 

 with an individual who has it not, then a proportion of the 

 descendants inherit it and a proportion do not ; there is rarely a 

 blend. If it is unfavourable, the descendants that are of the new 

 type are eliminated, and it disappears from the race ; if it is 

 favourable, the ancestral type may be eliminated in the struggle for 

 existence and the new type established in its place ; if the variation 

 is neither favourable nor unfavourable, the two types exist side by 

 side. According to the Mendelian and mutationist view, therefore, 

 evolution does not proceed continuously (along a smooth incline, 

 as it were), but by a series of steps of varying but usually con- 

 siderable magnitude. The theory of evolution by discontinuous 

 variations finds its strongest advocate in Professor Hugo de Vries, 

 the distinguished Dutch botanist. His opinions differ, however, in 

 some very important particulars from those of the more extreme 

 Mendelians. 2 



250. The theory of continuous evolution (that is, evolution 

 founded on ordinary ' fluctuating ' variations) is sometimes termed 

 the ' selection ' theory, that of discontinuous evolution the ' muta- 

 tion J theory. Both theories, however, are doctrines of evolution 

 by the selection of favourable variations. They differ in that they 

 attribute evolution to the piling up of different classes of variations. 

 It is true, however, that mutationists often lay much less stress 

 on selection and adaptation than selectionists. 



251. Sir Francis Galton, who may be said to have founded the 

 biometric school, distinguishes three kinds of inheritance: (i) 

 Particulate, or inheritance " bit by bit, this element'from one pro- 

 genitor that from another," 3 as when black and white poultry are 

 crossed and the offspring have in varying degrees some feathers 

 black or barred, and others white. (2) Exclusive, when the 

 character of only one parent is reproduced to the total exclusion 

 of the other, as when the plumage of the offspring of black and 



1 Doubtless the opinion that discontinuous variations are absolutely stable 



is not held by all mutationists. Apparently, however, it is held by many of 

 them. See 285. 



8 See 282. 3 Natural Inheritance, p. 7, 



