THE PURE LINE HYPOTHESIS. 539 



lance — and there must be an underlying principle con- 

 necting these two types together. As seen above, some 

 Mendelians attempt to solve the difficulty by denying the exis- 

 tence of genuine blended inheritance, but some, including the 

 writer, cannot accept, saving in an almost metaphysical sense, 

 the view that the blending is merely apparent, and arises through 

 multiple factors. The whole notion of unit-characters has, as a 

 matter of fact, an only very limited application, for there is 

 reason to suppose with de Vries that a change in any one factor 

 involves some change in every other factor of the organism, and 

 thus no single factor can be identical in two individuals. 



5. Small or Fluctuating^ Variations. 



It is maintained by some that small so-called " fluctuating 

 variations," observable in any given character among brothers 

 or sisters having the same gametic nature in a Mendelian 

 sense, are non-inheritable. With such a view, arguing in a 

 vicious circle is singularly easy in interpreting experimental 

 results, since, if any particular expression of a character is shown 

 to be inheritable, it can be retorted that the fact of it being so 

 proves that that particular expression was not a fluctuating 

 variation, but a variation of an inheritable nature, which is 

 known as a mutation. 



The subject is, of course, of fundamental importance, for 

 if small, casual variations are not usually inheritable, it becomes 

 almost impossible to picture a gradual evolution of organic forms, 

 and the origin of the adaptation of organisms to their environ- 

 ment is well-nigh inscrutable. It may be assumed that 

 " fluctuating variations " are the variations in the expression of 

 a character in all those individuals of a family which possessed, 

 in the fertilised egg-cells from which they sprang, the essentially 

 unchanged factor governing the unit-character. 



In other words, a " mutation " denotes that there has been 

 a change in the factor, while in the case of a " fluctuating varia- 

 tion " it is supposed that the factor has undergone no change 

 whatever. 



The varying vicissitudes to which embryos are subjected 

 may modify the physical appearance or expression of a character 

 in the adult, and some consider that all small, fluctuating varia- 

 tions arise in this manner. Those that hold this view suppose 

 that, when there is no mutation, the factor remains entirely 

 unchanged, and consequently the power of hereditary trans- 

 -mission is identical, notwithstanding the fact that the visible 

 character in the different individuals may vary considerably. The 

 great contention of all who accept this interpretation is that 

 under ordinary circinnstances the factor, or that which repre- 

 sents the character in the fertilised egg-cell is, broadly speaking, 

 unmodifiable. and it can only be changed by some unusual and 

 at present unknown influences : and then the visible character as 

 a rule is profoundly changed and a mutation is the result 



Careful experimentation is necessary in this connection, since 



