xvni ORGANISATION AND EVOLUTION 419 



view. But we must observe : (a) the only ascertained 

 cause of determinate modification in the individual is 

 conation. () Conation modifies the individual structure. 

 (c) The cause of germinal variation remains wholly 

 unknown except in so far as it is due to the combination 

 of different gametes, and this leaves the variations of the 

 individual gamete unexplained. (a) The denial that 

 modification due to conation may affect the germ plasm 

 has been erected into a dogma on account of the difficulty 

 of understanding it, experimental evidence leaving the 

 question open. The early origin of the germ cells within 

 the body is not decisive because we do not know all the 

 ways in which distinct cells affect one another. It is just 

 possible that modifications impressed from without (e.g. 

 mutilations) should not affect the germ cells while 

 conations arising from within do in some way work upon 

 them. It is clear that any considerable change of the 

 parent organism must affect the environment of the germ 

 cells, and therefore call forth from them new efforts to 

 maintain themselves within the body tissues, and it is 

 conceivable that such effort should involve a modification 

 of structure congruous with and comparable to the change 

 in the tissue which surrounds them. We must not 

 therefore exclude the possibility that conation is at work 

 through the generations building up the organic structure, 

 and if that is the case the whole of organised life would 

 be referable ultimately to the working of Mind, the 

 " mechanical " factors being merely conditions prescribing 

 the limits of its operations. Such an explanation, which 

 would unify the whole evolutionary process, must, how- 

 ever, be left for the present as a mere possibility. 1 



1 Apart from this possibility it must be maintained that later research 

 on heredity has made any evolutionary theory of Instinct far more difficult 

 than before. According to the later lights we are to rule out the small 

 variations, the " fluctuations " about the parental mean that occur in every 

 individual. We are to rely on certain definite mutations of appreciable 

 magnitude occurring once for all and founding each time a new and 

 relatively fixed type. In relation to instinct this theory postulates a suc- 

 cession of divine chances. What first causes a spider to spin a web on 

 this view ? The addition to or the omission from the fertilised ovum of a 

 physical element How any physical element destined to produce web 

 spinning comes into existence plumply before any web has been spun is a 

 question which admits of no answer unless we appeal to special creation. 



E E 2 



