I/O C. M. CHILD. 



since the real formative activity would in this case reside in the 

 environmental conditions not in the substance. 



Assuming then that the primitive germ-cell contains a complex 

 of formative substances all that has been said in the discussion of 

 preformation hypothesis regarding the nature of ultimate units 

 of organization would apply here. The complex must be physico- 

 chemical in nature and the formative factors must reside in the 

 relations between the constituents of the complex and in the 

 environment, not in the single constituents themselves, for a 

 single substance cannot of itself give rise to unlike substances, 

 definitely arranged. 



Most hypotheses of this character assume further that the ele- 

 ments of this ultimate complex are definitely and characteristically 

 arranged. This arrangement must be the result of physico- 

 chemical conditions past or present, /. c., the cause of the organ- 

 ization is to be found in relations not in particular substances, or 

 else we must again take refuge in vitalism and assume the exis- 

 tence of some organizing principle ultra-physico-chemical in 

 nature. 



The next question to be considered is how do the new forma- 

 tive substances which appear at any given stage and in typical 

 space relations arise from those previously existing? Here again 

 we must point to the physico-chemical conditions in the complex, 

 /. e., to the relations and interactions of the elements in a given 

 environment as the real formative factors, since a single substance 

 cannot of itself give rise to unlike substances in typical space- 

 relations. 



And finally, how do the last series of formative substances give 

 rise to the definitive structures? Here three points of view are 

 possible. We may assume as in the preformation hypothesis a 

 process of crystallization or something similar but by so doing 

 we account only for the form of the elements of structure not for 

 the characteristic arrangement of these elements. We may 

 attempt to account for this by assuming that the formative sub- 

 stances existed in a similar characteristic arrangement, but this 

 only throws the difficulty back to earlier stages. We must turn 

 sooner or later from substances to relations or else to vitalistic 

 assumptions. Secondly we may assume some grouping principle 



