SO-CALLED FORMATIVE SUBSTANCES. 167 



ULTIMATE UNITS MOLECULES OR GROUPS OF SIMILAR 



MOLECULES. 



If the ultimate units are molecules or groups of like mole- 

 cules, then their properties as formative substances must be the 

 result of their physical and chemical properties as molecules. If 

 this be admitted the hypothesis based on these assumptions is 

 identical with or similar to the crystallization hypothesis. For 

 structure is primarily a definite and characteristic space-relation 

 of parts and an aggregation of like molecules cannot, so far as 

 we know, give rise to anything like definite and characteristic 

 fixed space- relations except by an arrangement of its molecules 

 according to physical laws. According to this view organic struc- 

 ture is made up of a multitude of component parts, each consisting 

 of like molecules, but each different from the others in compo- 

 sition and each arising from its corresponding ultimate unit. 

 But in the fully developed structure these different components are 

 definitely arranged, and indeed it is this feature the definite 

 space-relations of heterogeneous parts --that is the fundamental 

 characteristic of all structure. To what is this arrangement 

 due ? If we say it is due to the physico-chemical interactions 

 and relations to each other of the substances in a given environ- 

 ment, then we have departed from our premises, for in this case 

 the relations and the environment are more directly and imme- 

 diately formative factors than the substances. The substances 

 have formed the various components, but there is no substance 

 which can be made to account for their arrangement. 



If, on the other hand, we assume that the characteristic ar- 

 rangement of the heterogeneous parts is due to something 

 else than their physico-chemical interactions and relations in a 

 given environment, then we pass beyond the realm of physics 

 and chemistry and become "vitalists." 



In short if we start from a series of molecules or groups of 

 like molecules as the ultimate formative substances, \ve can, by 

 assuming a process of crystallization or something similar in each 

 substance, account for the development of each physico-chem- 

 ically homogeneous element of structure, but we cannot account 

 for the most important feature of structure, viz., the characteristic 

 arrangement of heterogeneous elements except by assuming rela- 



