ever, we determined that the theory was complex from the fact that 

 the relation between mechanical structure and chemical quality was 

 unintelligible. We do not know such a relationship to be a fact; 

 that is what the theory invites us to infer. We are required to infer 

 a relationship, although at present the relationship can in no way be 

 explained. If we knew that the relationship existed, we would not 

 be concerned with showing by what means it is accomplished. But 

 we have more difficulty in inferring the relationship to be a fact, be- 

 cause the relationship is at present unintelligible. Hyslop's theory 

 illustrates this point well. If we knew that spirits existed and could 

 communicate with each other and with us, the theory would not be 

 concerned with showing how they do it. The existence of super- 

 normal knowledge could be explained without such a demonstration. 

 But the theory asks us to infer that spirits have powers of communi- 

 cation between themselves and human beings. This inference is the 

 more difficult for us to make because the manner in which spirits 

 perform these functions is unintelligible to us. A theory is only 

 complex then, when it asks us to infer an unintelligible thing to be a 

 fact. 



Having considered the simplicity of the various theories, we shall 

 now study the relationship between the method by which the theories 

 were derived, and their simplicity. We have found that simplicity 

 may mean any one of four things. It may mean that the theory has 

 few concepts and few relations between them. Secondly, it may 

 mean that the materials with which the theory deals are things which 

 can be presented in experience, and not such as must be inferred 

 from it. Thirdly, it may mean that the reasonings of the theory are 

 easy to follow. And, fourthly, it may mean that the theory pos- 

 sesses few or no unintelligibilities. Now we shall find little relation- 

 ship between method and simplicity in the first three senses. We 

 have found that none of the theories here considered possesses many 

 concepts, and that, with the exception of the theory of ions, their 

 reasonings are easy to follow. Yet Locke's and Hyslop's theories 

 are derived by the method of assumption and Darwin's theory, the 

 theory of inorganic evolution, and the theory that matter is electrical 

 are based upon the experimental method. It is very evident that 

 the method of assumption might lead to a theory complex in struc- 

 ture, and containing difficult reasonings by increasing the number of 

 assumptions and the intricacy of their implications. And so the ex- 

 perimental method may lead to a theory involving many concepts and 

 difficult reasonings. The theory of ions is such a theory. The con- 

 ceptions involved in it are comparatively numerous. It involves the 

 conception that matter is made up of ions of different kinds; that 



[26] 



