tances are ideas. In order, therefore, to verify Locke's theory, the 

 mind itself must be directly investigated. We discover, then, that the 

 theories based upon the method of assumption and deduction are pe- 

 culiar in this respect, that their verification requires a direct investiga- 

 tion of the thing assumed to exist. No inference can be made con- 

 cerning the assumed objects, until their qualities are known. It is 

 notable that in order to verify these theories, they must be given a foun- 

 dation in experiment and that when verified they will no longer be 

 theories of assumption and deduction. If this conclusion is sound, 

 we can further affirm that there is only one method by which a veri- 

 fied theory can be reached ; for we found from our investigation that 

 there were only two principal methods of theorizing: the method 

 based upon assumption and deduction and that based upon observa- 

 tion and experiment. 



We conclude then that whether a theory is based upon the 

 method of experiment or upon the method of assumption, it is verifi- 

 able in so far as its materials are subject to experimental investiga- 

 tion. And, considering the theories which we have thus far investi- 

 gated, there are none but what we can hope in time to verify or dis- 

 prove experimentally. The individual variations, inheritance, and 

 natural selection of Darwin's theory; the ions and their properties in 

 the theory of ions; the changes of temperature and elemental struc- 

 ture in the theory of inorganic evolution ; and the inertia of matter and 

 properties of electricity in the theory that matter is electrical, have all 

 been to some extent investigated, and apparently nothing prevents their 

 being investigated in greater detail. In Hyslop's theory the spirits 

 have not been experimentally investigated; but if spirits exist, we can 

 hope by experiment to discover their existence and properties; and if 

 they do not exist we can hope to explain super-normal knowledge 

 without assumptions concerning them. So in Locke's theory, the 

 mind which Locke assumed has not been revealed by experiment; but 

 if it exists there is no reason to doubt but that it may be; or, if it does 

 not exist we can hope to explain ideas and knowledge in some other 

 way. 



If then our investigation covered all types of theories, we could 

 conclude that all theories are verifiable and that there is only a ques- 

 tion of more or less difficulty in verification. We cannot, however, 

 securely draw this conclusion until we have investigated a type of 

 theories which Imanuel Kant affirmed to be true a priori, prior to and 

 independent of all experience. Poincare has recently called the same 

 type of theories conventions, neither true nor false; suggested by ex- 

 perience but not of a nature to be verified in it. Among such theories 

 are those, that there are lines which never meet, however far pro- 



[33] 



