MEASUREMENTS 219 



often as strong, and its investigations at least as necessary 

 to our welfare, as those of the natural philosophers, and 

 it would be unmeaning to call even its uncertainties (which 

 are many) a defect. 



These are the main conclusions, but, before dismissing 

 the argument, I may perhaps be allowed to add one or two 

 corollaries. The first of these is that, in the absence of 

 exact measurements of subjective states, it is impossible 

 to demonstrate any exact correspondence between them 

 and events in the external world. No theory of psycho- 

 physical parallelism is susceptible of proof. It requires 

 quantification in both series, and this is only obtainable in 

 one. 



The same consideration forbids us to expect that the 

 method of residues will help us to obtain evidence for or 

 against the reality of a self-determining activity of the soul. 

 Attention is an essentially subjective process, and its direc 

 tion means the completeness with which the mind is diverted 

 from other objects, and concentrated on the object to be 

 attended to. Stated in this way, the problem is strictly 

 quantitative. But we have no mental quantities ; and no 

 degree of accuracy in the definition of the physiological 

 factors (even if it be granted that they are the real deter 

 minants) will ascertain the exact share in the direction 

 which is contributed by each. Without this knowledge, 

 we must always remain uncertain as to whether there is 

 a residue or not. 



The bearings of these conclusions on theories of agnos 

 ticism and naturalism are obvious, and do not require to 

 be further insisted on ; but it may be submitted that the 

 pretensions of science to universal dominion are better met 

 by an indication of the limits within which its methods 

 are applicable than by a general attack on the validity of 

 its first principles. A successful defence to all attacks of that 

 kind may, I think, be found in a reference to the triumphs 

 which have already been won by the use of the impugned 

 principles, and by the admission that they have no claim 

 to be regarded as anything more permanent than working 



