iv MENTAL DEVELOPMENT 



43 



of actions by which I bring my past reading to bear on the 

 whole task of writing this book. It helps to correlate a 

 long train of experience with a long series of co-ordinated 

 activities. This is a correlation effected by that and other 

 acts of consciousness together. Now the whole of both 

 trains may also be grasped by consciousness more or less 

 adequately in a single act and in that sense become a corre- 

 lation in consciousness. But in the first place, that will be 

 another act of consciousness quite distinct from the first 

 and more comprehensive ; in the second place, a mind 

 capable of the lesser, simpler synthesis might not be capable 

 of the wider one, so that it might build without ever 

 knowing what it is building or reviewing what it has built. 

 The elements correlated in consciousness then do not 

 necessarily coincide with the factors of life correlated by 

 consciousness, and in comparing different phases of corre- 

 lation we must take account both of what goes on within 

 the conscious area and of what is effected thereby. Again, 

 if the two things do coincide, it may be only the correlation 

 effected by consciousness that is susceptible of proof. We 

 have no direct knowledge of that which passes in the mind 

 of another. We judge analogically on the basis of our 

 own experience and of the behaviour of others. In the 

 case of animals, their behaviour differs so far from the 

 human as to throw a shadow of doubt on all interpretations 

 of what is actually passing in the animal mind. The solid 

 basis of our argument is always the correlation which the 

 mind actually effects. We find, for example, certain 

 external stimuli affecting the organism. We find subse- 

 quently a certain modification of behaviour conducing to 

 a result beneficial to the organism and bearing a definable 

 relation to the stimuli. That in such a case the effects 

 of certain experiences are so brought into relation by forces 

 acting within the organism as to conduce to its benefit, is 

 then a hypothesis susceptible of the ordinary methods of 

 inductive proof or disproof, and the result is independent 

 of any theory of the precise mechanism by which the corre- 

 lation is effected. Generally, then, our problem is to dis- 

 tinguish forms of correlation according to the data which 

 may enter into them in each case and the use made of 



