692 F8YCH0L0OT. 



awakens the idea h of its past contiguous associate B, lio-w 

 can this occur except through first reviving an image a of 

 its own past occurrence. This is the term directly con- 

 nected with h ; so that the process instead of being simply 

 A — h is A — a — h. Now A and a are similars ; therefore no 

 association by contiguity can occur except through a previ- 

 ous association by similarity. The most important su])po- 

 sition here made is that every impression on entering the 

 mind must needs awaken an image of its past self, in the 

 light of which itis'apperceived' or understood, and through 

 the intermediation of which it enters into relation with the 

 mind's other objects. This assumption is almost univer- 

 sally made ; and yet it is hard to find any good reason for it. 

 It first came before us when we were reviewing the facts of 

 aphasia and mental blindness (see p. 50 ff.). But we then 

 saw no.need of optical and auditory images to interpret opti- 

 cal and auditory sensations by. On the contrary, we agreed 

 that auditory sensations were understood by us only so far 

 as they awakened wow-auditory images, and optical sensa- 

 tions only so far as they awakened now-optical images. In 

 the chapters on Memorj', on Keasoning, and on Percep- 

 tion the same assumption will meet us again, and again 

 will have to be rejected as groundless. The sensational 

 process A and the ideational process a probably occupy 

 essentially the same tracts. When the outer stimulus 

 comes and those tracts vibrate with the sensation A, they 

 discharge as directly into the paths which lead to B as 

 when there is no outer stimulus and they only vibrate with 

 the idea a. To say that the process A can only reach these 

 paths by the help of the weaker process a is like saying 

 that we need a candle to see the sun by. A replaces a, 

 does all that a does and more ; and there is no intelligible 

 meaning, to my mind, in saying that the weaker process 

 coexists with the stronger. I therefore consider that these 

 writers are altogether wrong. The only plausible proof 

 they give of the coexistence of a with A is when A gives us 

 a seTise of familiarity but fails to awaken any distinct 

 thought of past contiguous associates. In a later chapter 

 I shall consider this case. Here I content myself with say- 

 ing that it does not seem conclusive as to the point at issue ; 



