36 PROCEEDINGS OF THE CANADIAN INSTITUTE. 



down in this express use, and no valid reason can be given why it 

 should not be continued in that meaning, which makes it include 

 every principle whereby one mental or bodily state or action suggests 

 another. These principles comprise similarity as well as contiguity. 

 (For present purposes, as not relevant, we may leave out of cou- 

 sidei-ation the principle of contrast.) Now, if the expression 

 "association of ideas," allowing to the word "idea" the necessary 

 widening of its application before referred to, be still used as 

 oi'dinarily, the association of ideas cannot be resolved into or 

 explained by the power of habit, for association by similarity cannot 

 be so explained or resolved. It stands alone as a distinct and 

 ultimate principle of our nature. Its working, though inextricably 

 intermingled with that of contiguity, is logically prior to it. The 

 absolute and indelible distinction between this principle of association 

 and the power of habit is seen in the fact that association by similarity 

 is primitive, whereas habit has express reference to the moulding 

 power of repetition. Only after repetition can a habit be formed ; 

 but similarity may produce its effects at once. It is possible to cite 

 instances of its working where there has been no previous experience, 

 and therefore no possibility of the formation of any result through 

 repetition, as e.g. in the identification hit upon by Franklin of the 

 similarity in nature between lightning in the sky and the phenomena 

 of electricity. 



Evidently then association of ideas cannot be resolved into habit, 

 unless at least the principle of similarity be left out of consideration. 

 The question then remains : — Can we bring habit in under association 

 of ideas as but a phase of the working of the principle of con- 

 tiguity "? We answer : Yes, in all probability. The hypothesis that 

 all the varied manifestations of habit can ultimately be shown to 

 be essentially similar in the woi'king of contiguous association, is 

 that which, in the present state of knowledge of the facts, most 

 commends itself to our acceptance. We note, however, that it yet 

 remains a mere hypothesis, not having been brought to the test of a 

 complete scientific induction. We accept it, therefore, in the mean- 

 time only provisionally, awaiting complete verification from further 

 reseai'ch. Its commendation rests, in brief, on the following con- 

 siderations : — 



{a) Habit and contiguous association alike have express reference 

 to the plastic power of repetition. Both recognize its necessity as a 



