iH. XV.] THE COMPOSITION OF illND. 117 



Bomewliat wider range of meaning than the term "sensation." 

 Nevertheless the current use of the word " feeling" to desig- 

 nate indifferently a sensation or an emotion bears unconscious 

 witness to the fact that the two kinds of psychical state differ 

 only in their modes of genesis and of composition. The con- 

 trast between a peripheral sensation, as of. colour or touch, 

 and an emotion, is chiefly a contrast in degree of definitenesa 

 and of localization. But this contrast holds also between 

 peripheral sensations and such vague internal sensations as 

 hunger, which, being known as cravings, are assimilated to 

 the lowest orders of emotion. From this difference in defi- 

 niteness arises the fact that the peripheral sensations admit 

 of being definitely grouped according to their relations of 

 likeness and unlikeness, and thus afford the material for per- 

 ception and reasoning, while emotional states admit no such 

 definite grouping, but arrange themselves variously in clusters, 

 the particular character of the cluster being determined by 

 certain contemporaneous perceptions or ideal reproductions of 

 past perceptions. For these reasons the ultimate psycho- 

 logical nature of emotion can be reached only through a syn- 

 thetical interpretation which starts by recognizing the fact 

 that, along with that classifying of conscious states which 

 occurs in perception and reasoning, there goes on a recogni- 

 tion of certain states as pleasurable or desirable to retain in 

 •jonsciousness, and a recognition of certain other states as 

 painful or desirable to expel from consciousness. Thus in 

 practical experience emotions are, in however slight a degree, 

 inseparably associated with perceptions and inferences, as the 

 vague, internally-initiated feelings accompanying the definite 

 peripheral feelings in the classifying of which the perceptions 

 and inferences consist. 



Looking back, now, over the region already traversed, we 

 find that we have passed in review a large number of mental 

 operations wliich differ immensely in complexity, some of 

 ihem being performed only by the most highly-educated adult 



