CH. xv.] THE COMPOSITION OF MIND. Ill 



together consists solely in the heterogeneity and definiteness 

 of the groups which are classified, and in the extent to which 

 the classifications are compounded. 



To such a statement, however, there is one essential qualifi 

 cation to be added. It is not strictly correct to say that the 

 classification involved alike in the most complex act of rea 

 soning and in the simplest act of perception is a classification 

 of groups of sensations and of the relations between them. 

 For, when an object is perceived, along with the sensations 

 actually present, there are remembered or internally-revived 

 sensations which enter into the classification, and these inter 

 nally-revived sensations are what we call ideas or images. 

 For example, &quot; when passing the finger over a rough surface, 

 the perception contains very much uiv,re than the coordinated 

 sensations immediately experienced. Along with these there 

 go the remembered visual impressions produced by such a 

 surface, which cannot be kept out of the mind, and in the 

 suggestion of wnich the perception largely consists ; and there 

 are automatic inferences respecting the texture and density 

 of the substance.&quot; So when we see an orange lying on the 

 table, the only sensation actually present and entering into 

 the case is the sensation of a patch of reddish-yellow colour 

 surrounded by other unlike patches of colour. The other 

 elements in the classification of which the perception consists 

 are ideas or internally-revived sensations of position, shape, 

 bulk, texture, juiciness, and so on. And now we discover 

 another point of difference in d::gree between perception and 

 .Basoning. While in perception some of the elements classi 

 fied must be sensations actually present, in reasoning all the 

 elements classified may be ideas or internally-revived sensa 

 tions. The sides and angles of the isosceles triangles which 

 the astronomer compares in estimating the moon s distance 

 are ideal sides and angles ; and the naturalist, in writing 

 about the classification of ruminants, deals solely with in 

 ternally-revived impressions of horns, hoofs, and multiple 



