tH. XV.] THE COMPOSITION OF MIND. 103 



animal is like the relation between homologous attributes 

 in certain other animals." To confine onrselves to the first 

 clause of the description — "the attributes implied by the 

 term ruminant can be known only as previously observed or 

 described ; and the predication of these, as possessed by the 

 animal under remark, is the predication of attributes Hie 

 certain before-known attributes. Once more, there is no 

 assignable reason why, in this particular case, a relation of 

 coexistence should be thought, between * such attributes as 

 the possession of four stomachs and the possession of horns 

 and cloven hoofs,' unless as being like certain relations of 

 coexistence previously known ; and, whether the thinking of 

 this relation can be otherwise accounted for or not, it is clear 

 that the predication cannot otherwise have any probability, 

 much less certainty." ^ The case is the same with the re- 

 maining clauses of the description. In each instance the 

 mental operation performed by the naturalist is the recogni- 

 tion of the likeness between certain groups of relations 

 observed in this giraffe and certain other groups of relations 

 previously classified as pertaining to ruminants, ungulata, 

 mammals, and vertebrates. Obviously, therefore, the reason- 

 ing by which the places of animals in the zoological scale 

 are determined, consists in the compounding of cognitions 

 of likeness or unlilceness between certain given groups of 

 relations. 



So far, then, the mental operation performed by the natu- 

 ralist seems to be not unlike that performed by the astro- 

 nomer. And indeed, in spite of the superficial difference 

 which seems so widely to separate the classification of 

 animals from the measurement of celestial spaces, it will 

 appear, on a moment's reflection, that the only real difference 

 between the mental processes involved in the former case, 

 and those involved in the latter, is the extent to which like- 

 ness is predicated of the relations concerned. Deeply con 

 * fjpencer, Principles of Psychology, voL u. p. 69, 



