106 COSMIC PHILOSOPHY. [ft. ii. 



whicli tlie?e rtlations are cognized as equal or unequal, like 

 or unlike — let us now advance a step farther. Already, in 

 the course of the foregoing analysis, the essential similarity 

 between reasoning and classification has been vividly brought 

 before us. We have now to scrutinize this similarity some- 

 what more closely. 



To cite an example with which we are already familiar : 

 when our astronomer, some thirty years ago, observed 

 that certain irregularities in the motions of Uranus still 

 remained unaccounted for, after calculating the combined 

 effects of all the interior planets in producing such irregu- 

 larities, it occurred to him that the unexplained irregularities 

 could only be due to the gravitative force of some undisco- 

 vered planet outside of Uranus ; and the discovery of Nep- 

 tune was the result of this most brilliant hypothesis. Now, 

 the mental act involved in this deduction was essentially a 

 classification of cases. The case of the unexplained pertur- 

 bations was mentally ranked along with the several cases of 

 explained perturbations presented by the solar system, as 

 being similarly due to gravitative force ; and to the number 

 of known cases in which planets deflect each other from the 

 regular paths in which they would otherwise move, a new 

 hypothetical case was added. Comparing, now, this mental 

 operation with that of the naturalist who, by virtue of certain 

 observed likenesses of structure and function, ranks together 

 lions, and elephants, and seals, in the class of mammals, we 

 may conclude roughly that the one process consists in the 

 formation of a group of like cases, while the other consists in 

 the formation of a group of like tilings. And since by the 

 expression " like cases" we mean merely "like sets of rela- 

 tions among two or more given groups of things," it follows 

 that we may characterize Eeasoning as the classification of 

 relations, while Classification, ordinarily so called, is the classi- 

 Hcation of things. When, for example, on perceiving two 

 similar triangles se'' side by side, we proceed to make some 



