AS A MENTAL OPERATION 3 



Moreover, within rational inferences there are further 

 differences between induction and deduction, but espe- 

 cially that induction leaves us in a mere state of indefi- 

 nite generality, while deduction brings us down to 

 definite particulars. Take, for example, the prediction 

 of eclipses. Induction will tell us that, whenever an 

 opaque body intervenes between an illuminating and 

 an illuminated body, there will be an eclipse. But to 

 apply this mere generality, deduction is required to com- 

 bine it as a major premiss with a particular as aYninor 

 premiss that, on a particular day, hour, and minute, 

 a body, say the earth, will intervene between an illu- 

 minating and an illuminated body, say sun and moon, 

 and so to conclude that there will be an eclipse of the 

 latter at this particular time. In short, it is the function 

 of man as a rational animal to emerge from particular 

 to general inference, or from analogy to reasoning, and 

 therein to advance from induction to deduction. As, too, 

 action always takes place in particulars, induction must 

 be followed by deduction in order to carry our know- 

 ledge of laws into practice ; or, as Comte said, ' induire 

 pour de"duire afin de construire.' 



All three kinds of inference can give us knowledge 

 only by starting from experience, however indirect this 

 may afterwards become. All three, however little they 

 go beyond their starting-point, nevertheless carry the 

 mind beyond experience. Even analogy goes beyond 

 experience, as when a dog infers that he is going to get 

 his dinner before he gets it. General inference goes 

 farther : it rises to universals beyond experience ; for 

 nobody can experience all the members of a class, or 

 all the cases of a law in nature. What goes furthest of 

 all is deduction, because it carries the mind from uni- 

 versals to new particulars beyond experience ; in the 



B 2 



