370 SPARKS FKOM A GEOLOGIST^ HAMMER. 



overlook a metaphysical principle involved in the com- 

 parison and classification of concrete phenomena; but since 

 the principle clearly reveals itself to critical attention, 

 we must frankly acknowledge that the entire fabric of 

 physical science rests upon a truth grounded in the realm 

 of metaphysics; and that this is not for such reason a 

 truth " merely speculative " in the reproachful sense, but 

 a truth which is self-evident, and surer than any scien- 

 tific conclusion. To a certain class of minds such a 

 statement may not address itself with all the cogency of 

 a concrete proposition, but it may impress the necessity 

 of caution in vaunting scientific conclusion from sensible 

 phenomena as the most certain kind of knowledge, and 

 incomparably more substantial than the ethereal abstrac- 

 tions of metaphysics. 



IV. When, in the progress of our scientific investiga- 

 tions, we reach the stage of inductive inference, the pro- 

 cess of concluding from a part to the whole is based on an 

 assumption of the uniformity of nature, which is only 

 the concrete form of the principle that like results pro- 

 ceed from like causes. If unobserved phenomena belong- 

 ing to the same group with those on which the infer- 

 ence is based are not ascribable to the same cause, or 

 same kind of cause, we have no right to extend the in- 

 ference from observed data to these. But the principle- 

 of the uniformity of causation is accepted as more valid 

 than any inference which we may induce from any array 

 of phenomena, however extended. The inference may 

 express a bond of connection running through the phe- 

 nomena observed, and no others: it is therefore not a 

 causal bond. It may express a causal bond, but not the 

 deepest and strongest bond. In any such case the infer- 



