584 SPECIAL METHODS OF DISCOVERY. 



either inductive or deductive ; we may either discover a 

 general law by inductive inference from facts, or deduc- 

 tively infer a new conclusion from a general law or theory. 

 We may draw new inferences from old knowledge some- 

 times, as well as from that which is new ; but new facts 

 constitute the chief source of discovery by this method. 



The method of discovering new truths by means of 

 inference is precisely the same in all cases, and consists 

 simply of rendering explicit (by means of a different form 

 of words) in a conclusion that which was implicitly 

 stated in propositions. In making discoveries by means 

 of this method, we either take the facts before us 

 just as they stand, and, without altering their form or 

 arrangement in any way, we at once draw conclusions 

 respecting them ; or we previously subject them to suit- 

 able preparatory intellectual operations. All scientific 

 knowledge (including the collection of various facts which 

 constitute the immediate results of an experimental re- 

 search) is liable to require previous mental treatment in 

 order to fit it for undergoing the process of inference. 1 

 The ideas must be selected by comparison, combined, ar- 

 ranged, and formed into definite statements or proposi- 

 tions ; we must also ponder upon them beforehand, and 

 consider them in every possible aspect. We must analyse, 

 combine, and arrange the evidence, in every possible way, 

 in order to extract from it the greatest number of new 

 inferences ; and the greater amount of knowledge or data 

 we possess respecting the subject, the more discoveries are 

 we thus enabled to make. In the act of pondering upon and 

 analysing the evidence, we add, subtract, multiply, divide, 

 combine, arrange, and permutate the ideas, transform 

 them either in part or whole into their inclusives or 

 equivalents, imagine contradictories of them, &c. &c., 2 



1 S e p. 333. z See p. 332 et seq. 



