INDUCTIVE AND DEDUCTIVE MODES OF DISCOVERY. 461 



upon them, as we proceed. We first, by means of suitable 

 experiments, separate one by one all the interfering and un- 

 essential circumstances until we isolate the pure effect, and 

 are able to obtain uniform results, and have discovered to 

 what particular force the purified action or phenomenon 

 belongs. Having found the force to which it is due, we 

 continue, by a similar process of hypothesis, experiment, 

 observation, and inference, to exclude all the various 

 modes of action of that force until we arrive at one (or 

 more) which cannot be excluded without preventing the 

 effect, and to which therefore it is due, and which fully 

 agrees with all the observed results. But in the investiga- 

 tion and discovery of the properties of a substance or 

 force, we proceed in the opposite or deductive manner : 

 we cause it to act upon a number of other substances or 

 forces, and under the greatest variety of conditions, and 

 note the results; we then compare and classify those 

 results in every possible way, and extract from them by 

 induction every possible truth and general conclusion that 

 we are able. In one class of cases we assume the possibility 

 of a new effect, and then devise means of producing and 

 observing it ; and in another we devise a new cause, i.e. a 

 new combination of matter and its forces (or we take a cause 

 already known), and then ascertain its effects. 



4 The correctness of synthesis is proportionate to that of 

 the preceding analysis ; and a doubtful analysis may be con- 

 firmed by a synthesis. In other words, correct induction 

 furnishes the premises for a sound deduction, and a doubtful 

 induction must be verified by deductions from it.' ' A cor- 

 rect analysis of Icipis lazuli was suspected to be erroneous, 

 because there seemed to be nothing in the elements 

 assigned to it, which were silica, alumina, soda, sulphur, 

 and a trace of iron, to account for the brilliant blue colour 

 of the stone ; accidental synthesis, which was followed up 



