MODE OF DISCOVERING A CAUSE. 425 



and thus still further narrow the extent of uncertainty. 

 This is only a crude illustration, by means of familiar 

 phenomena, of the very much more difficult and tedious 

 process employed in actual research. 



Or we may think of the most frequently occurring 

 causes of similar phenomena. For instance, if it is a case 

 of motion, we think of ordinary mechanical causes, vibra- 

 tion of the room, currents of air, &c., as being the most 

 frequent and therefore the most probable ; then of motion 

 produced by expansion by heat, or by electric or magnetic 

 attraction j &c. As the most probable cause of an event 

 usually is that circumstance which in the greatest number 

 of cases accompanies it, we also observe what are the con- 

 ditions and circumstances most frequently present ; and if 

 we perceive that a similar force is active, or similar con- 

 ditions exist in the phenomena under investigation as in 

 others which are known, we must at once infer that such 

 a force or conditions may be the cause, and we devise hypo- 

 theses which agree with this. If a phenomenon have one 

 antecedent which appears to be the only invariable one, 

 that one is probably the cause ; different antecedents may, 

 however, produce the same effect. 



In some cases, however, we arrive at this stage of the 

 enquiry at once and without sensible effort, because the 

 phenomena belong obviously to some particular force ; 

 and the next step is to discover the way in which the 

 particular form of energy operates to produce the effect. 

 To ascertain this, we think of each general mode or 

 principle of action, and each qualitative order and series 

 of constants of the particular force, in order to exclude 

 those which evidently have no relation to the effect, and to 

 find a case of resemblance. If a resemblance is found, 

 concordant hypotheses must be imagined, and suitable 

 ways of testing them invented ; but if no resemblance 



