DIFFICULTY OF CLASSIFYING MODES OF DISCOVERY. 455 



are made are so various that it appears almost im- 

 possible, in the present state of knowledge, to classify in 

 a systematic manner the numerous special ways in which 

 new scientific truths are found. An investigation is often 

 commenced for the purpose of testing an hypothesis ; less 

 frequently it is begun without any distinct preconceived 

 question, simply with the less definite object of wishing to 

 find what a given research or experiment will yield ; but 

 whether it is started with a speculation or not, hypotheses 

 are always raised during its progress. In nearly all cases 

 of discovery, we more or less investigate ; and as we cannot 

 command results, we obediently accept in all cases those 

 which nature yields, and rest content with such expla- 

 nation or conclusions as the results afford. Sometimes 

 we discover a new substance, occasionally a new action ; 

 at other times a new cause, effect, coincidence, or other 

 relation ; frequently a new fact, and occasionally a general 

 law or principle. In all these cases discovery commences 

 at the point at which old information ends and new know- 

 ledge begins, and the only two agents concerned in it are 

 external nature and the human faculties ; and by external 

 nature I mean anything external to the consciousness and 

 thinking power itself. 



The special modes of discovery are as varied as our 

 senses and mental powers, for it is by means of these and 

 their combinations all our discoveries are made. We dis- 

 cover a phenomenon or an existence by touching, tast- 

 ing, smelling, hearing, or seeing it ; by comparing and 

 classifying it with other phenomena or existences ; and by 

 means of reason and inference. They are also as varied as 

 the sciences themselves and as the properties and actions 

 of the forces and substances of which those sciences treat, 

 because every different force and substance must be 

 specially investigated according to a more or less different 



