PERSONAL QUALIFICATIONS FOR DISCOVERY. 241 



public or the most popular scientific man, nor he who is 

 apparently the most useful, who makes the greatest dis- 

 coveries, but the less known and the less appreciated pro- 

 found thinker and industrious original worker. A man's 

 real ability in science is much less to be measured by his 

 physical feats than by his intellectual ones ; and also much 

 less by the popularity accorded to him during his life by 

 partially scientific persons, than by the honour assigned to 

 him by succeeding scientific philosophers. The glory of 

 Newton and of Faraday has far outshone that of the most 

 popular expositor of science. 



A discoverer alone can best describe the mental con- 

 ditions essential to successful research, or by which disco- 

 veries are made, because he alone can best realise the 

 mental process passed through. But even he can only do 

 it imperfectly, because of the difficulty experienced by the 

 mind in observing its own actions. In discovering, much 

 of the success depends upon the man. The greatest dis- 

 coveries require a gifted and an active mind acutely im- 

 pressible by the slightest really exceptional circumstance. 

 The personal qualifications for achieving success in physi- 

 cal and chemical research are various ; the most important 

 are, an inherent sensitiveness to particular impressions of 

 similarity and difference, an ardent spirit of enquiry and 

 enterprise, suitable and sufficient knowledge of science, 

 strong scientific imagination and fertile invention, acute 

 observing faculties, accurate reasoning power, and apti- 

 tude in making experiments. A discoverer must be able 

 to imagine hypotheses, invent means of testing them, 

 manipulate in experiments, and infer causes and explana- 

 tions. The humble faculties of indomitable industry, 

 patience, and perseverance are constantly required to per- 

 form the drudging portion of searching for new truths ; 

 the faculties of imagination and invention are frequently 



R 



