FIRST PRINCIPLES n 



Let it stand as a metaphor. To the man whose knowledge 

 is partial and second-hand, the ease with which a specialist 

 who has the personal and intimate knowledge of his subject 

 which we describe can give the correct explanation of a 

 phenomenon which he observes for the first time, or can 

 judge between discrepant reports of observations, seems 

 to be too rapid for reason ; and the specialist himself finds, 

 when he attempts to give his reasons, that they hardly 

 justify the strength of his conviction. 



Perhaps it is permissible to use the expression "scientific 

 imagination" in a still larger sense. In the army of workers 

 who are advancing the boundary of knowledge there are 

 some who gain for it a furlong, while others move it forward 

 but an inch. And those who make the greatest advance 

 do so because they bring to bear upon their subject the 

 same mental qualities which constitute imagination in an 

 artist. The artist imagines new combinations of form, 

 colour, musical notes. The man of science imagines new 

 combinations of force, new conditions of action. Some men 

 are incapable of picturing anything outside the limits of 

 their experience others can devise new conditions, and 

 can foretell what would happen to inorganic matter or to 

 a living thing if placed in circumstances which, so far as 

 they know, have never concurred before. The develop- 

 ment of an acquaintance with Nature so sympathetic and 

 confidential as to allow the worker to share her secrets, 

 and to unite with her in designing new combinations, is 

 the highest result of scientific training. 



The Boundaries of Science. Science extends no 

 further than knowledge. Its agents the five senses collect 

 stores of facts upon which science lives and grows. It has 

 no traffic with the unknowable ; nor can it cross the border- 

 line which separates the world of the senses from the world 

 of consciousness, or barter its facts, gathered from the 

 external universe, for the equally real facts which the in- 

 dividual learns by self-examination. 



