246 PERSONAL PREPARATION FOR RESEARCH. 



science ignores mere authority and requires in its stead 

 reasonable evidence ; ifc requires us also to set aside all 

 human pride and approach the subject like a little child. 

 As scientific research cannot create new truths, but can only 

 reveal to us those which are consistent with the very 

 nature of things, we must accept those truths whether 

 they harmonise with our preconceived ideas or not. It is 

 an important qualification of a discoverer that he should 

 consider the limits of his own faculties, and not attempt 

 that which is unattainable, nor on the other hand be dis- 

 couraged by difficulties from attempting that which appears 

 to be within his power. The principles of rectitude and 

 the pursuit of scientific truth require the investigator to 

 treat that as certain which is certain, and that as un- 

 certain which is so ; to entertain only a cautious belief 

 of statements which cannot be proved or disproved, and 

 which therefore may be true or false ; and to rely upon 

 hypotheses only as far as they are supported by evidence. 

 An investigator should also be able to weigh the value of 

 scientific evidence, and to estimate to some extent the 

 relative degrees of generality and importance of different 

 scientific truths. In consequence of the great public 

 benefits which have already resulted from the applications 

 of science, the scientific mind is fast becoming recognised 

 as being pre-eminently the truthful one. Reject that 

 which is false, and hold fast that which is true, was a 

 great characteristic of Newton, and must be of every 

 scientific investigator. 



Great power of attention is another condition of success 

 in research. * The difference between an ordinary mind and 

 the mind of Newton, consists principally in this, that the one 

 is capable of the application of a more continuous attention 

 than the other ; that a Newton is able, without fatigue, to 

 connect inference with inference in one long series towards 



