224 GENERAL CONDITIONS OF SCIENTIFIC RESEARCH. 



coveries are governed by casual coincidences,' but 

 ' Newton had entertained the thought of the moon being 

 retained in her orbit by gravitation as early as 1665 or 

 1666.' 1 The discovery of Oersted has been spoken of 

 as a casual isolated experiment. Yet Oersted had been 

 looking for such an accident, probably more carefully 

 and perseveringly than any other person in Europe. In 

 1807 he had published a work, in which he professed 

 that his purpose was to ' ascertain whether electricity; 

 in its latent state, had any effect on the magnet.' And 

 he, as I know from his own declaration, considered his 

 discovery as the natural sequel and confirmation of his 

 early researches ; as, indeed, it fell in readily and imme- 

 diately with speculations on these subjects then very pre- 

 valent in Grermany. In was an accident, like that by 

 which a man guesses a riddle on which his mind has been 

 long employed.' 2 



6 No scientific discovery can, with any justice, be con- 

 sidered due to accident. In whatever manner facts may 

 . be presented to the notice of a discoverer, they can never 

 become the materials of exact knowledge, unless they find 

 his mind already provided with precise and suitable con- 

 ceptions by which they may be analysed and connected.' 3 

 It has been said, ' By the accidental placing of a rhomb of 

 calcareous spar upon a book or line, Bartolinus discovered 

 the property of the double refraction of light.' But Bar- 

 tolinus could have seen no such consequence in the accident, 

 if he had not previously had a clear conception of single 

 refraction. A lady, in describing an optical experiment 

 which had been shown her, said of her teacher, ' He told 

 me to increase and diminish the angle of refraction, and 



1 Whewell, History of tlie Inductive Sciences, 3rd edit. vol. ii. p. 451. 



2 Ibid. p. 72. 8 Ibid. p. 701. 



