226 GENERAL CONDITIONS OF SCIENTIFIC RESEARCH. 



space which he possessed, resolved the phenomena into its 

 geometrical conditions. 'A believer in accident would not 

 have sought them ; a person of less clear ideas would not 

 have found them. A person must have a strange con- 

 fidence in the virtue of chance and the worthlessness of 

 intellect who can say, 1 even in the heat of debate, or the 

 recklessness of anonymous criticism, that 'in all these 

 fundamental discoveries appropriate ideas had no share,' 

 and that the discoveries c might have been made by the 

 most ordinary observers.' 2 



Scientific researches are rarely made in a haphazard 

 way, but nearly always by men with specially-trained 

 minds, and for the purpose of solving definite questions. 

 Pascal did not verify his theory of the weight of the atmo- 

 sphere by accident, but purposely had a barometer carried 

 to the top of a mountain (the Puy de Dome, in France) to 

 test his conjecture. The great majority of important dis- 

 coveries are also laboriously sought for. Oersted sought 

 for the true relation between electricity and magnetism 

 during more than fifteen years before he found it. Volta 

 had studied electricity for nearly thirty years, and had 

 invented his electrophorus and electric condenser before 

 he discovered chemical electricity. Faraday began to 

 search for a relation of magnetism to light in the year 

 1822, and discovered it in 1845. He also sought 'for an 

 experimental connection between gravity and the physical 

 forces during a great many years, but did not succeed in 

 finding it. Discovery, unless it be that of an isolated 

 fact, is always a more or less gradual process. We do not 

 at once find a general law or principle, even though we 

 may have correctly predicted its existence by an hypo- 

 thesis, because a general law requires a great variety and 



1 See Edinburgh Review, No. cxxxiii. p. 122. 



* Whewell, Philosophy of the Inductive Sciences, vol. ii. pp. 189-192. 



