386 PREFACE TO 



September, 1847, and appears to contain his deliberate judg- 

 ment as to the practicability of making a collection of natural 

 history, such as would be available for scientific purposes, in the 

 manner in which Bacon proposed to have it made. 



" That it is impossible (he says) to sever the business of experi- 

 ment and observation from that of theorising, it would perhaps be 

 rash to affirm. But it seems to me that such a severance could 

 hardly be effected. A transcript of nature, if I may so express 

 myself, that is, such a collection of observed phenomena as 

 would serve as the basis and materials of a system of natural 

 philosophy, would be like nature itself infinite in extent and 

 variety. No such collection could be formed; and, were it 

 formed, general laws and principles would be as much hidden in 

 a mass of details as they are in the world of phenomena. 



" The marshalling idea, teaching the philosopher what ob- 

 servations he is to make, what experiments to try, seems ne- 

 i cessary in order to deliver him from this difficulty. Can we 

 V conceive that such experiments as those of Faraday could have 

 preceded the formation of any hypothesis? You allude, I 

 think, to what has been done in the way of systematic observa- 

 tion with reference to terrestrial magnetism. And beyond all 

 doubt the division of labour is possible and necessary in many 

 scientific inquiries. But then this separating of the observer 

 from the theoriser is only possible (at least, in such a case as 

 that of magnetism) when the latter can tell his "bajulus" what 

 experiments he is to make, and how they are to be made. As 

 a matter of fact, the memoirs of Gauss, which have done so 

 much to encourage systematic observation of terrestrial mag- 

 netism, contain many results of theory directly bearing on ob- 

 servation ; e.g., the method of determining the absolute measure 



of magnetism. 



***** 



" Of course I remember that Bacon speaks of experiments to 

 be suggested by theory : as for instance in Solomon's house ; 

 all I mean is, that is seems doubtful whether a large collection 

 of facts can in most sciences be made useful, unless some theory 

 has guided its formation." 



Now I am quite willing to accept this judgment as perfectly 

 sound and just ; as pointing truly at the practical difficulties 

 involved in Bacon's scheme, and proving that it could not be 



