42 THE LIMITATIONS OF SCIENCE 



tension or modification of their ideas. And of the 

 three, the most attention in an essay of this character, 

 which attempts a general discussion of the philosophical 

 basis of scientific theory, should be devoted to Pro- 

 fessor Larmor's ideas. In his treatise, lEther and 

 Matter, published in 1900, we have the rather rare 

 example of a scientific theory with a philosophical back- 

 ground clearly expressed and discussed. 



The main thesis of his essay is that a purely me- 

 chanical theory of discrete atoms moving in empty 

 space has failed to account for the phenomena of 

 nature, and especially for the recent discoveries made 

 in electricity. We may, however, by altering this con- 

 ception of atoms, by separating electricity and matter 

 into two entities, or even by considering matter as an 

 attribute of electricity, again reconcile fact and hypoth- 

 esis. Since Faraday's time the drift of opinion has 

 been in the direction of this separation, so that it is 

 merely necessary to crystallize it into a definite scien- 

 tific postulate. No idea of finality is expressed in 

 Professor Larmor's theory, since he thinks it should 

 endure only so long as it agrees with facts in our 

 possession; on the other hand, he considers it not to 

 be effective criticism to make a charge of incomplete- 

 ness without indicating a better way, as an hypoth- 

 esis may be valuable not only when imperfect, but when 

 quite wrong, providing it serves as a useful instru- 

 ment for the progress of natural philosophy. As an 



