SCIENCE AS A SYMBOL AND A LAW 19 



luminous bodies. Many times it has been announced 

 that the question has been decided experimentally ; for 

 instance, when Foucault found that the velocity of 

 light decreased when passing through transparent 

 bodies, instead of increasing as Newton's hypothesis 

 had predicted. But, on the other hand, the aberration 

 of light discovered by Bradley is a stumbling-block 

 to Huygens' theory and is a simple consequence of 

 Newton's. Neither of these experiments caused the 

 abandonment of a theory. A Newton could have 

 altered the properties of the corpuscle and a Huygens 

 is sure to arise who can alter the character of ethereal 

 waves so as to explain aberration; in fact, it has been 

 done with considerable success by Sir Joseph Larmor. 

 We may then take it as established, both on theoretical 

 and on historical grounds, that no experiment has 

 been, or will be, devised to decide finally between the 

 claims of the two hypotheses, yet the corpuscular 

 theory was abandoned. The reason was not that either 

 was impossible, but that the corpuscle, with the accre- 

 tions added to it as new facts were discovered, became 

 too unmanageable. Huygens' mechanical wave theory 

 having outgrown its usefulness has suffered the same 

 fate. He ascribed light to a series of mechanical 

 waves propagated through an elastic ether, but the at- 

 tributes necessary to the medium became so contra- 

 dictory that a new theory, advanced by Maxwell, was 

 accepted as a great relief. In this theory, the ethereal 



