EXPERIMENTS ON ETHER-DRIFT 39 



always impossible terrestrial detection of ether-drift. There is 

 another alternative which, after all, may be the correct one, 

 namely, that it is not justifiable, particularly as so little is 

 known about the ether, to attribute to it all the properties of a 

 material medium, and expect it to behave precisely as we know 

 the latter would do under similar circumstances. Nothing can, 

 however, be stated with certainty, and the problem must be 

 regarded as not yet solved. The experiments described, with 

 the exception of Fizeau's, have not disclosed the information 

 expected and desired ; but they are surely interesting, if for 

 no other reason, because they show to what extreme accu- 

 racy measurements can now be pushed when circumstances 

 require it. 



