6 



Stokes, on the Nature of the Rontgen Rays. 



very great difference of nature between them and rays 

 of light that a very great difference in properties might 

 reasonably be expected. But assuming that the Rontgen 

 rays are a process which goes on in the ether, are the 

 vibrations belonging to them normal or transversal ? 

 If we could obtain evidence of the polarisation of those 

 rays, that would prove that the vibrations were not 

 normal but transversal. But if we fail to obtain evidence 

 of polarisation, that does not at once prove that the 

 vibrations may not after all be transversal, because the 

 properties of these rays are such as to lead us a priori 

 to expect great difficulties in the way of putting in 

 evidence their polarisation, if, indeed, they are capable 

 of polarisation at all. Several experimentalists have 

 attempted, by means of tourmalines, to obtain evidence 

 of polarisation, but the result in general has been nega- 

 tive. Of the two photographic markings that ought to 

 be of unequal intensity on the supposition of polarisa- 

 tion, one could not say with certainty that one was 

 darker than the other. Another way of obtaining 

 polarised light is by reflection at the proper angle from 

 glass or other substance; but, unfortunately for the suc- 

 cess of such a method, the Rontgen rays refuse to be 

 regularly reflected, except to a very small extent indeed. 

 The authors of the paper to which I have already 

 referred appear to have had some success with the tour- 

 maline. Like others who have worked at the same 

 experiment, they took a tourmaline cut parallel to the 

 axis and put on top of it two others, also cut parallel 

 to the axis, and of equal thickness, which were placed 

 with their axes parallel and perpendicular respectively to 

 that of the under tourmaline. But they supplemented 

 this method by a device which is not explained in the 

 paper itself, although a memoir is referred to in which 

 the explanation is to be found — at least by those who 



