EXPERIMENTS ON ETHER-DRIFT 29 



draughts, and heating effects. These eventually having been 

 practically eliminated, times were chosen when the supposed 

 ether-drift was horizontal, and the plane of the condenser placed 

 at 45 with its direction. It was then charged to a difference 

 of potential of 2,000 volts, and a deflection looked for. The 

 calculated value of the deflection, taking into consideration the 

 earth's orbital motion only, was 3*4 centimetres, or, allowing 

 also for the sun's proper motion, and experimenting in March at 

 times when the resultant motion is horizontal, 6*8 centimetres. 

 No deflection approaching either of these values was obtained, 

 the largest barely exceeding 5 per cent, of the calculated 

 value, and the authors considered the observations as being 

 conclusive against the existence of any such effect as that 

 sought. 



Here, then, are two negative results which, on the face of 

 them, point to the conclusion that the ether near the earth is 

 dragged with it. But if this be admitted, difficulties at once 

 arise in the ordinary explanation of aberration. Regarding the 

 ether as behaving like a material medium, a velocity varying 

 with the distance from the earth would produce an aberration 

 of the stars differing for different latitudes. There is no record 

 of a variation attributable to this cause having been observed, 

 and, apparently, this cannot be due to lack of accuracy in 

 measurement. The alteration of the angle of aberration pro- 

 duced by the motion of the earth on its axis has been detected, 

 and, as it only amounts to about 2 per cent, of the iwhole at 

 the equator, it is quite unlikely that the comparatively large 

 variation which one would expect to be caused by a viscous 

 ether should have been overlooked. Besides this consideration, 

 there is the result of the experiment performed by Sir Oliver 

 Lodge, 1 in which he aimed at detecting a drag on the ether near 

 rapidly revolving steel discs. There is no space to fully describe 

 the method of procedure here. Suffice it to say that two parts 

 of a beam of light, after passing by opposite cycles round the 

 interspace between two parallel discs, were caused to produce 

 interference bands. If, now, the discs were rotated, and in 

 doing so dragged the ether with them, the one part of the 

 beam would be accelerated and the other retarded. This would 

 produce a shift of the bands, but so little was observed that it 

 was concluded that the velocity of light between the steel plates 



1 O. J. Lodge, "Aberration Problems," Phil. Trans, vol. clxxxiv. 1893. 



