81G Relative Motion of the Earth and the /Ether. 



so that it ought to be possible to detect an effect on the velocity 

 of light due to a magnetic field. The absence of any such 

 effect shows, therefore, that a magnetic field does not consist 

 of an aetherial stream along the lines of force, or else that the 

 aether is carried along by the earth. This latter supposition 

 is, however, generally allowed to be inadmissible, so that it 

 appears certain that a magnetic field is not an aetherial stream. 

 It is permissible, therefore, to regard the motion of the aether 

 as influenced by matter in the way suggested in this paper. 



The variation of the mass of a negative electron with its 

 velocity appears to agree with the view that its length alters 

 with its velocity in accordance with the principle of relativity. 

 It is, however, very doubtful whether the experiments done 

 on the variation of the mass with the velocity are of sufficient 

 accuracy to decide between the different theories as to tie 

 constitution of electrons. This is a point in favour of the 

 change of length, but it may be that only electrons change in 

 length and not pieces of matter, or it may be possible to 

 imagine a constitution for the electron giving the proper 

 variation of mass with velocity and not involving any change 

 of length. We know nothing about the actual constitution 

 of an electron, so that we cannot say that the observed 

 variation of the mass requires a change of length in matter 

 which certainly is not wholly composed of negative electrons. 



The absence of first order electrical and magnetic effects 

 due to the earth's motion is in agreement with the theory 

 here proposed ; and in the second order experiments, such as 

 Trouton's on the resistance of a wire, the apparatus was turned 

 about a vertical axis, and was inside a building, so that the 

 absence of any effect is again in agreement with the theory. 

 The extreme accuracy required to detect a second order effect 

 requires the apparatus to be only moved in a horizon! al plane, 

 for any rotation about a horizontal axis would certainly 

 involve distortion of the apparatus in consequence of its 

 weight. It seems, therefore, almost hopeless to attempt to 

 detect a vertical motion of the asther by any experiment of 

 second order delicacy. It may, however, be possible to detect 

 the horizontal motion in the neighbourhood of a body with 

 vertical sides. This could be tried by setting up a Michelson 

 and Morley apparatus underneath a thin horizontal roof of 

 large area, and then bringing a large mass near it on one 

 side, which ought to cause a shift of the bands varying with 

 the time as the earth turns round. The shift could be calcu- 

 lated on the supposition that the aetherial permeability of the 

 mass was large. The supports for the mirrors would have to 

 be arranged so as not to screen off the aetherial stream fro?n 

 the light-rays. This could perhaps be done by supporting 



