EXPERIMENTS ON ETHER-DRIFT 35 



meter was so rapid that it was impossible to observe the effects 

 of rotation. Even now, when the wire is uniform throughout 

 and thickly covered with gutta-percha, and the stand and 

 enclosure are rotated bodily, variation of the current is not absent. 

 It is, however, sufficiently slow to admit of an effect of the 

 calculated amount being observed. No attempt is made to 

 reduce the galvanometer current actually to zero ; the effect 

 looked for is a change of the current then flowing upon rotation 

 of the apparatus. Another serious disturbance — and one which 

 at one time led us to suspect a positive result — is that due to 

 changes of resistance caused by stresses in the wires. At that 

 time the apparatus was fitted up in such a way that the slider 

 and balancing bridge did not move when the coils were rotated, 

 with the result that parts of the wires whose resistances were 

 being compared became twisted upon rotation, balance being 

 thus destroyed. This effect has now been removed by turning 

 the apparatus bodily, with the exception of the galvanometer; 

 and the whole strain now occurs in the wires leading to the 

 latter. It is now possible to say definitely that the motion 

 of the earth through the ether does not produce a change 

 of resistance of as much as 5 per cent, of the calculated 

 value. The result of the experiment, therefore, like those 

 previously described, fails to give any indication of the existence 

 or non-existence of ether-drift. All that may be stated is that, 

 if any alteration of dimensions of the wires does occur, the 

 specific resistance of the material forming it changes also in a 

 compensatory fashion. 



All the above experiments have been devised for the measure- 

 ment of quantities of the second order, i.e. quantities of order 



of magnitude (y J . There is one experiment, however, which 



has been carried out by Lord Rayleigh, 1 in which a first order 

 effect might be expected. Like the others, its result is negative ; 

 and it is to be noticed that the Fitzgerald-Lorentz shrinkage 

 theory will not account for this. It is well known that if a 

 beam of plane polarised light be incident upon a plate of 

 crystalline quartz, and pass through it parallel to its optic axis, 

 the light, on emerging, is still plane polarised ; but the plane of 

 polarisation is rotated through an angle depending upon the 

 thickness of the quartz plate used. It is usual to account for 



1 Lord Rayleigh, Phil. Mag. 6th Scries, vol. iv. 1902. 



