EXPERIMENTS ON ETHER-DRIFT 27 



other beam. The time difference will now be in the opposite 



sense, and therefore the total time difference introduced by 



?D ir 

 turning through 90 will be "Vt-tti- In the later experiments of 



Michelson and Morley the shift of the interference bands to 

 be expected was 0*4 of the width of a band. The extreme 

 difficulty of carrying out an experiment of such delicacy can 

 be imagined. It is obviously essential that the distance D 

 should remain constant during the rotation of the apparatus, 

 and such rotation is apt to call into play forces which produce 

 a distortion sufficient to mask entirely the small expected 

 effect. In ordinary work disturbances of this kind are of no 

 consequence, but in experiments of this order of accuracy 

 they become so serious that every precaution has to be taken 

 to eliminate them. Eventually, the whole frame bearing the 

 apparatus was floated on mercury, and could then be turned 

 slowly by the application of very small forces. Many other 

 difficulties had to be overcome ; but at last spurious shifts 

 of the interference bands were reduced to a minimum, and it 

 was possible to state definitely that no movement amounting to 

 5 per cent, of the calculated value could be attributed to the 

 supposed drift of the ether past the earth. The natural way of 

 interpreting this negative result is to conclude that the ether in 

 the near neighbourhood of the earth is in reality carried with it ; 

 but, as will be seen later, there are other means by which it 

 could be accounted for. For the present, however, let us 

 consider other and subsequent attempts to detect the supposed 

 relative motion. 



The following experiment is probably not so well known as 

 the one just described, but it is extremely interesting because 

 it depends upon a property of the ether other than that of 

 conveying light. The method suggested is due to Prof. Trouton, 

 and the work was carried out by him in collaboration with Mr. 

 Noble 1 in 1903. The principle of the experiment is as follows: 

 Imagine a charged electric condenser to be moving relatively to 

 the ether in a direction parallel to its plates. A moving electric 

 charge is equivalent to a particular current of electricity, and it 

 is therefore to be expected to produce a magnetic field. When 

 the condenser is moving as indicated, we have two equal 

 opposite charges travelling in the same direction along parallel 



1 F. T. Trouton and H. R. Noble, Phil. Trans, vol. ccii. 1903. 



