28 SCIENCE PROGRESS 



planes separated by the thickness of the condenser. Owing to 

 this separation, the magnetic effects of the two charges will not 

 cancel one another, but there will be in the dielectric a resultant 

 magnetic field perpendicular to the direction of motion and 

 parallel to the plane of the condenser. If, however, the 

 direction of motion be perpendicular to the latter plane, equal 

 and opposite charges now travel along exactly the same paths ; 

 and, hence, are on the whole equivalent to a zero current. In 

 the former case, therefore, there is associated with the condenser 

 a certain amount of magnetic energy which is absent in the 

 latter. On this account it was expected that such a condenser, 

 freely suspended with its plane parallel to the ether-drift, would 

 tend to turn through a right angle, so as to transform this 

 surplus of energy ; it could, however, be prevented from doing 

 so by exerting a couple of the proper amount in the opposite 

 sense. The detection of such a couple was the aim of the 

 experiment, but, as in the work of Michelson and Morley, the 



quantity to be measured was of the order (^ j , u and V having 



the same significance as before ; and it was only after much 

 work and many failures that definite information was obtained. 

 To be exact, if N is the electrostatic energy of the condenser, 

 the magnetic energy associated with it, when it is moving 



parallel to its plane with velocity u, is N . (^ J and the couple 



tending to turn it, which attains a maximum value when the 

 plane makes an angle of 45 with the direction of motion, is in 



the latter position also measured by N . (^) . 



With regard to the details of the actual experiment, the 

 condenser, composed of successive sheets of mica and tinfoil, 

 was suspended by a vertical phosphor-bronze strip, which 

 served, not only as the means of controlling the movements of 

 the condenser, but also for charging one set of plates. The 

 other plates were earthed by means of a wire dipping into 

 sulphuric acid, and a difference of potential of 2,000 volts was 

 maintained by means of a Wimshurst machine. A mirror 

 attached to the condenser viewed with a telescope and scale 

 indicated its movements. Many precautions had to be taken to 

 avoid disturbing influences, such as electrostatic attractions 

 between the condenser and the walls of the enclosing case, air 



