hy a Charged Condenser moving through Space. 157 



did not affect the steady position of the zero, but nevertheless 

 it was convenient to reduce them to a minimum. 



Gradual leakage of electricity from the charged plate could 

 not result in any permanent alteration in zero, even when 

 the coil was rotating, for, in view of the fact that the charging 

 connexion was made to the central plate at the top, the 

 magnetic flux through the coil due to such leakage would be 

 parallel to the brass base of the apparatus, and even if its 

 magnitude were worth considering, it could produce no effect 

 on the rotating coil. Only magnetic flux perpendicular to 

 the brass base could produce any effect on the rotating coil, 

 since the commutator was fixed so that when the planes of 

 the coils were parallel to the brass plate, a line joining the 

 commutator slits was perpendicular to it. 



Method of making the experiment. 



It was first necessary to set the apparatus so that a line 

 parallel to the brass base, and perpendicular to the axis of 

 rotation, should be parallel to the direction of the aether drift. 



Professor Trouton, in connexion with his work on the 

 " Couple exerted on a Charged Condenser moving through 

 Space"*, has determined the time of the day throughout the 

 year at which the rether drift is parallel to the earth's surface, 

 and he also gives the azimuth of the drift, and its magnitude, 

 taking into account the Sun's orbital and proper motions. 

 Though his calculations refer to London, it is sufficient for 

 the purpose in hand to take them as applying to Sheffield. 

 It is easy to find from Trouton's data the time at which the 

 drift is vertical, and the experiments in this paper were 

 always performed when the drift was either horizontal or 

 vertical. 



The apparatus having been adjusted to the correct position, 

 and the commutator having been well wetted with paraffin 

 oil, the motor was started. Any deflexion produced by want 

 of complete compensation of the coils was balanced by magnets 

 placed slightly asymmetrically, about 1*5 metres away. The 

 central plate was then charged, say positively, to about 

 4000 volts, as recorded by the electrostatic voltmeter. The 

 zero of the galvanometer having been read, the plates were 

 discharged by two knobs placed some distance away, and any 

 alteration in the galvanometer reading was noted. The 

 experiment was repeated several times, and another series 

 was then taken with the central plate charged negatively. 

 The whole apparatus was then turned through 180 degrees, 

 so that any magnetic field due to the earth's motion was 

 * Phil. Trans. A. 1904, p. 165. 



