202 . Mr. H. Pender on the Magnetic 



this corresponds to 39,800 turns of the disk. The speeds of 

 the two disks agreed very well. The average speed of the 

 two is the quantity recorded below under N. 



(3) The sensibility of the galvanometer. This was deter- 

 mined as already described. 



(4) The difference of potential between the disks and the 

 condensing- plates — the condensing-plates were always earthed. 

 This was determined by means of an absolute guard-ring 

 electrometer, the same instrument as that used by Rowland 

 and Hutchinson in 1889. The disk attached to the balance- 

 arm and its guard -ring were earthed. The lower plate was 

 connected to the disks through the reverser. By a simple 

 device it was so arranged that this connexion was made only 

 when the disks were charged in one way — positively, say — 

 so that the disk of the electrometer did not fluctuate up and 

 down as it would have done had the lower plate been con- 

 nected permanently to the disks. Had the lower plate been 

 connected directly to a pole of the Voss machine, the potential 

 measured would have been higher than that of the disks, 

 since in the interval during which the disks were disconnected 

 from this pole, its potential would have risen, in consequence of 

 the storing of the charge in the leyden-jars, as the Yoss machine 

 was kept rapidly rotating all the time. The potential then 

 measured by the electrometer was that to which the disks 

 were brought when charged positively (say). By reversing 

 the switch S, the potential to which they were brought when 

 charged negatively was similarly determined. The mean of 

 the positive and negative values of the potential thus deter- 

 mined is the quantity recorded below under V. 



The difference of potential between the two plates of the 

 electrometer is 



In this laboratory # = 980*1. A, the area of the disk, cor- 

 rected for the annular space between it and its guard-ring, is 

 30-11. Hence V= 17*56 \/M. 8 is the distance apart of the 

 plates, read from a scale attached to the lower plate. M is 

 the mass on the scale-pan ; 8 and M are the quantities deter- 

 mined for each measurement. 



A series of measurements was made as follows : — First one 

 observer determined the sensibility of the galvanometer. The 

 zero was noted, the test-current 2 x 10~ 8 amperes sent 

 through, a reading made, the connexion broken, and a zero 

 again noted. This was usually done three times, and the 

 mean deflexion taken as the sensibility. The commutator 



