200 Mr. H. Pender on the Magnetic 



This quantity therefore depends on the number of reversals 

 of the current per second, and the sensibility of the gal- 

 vanometer. The first idea which suggested itself was to 

 arrange a test-circuit, so that the deflexion produced by a 

 given current thus reversed in the test-coil T could be 

 determined alternately with the deflexion produced by 

 reversing the charge on the rotating disks, keeping the speed 

 of the reverser and the sensibility of the galvanometer 

 unchanged. This, however, was not practicable. In order 

 to insure a steady contact between the brushes of the reverser 

 and the brass sectors, it was necessary that the brushes bear 

 with some little pressure on these sectors. With the brushes 

 thus adjusted, however, the particles of brass gradually 

 rubbed off: soon formed a thin metallic bridge across the 

 intervening ebonite sectors, thus destroying the insulation for 

 high potentials. When the reverser was being used to 

 reverse the charge on the disks, it was therefore necessary to 

 have the brushes so adjusted that they barely touched the 

 brass, in which case the contact for the test-current was 

 destroyed. Moreover, when the brushes were adjusted first 

 one way and then another, the speed of course changed. 

 Hence another method of procedure was adopted. 



By a series of preliminary experiments it was found that, 

 on reversing a given current in the coil T as above described, 

 the factor A was directly proportional to the sensibility of the 

 galvanometer and to the number of reversals per second, 

 at least within the limits used, — 8 to 20 reversals per second. 

 Consequently, instead of determining A directly for each set 

 of readings, the sensibility of the galvanometer S, and the 

 number of reversals per second C, were determined instead. 

 Writing 



A=ASC, 



the above formula for v becomes 



The factor A was determined from time to time throughout 

 the experiment as follows : — The sensibility of the gal- 

 vanometer S was first determined. The brushes were then 

 properly adjusted, a known current i sent through the coil 

 and reverser, and the reverser and commutator set running. 

 The deflexion due to the current induced in I by this current i 

 in T was then determined ; also the number of reversals per 

 second C. The double deflexion 2D was the quantity measured 



