Resistance of the Electric Arc. 347 



the number of cells in circuit producing the light, increasing 

 therefore the current without interrupting it, and without 

 changing the distance between the carbons, as the lamp 

 had no automatic adjusting arrangement. Readings of A 

 and V were then quickly taken and the operation reversed — 

 that is, the bridge-piece put into D and that in C quickly 

 withdrawn, which had the effect of again reducing the 

 current ; and if the change back again were effected not long 

 after the first, the carbons were not sufficiently burnt away 

 by the stronger current to make the light go out when the 

 current was reduced, so that a third set of readings of A and 

 V could be taken. In this way, for the same distance between 

 the carbons two readings of the lower current and its corre- 

 sponding carbon difference of potentials, and one intermediate 

 reading of the higher current with its carbon difference of 

 potentials, were obtained. The whole experiment was now 

 repeated with the cells P C and P B instead of P D and P C. 

 The following is a sample of the results obtained from a 

 number of tests with 30, 40 and 50 Grove cells : — 



Difference of Work in foot- 

 Number of Current, in potentials between pounds per 

 cells. amperes. carbon, in volts, second in arc. 



30 6-52 30-4 146-2 



40 10-16 30-4 227-8 



50 11-92 30-4 267-2 



The last current is therefore nearly double the first, but the 

 difference of potentials between the carbons is not materially 

 altered by the increase of the current and the light. 



Subsequently a large number of experiments were made 

 using a Brush dynamo in place of the Grove cells, and in- 

 creasing and diminishing the current by suddenly increasing 

 and diminishing the resistance in circuit without stopping the 

 current. 



In the earlier experiments for each current its value w T as 

 read on the ammeter as well as the difference of potentials 

 between the carbons on the voltmeter; but since even with a 

 very dead-beat ammeter some little time must elapse when 

 the currents are alternately doubled and halved by taking 

 out and inserting resistance in circuit, and since even with 

 a slight delay the stronger current burns away the carbon 

 points very rapidly, and so makes the distance between them 

 for the stronger currents greater than for the weaker, it 

 was thought better in the later experiments merely to take 

 readings of the voltmeter when the resistance was altered 

 backwards and forwards sufficiently to alternately treble and 

 diminish to one third the current as shown by the earlier ex- 



