Electric Arc between a Ball and Point. 129 



curve C shows the character of the cycle when no arc was 

 formed, and D, when the arc was in operation ; C, like A, is 

 approximately a curve of sines. The irregularity at its posi- 

 tive crest, which also appears in D, is probably due to the 

 imperfect performance of the contact brush, and, having no 

 bearing upon the phenomena which the curves are intended 

 to elucidate, may be disregarded. Since no current was pass- 

 ing between the ball and point when A and C were taken, 

 they represent the fluctuations in successive portions of the 

 same circuit. The lag, due to self-induction, however, is very 

 marked, amounting to almost 90° of phase. Curve D, which 

 shows the influence of the arc, is especially instructive. The 

 potential rises during the first part of the cycle (ball positive); 

 then follows a very sharp oscillation, occupying about one 

 twentieth of the entire period or 1/4600 of a second of time. 

 The potential then reaches a small positive value, which it 

 maintains without fluctuation for at least four tenths of a com- 

 plete cycle, when it suddenly becomes strongly negative. 



Fig. 3. 



/ 



\ 



A 



F V 



1 2 3 ■» 5 'P 8 S 10 11 13 13 14 15 10 ° v TLB \ 30 



! 



* t 



\ // 



\ \ 



°v 



To obtain curves of electromotive force between ball and 

 point directly, a non-inductive resistance was substituted for 

 the indicating galvanometer, and the measurements from which 

 curves C and D had been drawn were repeated. Of the two 

 curves thus determined, the one taken when the arc was not 

 playing (E, fig. 3) is a sine curve closely coinciding in phase 



