MRS. H. AYRTON ON THE MECHANISM OF THE ELECTRIC ARC. 
317 
alternating current, of two frequencies, differing by many thousands of alternations 
per second, that would both give the same value of SV/SA. This would show that 
the horizontal part of the SV/SA frequency curve (fig. 10) had been found. 
A very good first test, however, is furnished by the curve connecting the P.D. 
between the carbons with the time, for this curve is unsymmetrical with respect to 
the corresponding current curve, when the resistance is affected, for the following- 
reasons. 
I have shown that the change in the area of the volatilising surface of the crater 
that is due to any change of current follows after the change of current and requires 
time for its completion. If, therefore, a superimposed alternating current is affecting 
the resistance of a direct current arc, the P.D. required for any given current must 
Fig. 10. Suggested curve connecting 8V/SA with, the frequency of a superimposed alternating current 
of constant root mean square value, when the direct current and the length of the arc are constant. 
be higher when the current is increasing than when it is diminishing. A current of 
10 amperes, for instance, would require a higher P.D. when it came after 9 and 
before 11 amperes than when it came after 11 and before 9 amperes, because in the 
first case it would be flowing through an arc of which the cross-section had been 
made by some current less than 10 amperes, and in the second by some current 
greater than 10 amperes. 
I have applied this test, with very satisfactory results, to some experiments in 
which it is quite certain that the alternating currents must have affected the 
resistances of the arcs, because they had frequencies of only 47 and 115-alternations 
per second respectively. 
The experiments formed part of a valuable series carried out in 1896 by Messrs. 
