MRS. H. AYRTON ON THE MECHANISM OF THE ELECTRIC ARC. 
331 
cored, and never a greater negative value when both are cored than when the 
negative alone is cored, the conditions as to current length of arc, &c., being the same 
in all cases. But SV,/SA is zero when both carbons are solid, is greatest when both are- 
cored, and has always some positive value, however small, when either carbon alone' 
is cored. Consequently, when the superimposed alternating current alters the 1 
resistance of the arc, if all other things are equal, the sum of these two, i.e., SV/SA, is- 
more positive when either carbon is cored than when both are solid, and most positive 1 
when both are cored. 
The general effect on SV/SA of coring either or both carbons is given in the 
preceding paragraph, but with a given root mean square value of the alternating 
current SV/SA depends not only on the nature of the carbons, but also on the 
frequency of the alternating current, the magnitude of the direct current, and the 
length of the arc. To complete our knowledge of the influence of cores on the value 
of SV/SA therefore, we must examine the effect they produce on the curves 
connecting each of these variables with SV/SA when the others are constant. Take 
first the curves connecting SV/SA with the frequency of the alternating current. 
The Change Produced in the Curve Connecting SV/SA with the Frequency of the 
Alternating Current , by Coring Either or Both Carbons. 
ABC (fig. 14), which is copied from fig. 10, is the curve, connecting SV/SA with the 
current-frequency for solid carbons. Since for moderate frequencies SV/SA is always 
most positive when both carbons are cored, and more positive when one is cored than 
when both are solid, the curve when both carbons are cored must resemble D E F, 
and the curves for one carbon cored and the other solid must lie between ABC, and 
D E F, but we have no means of knowing which of the two will start the higher. It 
follows, therefore, that the frequency with which SV/SA becomes positive, if it is not 
already positive, for normal changes of current (i.e., for frequency 0), must be lower 
when one carbon is cored than when both are solid and lowest when both are cored. 
Thus, with the same direct current and length of arc, SV/SA may be positive for all 
four sets of carbons, as at the points C, P, K, and F, or positive for some and negative 
for others as at B N H and E, or negative for all. Moreover, since the true resistance 
of the arc is greatest when both carbons are solid and least when both are cored, and 
smaller when the positive alone than when the negative alone is cored, the horizontal 
part of the curve, which shows the true resistance of the aj/c, must be highest when 
both carbons are solid, next highest for + solid — cored carbons, lower for -j- cored 
— solid carbons, and lowest of all when both are cored. Hence, since the curve for 
two solid carbons starts lowest, it must cut all the others at some fairlv high 
2 u 2 
