48 
MESSKS. G. F. C. SEARLE AND T. G. BEDFORD 
Let the poles of a battery of voltage E be joined to the points F, G (fig. 2), and 
let the resistance between F and G through the battery be T, Let F and G be also 
connected through the jDrimary coil, the fixed coils of the dynamometer, and a 
choking coil, the resistance of this part being It. Let the steady current, when F 
and G are not in contact, he Cq. Now let F and G be put into contact. The current 
in the primary coil then has the characteristic 
KdC/d^ + RC = o.(21), 
where K depends upon C, since it involves the value of cZB/c/H for the core of the 
choking coil (^ 33). 
After the current has sunk from Cg to some very small value, let F and G be 
separated again, hut before the separation let E he reversed. The characteristic is 
now 
K dOldt + (Pv + T) C = - E = - (R + T) Cg . . . . (22). 
The current now gradually attains the value — Cg. The whole process is joractically 
complete in a fraction of the time of vibration of the dynamometer coil. 
The key described in § 32 carries out this j^rocess exactly, provided that the 
I'esistances D are made zero. We find that the time for which no E.M.F. acts on the 
circuit through the j^rimary coil is about second, while a rough calculation shows 
that with our apparatus the current sinks very nearly to zero in that time. 
Suppose now that E, R and T are all increased ^9-fold. Then Cg remains 
unchanged, and for any given value of C, during the whole time of variation, dQ dt is 
also increased _^:>-fold. Thus, if the ratio R/T be kept constant as well as the 
maximum current Cg, dCldt for any value of C is jDroportional to E. Hence 
Z = X + Y = sE.(23), 
where 2 is a quantity independent of E. A method of finding Z similar to that 
employed for Y is now available. For if U, Z correspond to E, and Z' to E', then 
\j' — \J = Z' — Z, since W remains unaltered, unless the hysteresis depends upon the 
speed at ivhieh the magnetic changes occur. But Z/E = Z'/Y/, and hence 
Z (U' - U)E/(E' - E).(24). 
Thus, by observing the values of U found in two experiments with two values for E, 
the values of Z and Z' can be determined. 
Practical examples of these two methods are given in §§ 33, 41, 42. 
