162 
MR. A. NORMAN SHAW: A DETERMINATION OF THE ELECTROMOTIVE 
The method of determination finally adopted consisted of a set of observations 
taken in the following order. It is thought that by this system of measurement we 
have been able to eliminate or correct for all appreciable extraneous effects. 
Order of Procedure. 
(i) A current equal in magnitude to that used in the deflection measurements 
(approximately an eighth of an ampere in each of the four windings, in parallel as in 
use) is passed through the suspended coils. 
(ii) r 2 and r 4 are adjusted until the magnetometer gives only a small deflection on 
the simultaneous reversal of both currents. This deflection is recorded. 
(iii) The magnetometer scale is calibrated in terms of a small variation in the shunt 
r 4 , and from the subsequent operation (xi), this can be expressed in terms of the Kelvin- 
Yarley slide. 
(iv) The connection C is replaced rapidly by the connections shown in dotted lines 
in the figure. Previous to the observations, the Kelvin-Yarley slide has been set at 
the reading nearest to the correct ratio; hence if the galvanometer circuit be con¬ 
nected immediately after, it will show only a small deflection. The current now 
flowing through the coils is smaller than in the first connection, and hence there is a 
slow change corresponding to the cooling of the suspended coils. The change in the 
fixed coils by themselves was found to be quite negligible. Galvanometer readings 
are taken 15, 20, 25, and 30 seconds after making the second connection. 
(v) The current is reversed and another galvanometer reading taken at 60 seconds; 
this will be in the opposite direction. 
