FORCE OF THE WESTON NORMAL CELL IN SEMI-ABSOLUTE VOLTS. 
183 
given showing a splendid agreement between the calculated and observed results for 
x. It was thus possible to calculate accurately the correction for readings, but it was 
not always practicable to wait until the wires had made a complete recovery from any 
long period of distortion. Any further superimposed effects were eliminated, therefore, 
by (l) reversing several times at intervals of five minutes ; (2) taking readings at the 
usual interval after the last reversal; (3) reversing and reading again at the right 
intervals ; and (4) reversing and reading for the first position again. The mean of the 
first and third readings were taken with the second, and the mean of these for both 
scales gave a mean value of S, the double deflection on the scales, which could now be 
corrected according to the formula (ll). 
(c) The Electrical Connections. 
Fig. 6 shows diagrammatically the arrangement of the circuit for the deflection 
experiments. 
AB represents two ohm standards in series (2'00368 international ohms, see §VIII.) 
and AD represents two-tenth ohms standards in series (0'199625 international ohms, 
see §IX.); S and F represent the suspended and the fixed coils respectively; r x , r 2 , 
r% 9 r 4 , r 5 , are suitable non-inductive resistances and shunts capable of very delicate 
adjustment; Gj is a sensitive low resistance galvanometer, and G 2 a sensitive high 
resistance galvanometer ; EF is the potentiometer wire ; b 1} b 2 , b s , are batteries-; and 
X is the standard cell. Suitable commutators, not shown in the diagram, are 
arranged so that the current may be reversed (l) between A and C; (2) through the 
suspended coils; (3) through the fixed coils; (4) through Gj ; (5) through G 2 , 
(6) through EF. 
The battery b x was composed of sixteen storage cells (each forty-ampere hour 
capacity) arranged in parallel four by four, to give an electromotive force of between 
