FORCE OF THE WESTON NORMAL CELL IN SEMI-ABSOLUTE VOLTS. 
161 
will be seen that whatever the pole strength may be, — for such dimensions is about 
7i 
unity in magnitude. It is not necessary to determine its value further, because 
substitution of our measurements show that is so small that we may assume 
that 
I 2 _ 2Gj 
lx 1',. 
( 6 ) 
and therefore from (5) and (6) we see that 
2Gj 
ia 
— 2 ttN 
■,0L 
A 
B 
{« a + {x-SfY 1 * {a 2 + {x + S) 2 }^ 
■ ( 7 ) 
We can now solve for a in terms of G 3 , I 2 /Ii, x, S, and %. G 3 is known, L/fi and 
x must be determined as accurately as possible ; $ is less than a tenth of x in magnitude 
and need not therefore be known to a high degree of accuracy, while f and >; 2 are 
needed only to rough approximations. (In the case of our experiment A = 1'001463 
and B = 1'001421, but the appearance of in the expression for g x lessens the 
apparent importance of and rj 2 . It can be shown that 
2<7i = 
2G 3 1 a 
+ 
{a 2 +(x-$yyi> {, a . 2 +{ x +syy l ‘' 
(1 + 0-001448) 1-iV 
9 1 
where (1 — came to (l — 0'001315), thus lessening the importance of £ 2 and tj 2 in 
\ 9\ / 
the determination of g x by this method.) 
(c) The Determination of the Ratio of the Currents f when the Magnets 
Undeflected. 
Ii 
are 
The electrical connections are shown in fig. 3. S represents the suspended coils, 
and F the fixed coils. They are always connected in such a way that the current 
flows through them in opposite directions. r x , r 2 , are suitable resistances with shunts 
r 3 and r 4 , capable of delicate adjustment. K represents a Kelvin-Yarley slide which 
had been calibrated for refined comparisons according to the method outlined in the 
Appendix. G is a galvanometer, and BB are batteries. 1, 2, 3, 4, 5, 6, represent 
the terminals in a key which enables a sudden change to be made from the connection 
CD to the connections shown in dotted lines. The resistance of C (a thick copper 
rod) in the one case, and of the connections between the terminals 1, 2, and the 
Kelvin-Yarley slide in the second case, are of such a value that they are quite negligible 
in our measurements. 
VOL. CCXIV.-—A. Y 
