101 



J5 and c have been very accurately measured in international ohms and their tem- 

 perature coefficients determined. TI' is a standard cell, either a Clark or a 

 Weston, and is connected in opposition to the storage cell A. As the current then 

 passes from A, if the resistance in B is properly adjusted, the E. M. F. of W will 

 just counterbalance the potential around the J5-branch and there will be no de- 

 flection of the galvanometer G when the key A' is closed. But increasing the re- 

 sistance in B, if A is constant, has the same effect upon the potential around the 

 ^-branch as decreasing the resistance in B would have. Hence we may choose 

 any resistance in the B branch that we may wish, and yet regulate the potential 

 about that branch by properly adjusting the resistance in B. This being true, let 

 B denote the total resistance of the -B-branch, including c and (/; let e denote the 

 resistance ot the wire a b; let E« denote the E. M. F. of the standard cell TT'^ and 

 Ec the potential about the wire a, 6 or c. Then, when the resistance in B^ and R 

 are so adjusted that we get no deflection of the galvanometer G when K is closed, 

 we have the proportion: 



B:C::Ew: Ec 

 Knowing the resistance c of a b and that of the total i?-branch, of course we know 

 the potential about c. 'Again, since the potential between any two points p y be- 

 tween a 6 increases directly as the resistance included between them, and since 

 the resistance increases directly as the distance between the points, we can find 

 any portion of the potential E,- by measuring off on the meter bar the appropriate 

 length along the wire a b. 



Another portion of the potentiometer consists of a third branch circuit 

 including two standard cells TFj and W^ which are to be compared. For short 

 we shall call this branch the JN'-branch. It starts from a movable contact p on 

 the wire a b and leads to the reversing commutator N, thence through TI', and W^ 

 to S through the galvanometer G back again through J\* to 7, another movable 

 contact on a b. p and 7 are knife-edged contacts and can be placed at any position 

 along the wire a h and the distance between them measured by means of the 

 meter rod. 



The two cells W^ and Wn, which are to be compared, are now placed in this 

 branch in opposition to each other. If, then, the E. M. F. of TT'^i is exactly equal 

 to that of Wo, that is if E] = E^, and if p and 7 are placed close together upon 

 C, then there will be no deflection of the galvanometer G when K is closed. If, 

 however, Ej is greater than Eo we can tind two points upon C such that the 

 difference in potential between jy ami 7 shall exactly equal the difference between 

 Ej and Eo and in opposite directions. The potentials in the N-branch will then 

 be at equilibrium, and there will be no deflection of the galvanometer. In other 



