MEASUREMENT OF A CONSTANT ELECTROMOTIVE FORCE. 547 



the armatures is A cos 27r</> ; for a contact made at this instant the 

 error is 



HS - A cos 



The condition C = is in reality never fulfilled; for the open 

 wire has a capacity of its own which may be appreciable if the coil 

 consists of a great number of windings. The influence of the 

 capacity of the wire may, however, generally be considered as 

 negligable.* 



This method has the great advantage that the wire whose resist- 

 ance is measured is not, as in previous methods, the wire of the coil 

 experimented with, but a separate wire, the temperature of which is 

 more easily ascertained. 



1125. Instead of making an instantaneous contact, the ends of 

 the wire might be connected with an electrometer arranged as in 



A 2 

 816.f We shall have in this way the mean square of the 



o> 2 H 2 S 2 



electromotive force, or sensibly - , which would be com- 



2 



pared by the same instrument to the difference of potential existing 

 between the two extremities of a resistance R traversed by a con- 

 stant current. The frame itself used as a galvanometer giving a 

 deflection 8 for this current, we shall have 



<oHS H 

 ^ = R tan 8, 



V/2 G 



or 



(oGS 



R = 



A/2 tan 8 



The greatest difficulty of this method is in getting an electro- 

 meter of small capacity but sufficiently delicate. 



1126. MEASUREMENT OF A CONSTANT ELECTROMOTIVE FORCE. 

 The method of M. Lorenz,| in which the electromotive force of the 

 disc is compensated by the fall of potential of a resistance traversed 



* LIPPMANN. Comptes rendus, Vol. xcni., pp. 813, 955. 1881. BRILLOUIN. 

 Ibid., pp. 845, 1069. 



f JOUBERT. Comptes rendus, Vol. xciv., p. 1519. 1882. 



J LORENZ. Fogg. Ann., Vol. CXLI., p. 251. 1873. JfW. Ann., Vol. xxv. 

 p. i. 1885. 



N N 2 



