﻿Measurement of Electromotive Force. 559 



the temperatures for all differences of phase between the two 

 pulsations thermoelectric-ally by a zero method. Here, there- 

 fore, the instantaneous telephonic registry would have advan- 

 tages over the galvanometer, aiding the ear something after 

 the manner in which the stroboscope assists the eye. 



Contrary to my expectations, however, great difficulty was 

 encountered in endeavouring to make the telephone sensitive 

 enough ; and the work resulted in showing that the range of 

 increments of the magnetic field on both sides of zero, within 

 which the telephone does not respond, is out of proportion 

 with the sensitiveness of the instrument. It is the aim of this 

 paper to find the extent and character of this silent interval, 

 in its relations to the method in question. 



2. Literature. — The earlier history of the subject is sum- 

 marized in Wiedemann's Elektricitdt* , vol. iv. p. 285 (1885), 

 and need not be repeated here. Recentbv a series of remark- 

 able researches have appeared, due respectively to Prof. C. R. 

 Cross and his pupils f and to Lord Rayleighf. In the 

 former the excursions of the telephone diaphragm are mapped 

 out in their dependence both on the intensity of the magnetic 

 field and of the actuating line-current. The occurrence of a 

 marked maximum of sensitiveness for a certain intermediate 

 value of the magnetic field is clearly shown among other data 

 of value. The line-currents, however, are in magnitude above 

 those of the present paper. 



Lord Rayleigh's recent research covers much of the ground 

 of my own work §, and it leads to the same order of results, 

 in so far as data somewhat different in character (§5) are 

 comparable. My chief reason for publishing the present 

 paper is, therefore, the special point of view mentioned at the 

 end of § 1. 



3. Apparatus. — A diagram of the connexions, with special 

 reference to the construction of the key, is given in fig. 1. 

 The two circuits in action are a RE BC^ra and a T e A Csra, 

 where e is the weaker cell to be compared with the constant 

 cell E. R and r are rheostats, T the telephone, s a coiled 

 wire or spring introduced to allow easy motion of the metallic 

 plate C of the key. 



* The works of Warren De La flue, Brough, Pellatt, Ferraris, and 

 others being referred to. The theoretical researches of Tait and of 

 Preece are mentioned by Lord Rayleigh. 



t Cross and Hayes, Proc. Amer. Acad. xxv. p. 233 (1800) ; Cross and> 

 Phillips, ibid. Jan.'l893; Cross and Mansfield, ibid. May 1892. 



X Phil. Mag. xxxviii. p. 285 (1894). 



§ My note in the American Journal, October 1894, p. 346, was in press 

 before the advance account of Lord Rayleigh's work (in ' Nature ') 

 reached me. 



