404 Dr. C. V. Drysdale on the Use of the 



in which the P.D. to be measured is applied between the 

 needle and one pair of quadrants, while that derived from 

 the potentiometer is connected between the needle and the 

 other pair of quadrants, in each case through a considerable 

 resistance, say one megohm. By connecting a key to the 

 opposite pairs of quadrants, which normally short-circuits 

 them, but disconnects them when depressed, a deflexion 

 should be obtained to right or left proportional to Vi 2 — V 2 2 

 or 2Y dY when balance is nearly obtained. Any want of 

 symmetry in the electrometer should be eliminated by con- 

 necting the standard cell as shown, so as to obtain a double 

 weighing method. 



The procedure in this case is very simple, and exactly 

 resembles ordinary direct-current potentiometry. The sole 

 difficulty lies in the electrometer. It should be possible by 

 multicellular devices, or otherwise, to obtain a fair amount 

 of sensitiveness with 1 or 1^ volts, the P.D. of the standard 

 cell, as has been done by Mr. Addenbrooke, but w T hen we 

 come to a tenth of a volt or less, as is so often necessary in 

 current measurement, electrostatic devices seem hopeless. 

 All other magnetic or thermal devices are incorrect in 

 principle as requiring an appreciable current to operate 

 them, which militates against the accuracy of the potentio- 

 meter ; but by employing a low-resistance potentiometer, 

 and a differential thermal device, such as could be made 

 from the Duddell thermo - galvanometer employing two 

 junctions and two heaters, or two crossed thermo-j unctions 

 in series, a sufficiently sensitive and accurate potentiometer 

 €ould perhaps be made. 



In view of the difficulty of making such devices, however, 

 the writer has turned his attention to the second method, i. e. 

 of balancing the alternate current P.D. to be measured 

 against a known alternate current P.D., by the interposition 

 of a sensitive alternate-current detector, such as a telephone 

 or vibration galvanometer. This of course implies that the 

 two voltages under comparison must have the same magnitude, 

 frequency, and phase, and, approximately, the same wave- 

 form *. 



The equality of frequency is of course secured by deriving 

 the test and comparison voltages from the same source of 

 supply, but some special device is needed for bringing the 

 phases into coincidence. In some experiments recently 



* Since this was written the writer has been informed that Dr. Sumpner 

 has somewhere suggested that an alternating-current potentiometer could 

 be made, if an}' means were available for bringing the phase of the current 

 in the slide-wire into coincidence with that of the P.D. to be measured. 



