36 



Mr. A. Campbell : Measurements 



higher arms, and hence had to be measured. This was done 

 with sufficient accuracy by passing a current through them, 

 and measuring the potential-drop on the galvanometer. To 

 eliminate thermoelectric effects the galvanometer circuit was 

 kept closed, and the balancing was done by reversing the 

 battery. 



^' Two- Step " Bridge. — This is shown in fig. 4. A suitable 

 small resistance, v/hose value need not be known exactly, is 

 inserted at U, and is adjusted by shunting until the gal- 

 vanometer balances in position (a). The galvanometer is then 

 brought into position (/>), and a balance obtained by another 



Fig. 4. 



X 



T 



'■^ 



K/\N\N\N\N\NV^' — AAA 



-;3 



\/WV^AA^' 



.i\AAAA J 



5 



shunt at S. With a suitable galvanometer this method is 

 considerably more sensitive than the Thomson Bridge, but it 

 is less convenient since it necessitates balancing twice. 



Differential Galvanometer. — The method o£ using this is so 

 well known* that very little description is necessary. I£ the 

 galvanometer circuits (100 ohms each) be called a. and y5, a had 

 in series with it a small adjustable resistance c, while in series 

 wdth ^ was a non-inductive copper coil of 900 ohms. This 

 coil was placed near the galvanometer, the w^hole being in a 

 double-w^alled box lagged with w^adding. The a and /3 cir- 

 cuits were first connected in series and opposition, their 

 magnetic effects on the needle being adjusted to balance in 

 the usual manner by a small coil in series with one of them 

 Then^ by adjusting ^, a and /3 were set to balance on the 10 

 and 1 ohm coils. Without altering this adjustment a balance 

 was now^ obtained on the coils (A and B) to be compared, by 

 shunting one of them ; and thus the ratio of A to B w^as 

 determined. This method w^as only used for the 0*1 ohm 



* See Heaviside's Papers, vol. i. ; also C. W. S. Crawley, Journ. 

 Instn, of Electrical Engineers, April 1901. 



