376 



ELECTRICAL MEASUREMENTS 



Kz is depressed and the unneutralized portion of the charge sent 

 through the galvanometer, the deflection of which will be to the 

 right or to the left according to which charge preponderates. By 



D 



successive trials, altering - the galvanometer deflection may be 



Ml/2 



reduced to zero. Then 



p Rl 



To carry out the test a special key which combines K\, K 2 and K^ 

 on a single base is usually employed. In manipulating this key 



FIG. 222. Connection for Thomson method of comparing capacities. 



care must be taken that it performs its functions properly and 

 that cross-contacts by the fingers of the observer are avoided. 



To obtain 'a good precision the variable resistances Ri and Rz 

 should be high so that their adjustment may be sufficiently flexible. 

 If resistance boxes are used the smallest step is usually 1 ohm, so 

 when comparing ordinary condensers the resistance which is ad- 

 justed should be at least 1,000 ohms. In submarine cable work 

 much higher resistances are employed; Ri -f R% may be as much 

 as 100,000 ohms. The galvanometer must be very sensitive and 

 the battery voltage as high as is consistent with safety of the 

 apparatus. With high voltages, to avoid throwing an unduly 

 large potential on either condenser, the known and unknown 

 capacities should be about the same. 



When cables are tested, the core is connected to a, and b then 

 becomes the common "ground"; in this case perfect insulation 

 of the battery is essential. 



