370 ELECTRICAL MEASUREMENTS 



When K is thrown to the right-hand stop and KI is depressed, 

 the condenser C x is charged through the ballistic galvanometer, 

 giving rise to a deflection BIX', this must be corrected for the 

 multiplying power of the shunt, m x . A similar observation with 

 K thrown to the left gives 01s, the multiplying power of the shunt 

 now being m s . If the damping of the galvanometer be the same 

 in the two cases, 



C* = Qx = 

 C s Qs 

 or 



If the damping is not the same in the two cases, the deflections 

 must be corrected (see page 116). Constant damping may be 

 attained by use of the Ayrton universal shunt, A. 



If the standard and the unknown are of very different capac- 

 cities, instead of shunting the galvanometer, different known 

 voltages may be used in the two tests, and an allowance made. 

 This procedure is convenient if an Ayrton shunt is not at hand. 



The proper value of the standard is one which will give a de- 

 flection about equal to that due to the unknown capacity. 

 Enough battery should be used so that the deflections may be 

 read with good precision. 



If the galvanometer be placed at a, the capacities may be com- 

 pared by discharging the condensers. 



In arranging the apparatus, care must be taken that the capac- 

 ities to earth of long leads and of the instruments, as well as the 

 capacities between the leads to the condensers, do not introduce 

 errors. For instance, the leads from KI to the condensers should 

 be short, of small wire and well separated from the leads to the 

 other side of the condenser, otherwise, a separate test must be 

 made to determine their capacity. To prevent errors from 

 leakage, the battery and all the wiring should be well insulated, 

 especially the keys K and K \ and the leads from KI to the con- 

 densers via K. 



Sources of Error. In reality this test is not so simple as might 

 appear, for the assumption has been tacitly made that the con- 

 denser is entirely charged or discharged before the needle of the 

 ballistic galvanometer has moved appreciably. Reference to the 



