84 BELL SYSTEM TECHNICAL JOURNAL 



values required to cover the possible variations in losses. A com- 

 promise is effected in this bridge by connecting a 10,000-ohm shunt 

 across each of the arms CD and AD. A slight difference in the losses 

 in these two arms can then be measured by varying one of these 

 resistances slightly. Since the standard condenser practically always 

 will have lower losses than the condenser tested, it is usual to place a 

 fixed 10,000-ohm resistance across CD and a resistance across AD 

 variable in 0.01-ohm steps to 10,000 ohms. A change of one ohm in 

 this resistance, when balancing a condenser, is equivalent to shunting 

 it with a resistance of 100 megohms or 0.01 micromho. Accordingly, 

 the conductance of a condenser may be measured in micromhos by 

 simply dividing the resistance change in ohms by 100. This, of 

 course, is only approximate in the case of large conductances, but is 

 correct to 1 per cent for values up to one micromho. 



Due to the condensers forming such an integral part of the bridge cir- 

 cuit, they are all built into the bridge. The complete bridge is shown in 

 Figs. 9 and 10. Fig. 9 is a top view showing the capacitance and 



Fig. 9 — Capacitance and conductance bridge 



resistance dials for effecting a balance, and Fig. 10 is a view with the 

 cover removed, showing the method of shielding the individual parts. 

 The range of capacitance is from 0.1 /xm f. up to three n f., and the fre- 

 quency range is from about 10 cycles up to about 150 kc, the only 

 modifications required in the bridges to cover this whole frecjuency 



