SfX WILLIAM SIEMENS, F.K.S. 35 



back to the battery at the opposite pole from whence it proceeded. 

 The current will, according to Ohm,'s law, divide itself between 

 two conductors, inversely as their conductivity. If the arms, a c 

 and a d, are equal in resistance, and the two branches b c, and I d, 

 are also made equal, then there would be as much current pass the 

 one way as the other ; and no current will pass through a channel 

 which connects the two points, c and d. In this cross channel 

 there is a delicate galvanometer, the needle of which would deflect 

 with any passage of current crossways. But suppose I were to 

 make this resistance, b d, very great as compared with the resist- 

 ance, b c, then the current flowing through the branch, a d, could 

 not all of it proceed through d b, but would divide itself again, 

 part of it passing through the cross branch, c d, and the inferior 

 resistance,' b c, deflecting the galvanic needle. If we now increase 

 the resistance of b c until it equals b d, then the two branches will 

 again pass an equal amount of current, and no deflection of the 

 needle will occur. 



If b c be an unknown resistance, and b d is one which we can 

 alter ad libitum by putting in or taking out stoppers, representing 

 each a known resistance, we can so adjust the latter that upon the 

 application of current no deflection takes place, in which case the 

 one must be the exact measure of the other. That is one method 

 of testing which we employ. I have here, on this table, a 

 complete testing board, which has been taken out of actual use 

 only to-day, in order to be reinstated to-morrow morning, which 

 we employ for testing our cables, and in which as many as twelve 

 different methods of testing are contained. By simply altering 

 these stoppers, we can use the one method or the other. I may 

 here state that this apparatus, in order to be efficient, must be 

 constructed with extreme care ; it is a matter involving consider- 

 able mechanical skill to make a coil representing an exact amount 

 of resistance ; and it also requires considerable skill to make these 

 connections so that there are no false currents or loss, in order to 

 get natural and correct readings. 



These different methods I shall endeavour to pass in review. 



There is the direct method which I have just alluded to. Now, 

 in order to measure with the greatest amount of accuracy, it is 

 natural that the two channels through which the electricity flows 

 into the balance should be about equal. If you wanted to weigh 



D 2 



