402 THE ELECTRIC TELEGRAPH. 



of cable partly on board the Great Easteen and partly in the 

 sea. The resistance of the cable in such a position 

 would be, perhaps, 0,25 millions of units (representing an 

 average of 500 millions per knot length). The current 

 deflecting the needle of the ship's galvanometer G' with 

 the battery B, due alone to this resistance of the cable insu- 

 lator, would be 0,000004 B ; whilst the current due to the 

 cable and resistance R would be 0,00000401 B, neglecting 

 the resistances of battery and galvanometers. The addition 

 which the introduction of the resistance or shunt R makes, 

 therefore, to the indication of the ship's galvanometer, will 

 only be about 0,25 per cent, of the entire deflection a dif- 

 ference far too minute to be appreciated. The current from 

 the ship passes from the battery through G' ; after which it 

 is divided between the cable leakage and circuit B, A, G, R, 

 and earth. That portion of the current which takes this 

 course only the 0,0025 of the whole is nevertheless amply 

 sufficient to give a readable deflection of the needle of the 

 galvanometer G on shore ; and this deflection will continue 

 practically unchanged so long as there occurs no alteration 

 in the battery on board and in the cable no fault, the resist- 

 ance of which is so small as to prevent neglecting the resist- 

 ances G' and B in calculation. While it can, however, be 

 said with approximate truth that G -f- B = 0, a small fault 

 in the cable will not make itself known at the shore end ; 

 because it follows from Kirchhoff's laws that when the battery 

 resistance in any circuit is infinitely small, the addition of 

 shunts between the poles makes no difference in the currents 

 circulating in those circuits which were there before. Yery 

 different, however, will be the behaviour of the instrument 

 on board ship, which will indicate immediately the magni- 

 tude of the fault. The electrician will then determine its 

 distance by means of the method which we shall describe 

 directly. In order to do this, however, he will have to 

 measure the resistance of the copper conductor through the 

 fault, which he can do without sensible error by regarding 

 the end on shore as insulated. He will then have to tele- 

 graph to the clerk on shore to put the end of the cable 



