SUBMARINE TELEGRAPHY 223 



the act of leaving the ship. That is the only portion which is 

 being disturbed, and it is hardly possible that a change can take 

 place elsewhere. If the fault be instantly detected, the ship 

 stopped and the cable arrested as speedily as is consistent with 

 safety, the fault may be retained on shipboard, or if it pass into 

 the sea, only a short length of cable will have to be hauled back 

 bef< re the faulty portion is recovered. As soon as it is quite 

 certain that a fault exists, and the necessary steps have been 

 taken to prevent the cable from running uselessly into the sea, 

 means must be adopted to ascertain where the fault is. One 

 rough method is to cut the cable in the hold near the part being 

 paid out, and then by examining successively the portions in 

 the ship and in the sea, to determine whether the fault is still 

 on board ; but electrical methods exist by which, before or after 

 the adoption of this simple examination, the position of the 

 fault can generally be fixed with considerable accuracy. Few 

 statements concerning telegraphy excite more surprise than this 

 does ; few people know that accurate measurement of electrical 

 phenomena is possible ; some even think that electricity is an 

 agent almost capricious in its action ; but those who have learnt 

 that the electrical properties of a conductor or an insulator are 

 susceptible of definite numerical expression should feel no sur- 

 prise on hearing that, when the electrical properties of a sub- 

 marine cable of uniform construction are observed to undergo a 

 definite change in virtue of some alteration at some one point, 

 it is quite possible to make such a series of measurements as 

 shall fix the position of that point. There are only two un- 

 known quantities, and whenever by experiment equations can be 

 obtained, including these unknown quantities, they can be de- 

 termined. Quitting generalities, let us try to show how this is 

 done. "We will first suppose that the simple insulation test has 

 shown that the conductor is no longer fully insulated. 



A measurement must be made of the resistance of the con- 

 ductor intervening between the ship and the sea at the fault or 

 earth, as, oddly enough, it is always technically called. If this 

 measurement give 40 units, and the resistance of each knot of the 

 cable is already known to be 4 units, the observer will know 

 that the fault cannot be more than ten miles off. It has already 

 been stated that the electrical resistance of a wire or conductor 



