.9/A 1 n'f/.fJAM SIEMENS, F.R.S. 45 



TESTING ON BOARD SHIP. Having tested the cable at the in- 

 sulating and at the sheathing works, we have now to coil it on 

 board ship into tanks filled with water, where it is again subjected 

 to daily electrical tests until it is finally submerged. It has 

 occurred frequently that the cable in running overboard has 

 suddenly shown a fault, and it is the duty of the electrician 

 instantly to arrest the operation. The paying out apparatus used 

 on board ship must be so arranged, that with the least delay of 

 time, the cable may be hauled back in order that the fault, which 

 in the meantime should be accurately determined by the elec- 

 trician, may be seen to, the moment it comes on board, and the 

 operation continued with the least possible delay. Formerly, no 

 tanks were used on board ship ; the cable was dry on board, and 

 it was impossible to know whether it was defective or not, until 

 it was submerged. Since the manufacture of the Malta and 

 Alexandria cable, when I had it in my power to apply water-tight 

 tanks for the first time, these tanks have been invariably adopted, 

 and the operation of paying out rendered more safe in conse- 

 quence ; the chance of any fault going overboard should indeed 

 be exceedingly small. Before the water-tight tanks were used, 

 the method of testing a cable during submergence consisted of 

 putting on the land station a clock, which made certain electric 

 connection, at given intervals of time, in the course of an hour ; 

 it made connection of the cable with the earth, with the receiving 

 instrument and for insulation at preconcerted intervals. On 

 board ship, the electrician would, during these intervals, be 

 prepared to test for insulation or for continuity, or to give the 

 necessary instructions to land. It is desirable to test during the 

 paying out as much as possible for insulation, because the im- 

 portant point is to have insulation perfect, the land clock was 

 therefore so arranged, that about three-fourths of the whole time 

 was reserved for insulation tests. At present, when a cable passes 

 from the water in the tank into the sea, it is not so necessary to 

 introduce these various tests. A method has lately been suggested 

 by Mr. "Willoughby Smith, which is very promising, of super- 

 seding the former ship tests. This consists in having always 

 insulation tests upon the cable ; and to connect at the land end a 

 length of insulated wire or condenser to the cable, which is in 

 connection with a very delicate galvanometer, such as a Thorn- 



