ABSTRACTS OF SCIENTIFIC PAPERS 349 



VII. On the Construction of Submarine Telegraph Cables. ' Pro- 

 ceedings of the Institution of Mechanical Engineers,' July 1862, 

 p. 211. 



After naming the essential features of a submarine cable, the 

 author points out that ' for even* given ratio between the cost of the 

 materials of the insulator and of the conductor there exists a corre- 

 sponding ratio between the diameters of the conductor and insulator 

 which will give the maximum efficiency at a minimum cost ; and 

 practically the thickness of the gutta-percha is almost always in ex- 

 cess of this theoretical thickness ; and if a constant ratio is main- 

 tained lietween the diameters of the conductor and insulator the 

 number of words per minute which can be sent through a given 

 length of core is simply proportional to the quantity of the materials 

 used.' The process of manufacture is described ; namely, the twist- 

 ing of several copper wires into a conducting strand, the filling of its 

 interstices with Chatterton's compound, the coating of it with suc- 

 cessive layers of gutta-percha, the serving of the core with hemp or 

 jute yarn ; the sheathing with iron or steel wires sometimes sepa- 

 rately covered with hemp and laid on spirally but without twist in 

 the individual wires ; and the final basting of the cable with Clark's 

 bituminous compound. The relative merits of gutta-percha and 

 india-rubber as insulators for submarine lines are discussed ; the 

 most important point in favour of india-rubber being its lower 

 specific inductive capacity. A number of actual cables are described, 

 and exhibited by diagrams, and certain novel proposed forms are 

 mentioned. Reference is made to the tendency which the spiral 

 lay in the sheathing has to make the cable, if laid too slack, form 

 bights which are drawn into kinks if the cable is picked up. 

 Tables are given with particulars of the dimensions and weights of 

 the principal submarine cables working at the date of the paper, the 

 depth of water in which they were laid, and the time during which 

 they had been in use. In a discussion following the paper, the 

 author referred to the failure of cables through faults in the insulated 

 covering, or through the destruction of the sheathing : in either case 

 the failure was, in general, complete only when the copper conductor 

 parted. The electrical currents did not initiate faults in the insulator, 

 but where small faults existed previously they were developed by 

 the currents. The faults would be partially sealed for a time by 

 the use of positive currents, but this was done at the expense of 

 the copper, which was gradually eaten away until at last it was 

 severed, and signals then failed. This he believed had been what 



