TKANSACTIONS OF SECTION G. 997 



cables have been picked up and repaired in up-vsards of 2,000 fathoms within the 

 course of two or three days. 



The maximum speed of working obtainable on the long- section (say 

 3,500 N.M.) -with a larg-e core would be low as compared with that attained 

 on the Atlantic cables. Nevertheless, with a core of the same proportions as that 

 adopted in the ' Anglo-American ' Company's last cable — GSOlbs. copper to 4001bs, 

 gutta-percha— under ordinary conditions a speed (in five letter words) could be 

 attained up to the minimum required by the Canadian Government in 1894. 



This would be by ordinary manual transmission, whereas all the latest 

 improvements in machine transmission with curbing arrangements (as well as 

 condensers) would naturally be applied with something like a 35 per cent, increase 

 in the speed. The adoption of JNIuirhead's Duplex system would nearly double 

 the working capacity of the cable. 



As a first ^ enture the above cable would probably be sufTicient to meet the 

 ends in view. It is conceivable that a larger core would be out of the question on 

 the ground of cost. Moreover, increases iu the dimensions made beyond this 

 wouhl have tlie effect of still further increasing the mechanical difficulties as 

 regards a suitable type of sheathing. 



Experiments have, it is believed, already been made on two Atlantic cables 

 looped together to test the possibility of working through a considerably greater 

 length than that to be dealt with here, with satisfactory results. 



In the construction of the first cable to India {vid the Persian Gulf) Messrs. 

 r»right & Clark, as engineers to the Government, adopted a conductor in which 

 four ordinary wires were drawn down so as to form the four quadrants of a true 

 circle with a ring outside to embrace the whole. 



Ey this means — with a given weight of copper — they reduced the electro- 

 static (inductive) capacity of the core without increasing the conductor resistance, 

 the result being a very material increase in the attainable speed of siffnalling, 

 whilst sufficiently maintaining the mechanical requirements as regards pliability, 

 &c. 



Messrs. Siemens Brothers have since, to a great extent, approached a similar 

 end in their compromise between the single solid wire and the ordinary strand 

 conductor, and this has proved a great success on a number of Atlantic cables. 



It would seem as though a plan of drawing the wires down to a smaller total 

 iirea might in this undertaking be advantageously adopted. In the case of the 

 Persian G ulf cable, an economical method of securing any particular speed was 

 not of the same vital importance. Moreover, in that instance the length not being 

 of an abnornuil character, the extra cost of the above type of conductor was 

 probably scarcely made good by increased working value. Here, however, the 

 -author is of opinion that this plan would be found to more than repay the increase 

 in initial cost — especially as it would also materially reduce the quantity of gutta- 

 percha reijuired to obtain the same miniuuim thickness throughout for fulfilling 

 mechanical as well as electrical qualifications, this being the hcte vioire in the cost 

 <if an ocean cable, as also in its design. 



By way of improving the present means of signalling upon cables, Mr. Oliver 

 Ileayiside, F.K.S., has advocated the introduction of leak' circuits and self-induc- 

 tion into cable lines; and electrical engineers— including Mr. Preece, Dr. Thomp- 

 son, and others— have devised new forms for the insulated conductor accompanied 

 bv devices which to a certain extent realised these theoretical advantages. The ap- 

 plication of leaks on the ordinary cable, when suitably disposed along the line, would 

 obviate the choking effect of retardation, and secure increased definition for the 

 signals, though not sufficiently to permit of any substantial increase in the working 

 speed, even with the battery power raised within reason. 



A number of other proposals have been made, but, so far as concerns any 

 further material increase in the speed attainable for submarine telegraphy under 

 given conditions, it seems probable that if this is to be effected it will be by an 

 entire revolution in the form of conductor, dielectric, and completed cable, rather 

 than in the signalling apparatus. The latter has probably reached its limits of 



