July 7, 1910] 



NATURE 



MODERN SUBMARINE TELEGRAPHY.' 

 'PHIS lecture relates to modern submarine telegraphy, 

 and, therefore, I shall omit the historical part of the 

 subject and start with the cable itself, as we deal with 

 it now. The signals to form the messages are sent over 

 the submarine cable as electric currents. The cable 

 consists of a central copper wire ; this is the conductor for 

 the current, and to prevent the electricity escaping from 

 the wire it is insulated along its entire length by gutta- 

 percha. 



Gutta-percha is chosen for submarine work because of 

 its very high insulating properties and its not being acted 

 on, or suffering chemical change, under water. The gutta- 

 percha-covered wire is called the core ; this core, before it 

 can be laid at the bottom of the sea, must be surrounded 

 by jute serving and steel wires for protection when being 

 laid and during its existence after. 



When dealing with the electrical properties of a cable, 

 the core only is considered, and for all practical purposes 

 it may be taken that the return conductor to the current 

 is the water immediately outside the gutta-percha. A core 

 of any given length has a certain time rate of signalling ; 



I take an Atlantic cable laid in 1894 (Fig. i) as having 

 the greatest size of copper for size of core ; I take this 

 core to illustrate the improvement that might result by 

 increasing the copper up to the largest size electrically 

 permissible : — - 



1894 Cable. 



Diameter of core 0466 inch 



Diameter of copper 0202 inch 



Resistance per nautical mile 1-684 ohms 



Capacity per nautical mile 0420 microfarad 



The cable is 1852 nautical miles long and its K.R. is 

 2-41, and its speed of working under the capacity block 

 system of duplex, about 205 letters per minute. 



CFiG. 2.) 



... 0-466 inch 



0-2S2 inch 



... 0-864 ohm 



0-700 microfarad 

 ... 2-06 



Atlantic 1894 Cable. 



that is to say, when a voltage is applied at one end, the 

 effective current, that as a consequence flows in the wire, 

 does not arrive at the distant end instantaneously, but 

 takes time to grow. 



The time rate of signalling is inversely proportional to 

 the product of the resistance of the wire and the electro- 

 static capacity of the core. This is termed the " K.R." 

 or capacity resistance law, a law first pointed out by Lord 

 Kelvin. It follows from this law that if you double the 

 length of any given kind of cable you reduce its speed for 

 signalling to one-quarter. 



The time rate is inversely proportional to the resistance 

 multiplied by the capacity. If you make a certain sized 

 core (size of gutta-percha) with a large copper, up" to a 

 certain point you decrease the resistance and increase the 

 capacity; but there is a critical value giving the minimum 

 K.R. This critical limit, or the point when the size of 

 the copper is reached to give the lowest K.R., is when the 

 diameter of the copper is to the diameter of the core as 

 I : 1-65. 



There is another advantage in keeping the resistance 

 low for any K.R. ; the time constant ortly deter- 

 mines the time when the current at the far end 

 reaches a certain percentage of the possible maxi- 

 mum after the application of the voltage at thr- 

 sending end. Of course, the quantity of current 

 after any given time is determined again by the 

 voltage of the sending' battery, and is inversely 

 as the resistance of the cable. 



For instance, if two cables were constructed of equal 

 K.R., but one had a larger copper of half the resistance 

 of the other, with equal sending batteries, the one with 

 the lower resistance would deliver twice the current at the 

 receiving end, at the ends of equal times, and could there- 

 fore be made to work at a faster rate. It should also be 

 a cheaper cable, because copper is less expensive than 

 gutta-percha. 



-Against these electrical advantages should be placed 

 several mechanical disadvantages ; the reduction of the 

 thickness of the insulation might result in a greater liability 

 to faults developing after the cable was laid. With such 

 a heavy wire, which would naturally have to be well 

 stranded,^ to reduce the stiffness, the liability of the 

 decentralisation during manufacture would be greater than 

 with existing cores. 



These mechanical dilTficulties could, I feel sure, be over- 

 conne, say, by greater care being taken in the manufacture 

 or by substitution for the present yielding gutta-percha of 

 dry cotton or similar ' material well impregnated with 

 gutta-percha compound. 



1 Discourse delivered at the Ro>-al Institution by Mr. Sidney G- Ilrovvn. 



NO. 2123, VOL. 84] 



The Ideal Core. 



Diameter of core 



Diameter of copper 



Resistance per nautical mile ... 



Capacity 



K.R. for 1852 nautical miles 



The speed of working with the same duplex 

 system is about 240 letters per minute, and the 

 current received with this speed would be twice 

 as strong as in the actual cable, so that a stil' 

 greater speed than that given would result, 

 perhaps a speed of 260 letters per minute, a 

 -.ending battery of 40 volts to be used on both 

 cables. 



The copper conductor offers resistance to the electric 

 currents that flow along it ; this resistance by itself would, 

 with sufficiently sensitive receiving instruments, not affect 

 the speed of signalling ; it produces what is termed 

 " attenuation," or a weakening of the signalling current. 



There is also a lateral storage of electricity along the 

 outside of the copper due to the capacity of the insulating 

 material to absorb a charge of electricity ; this property is 

 termed the electrostatic capacity of the core. 



To allow this to be more fully understood, I shall take 

 mechanical analogies. Resistance in electricity is equi- 

 valent to friction in mechanics, capacity to elasticity of a 

 spring, and self-induction to inertia. If I force water 

 through an iron pipe, the friction in the pipe offers resist- 

 ance to the flow of water ; the same quantity that is forced 

 in flows out at the receiving end, but the energy accom- 

 panying the flow of water suffers attenuation, as part is 

 wasted in overcoming the friclional resistance. 



Suppose that, instead of taking an iron pipe, I take a 

 soft india-rubber pipe, a new kind of phenomenon will be 

 noticed. .As I force the water in, the resistance that the 



water encounters in flowing along the pipe causes the 

 rubber to swell, and the rubber will continue to swell until 

 it has acquired sufficient strain to press with sufficient force 

 on the water to overcome the friction of the pipe. 



.At the sending end, that is, the end where we are 

 forcing in the water, the pipe will swell the most, because 

 the pressure on the water is there the greatest and the 

 frictional resistance offered by the pipe to its flow also the 

 greatest. .As we move along, the swelling will be less, 

 being least at the far end, that is, at the receiving end 

 where the water escapes. 



.At the instant that we start forcing the water in, prac- 

 tically none escapes at the receiving end, the pipe com- 

 mences to stretch and the water begins to flow out, 

 continuously increasing in quantity, until it obtains a steady 

 value ; this steady value is reached when the pipe has 

 ceased to expand. 



The time taken for the pipe to expand and for the water 

 to reach a steady value is termed the variable period. 

 The less the elasticity of the pipe and the less the resist- 

 ance to water flowing through it, the less the time taken 

 to reach the steady value. This is equivalent to our sub- 



