CO885-(6-7) — Page 315

71

320

PUBLIC RECORD OFFICE

Reference :-

TITLTIC.O. 885

6

PUBLIC RECORD OFFICE, LONDON

ALLY WITHOUT PERMISSION OF THE BE REPRODUCED PHOTOGRAPHIC-

COPYRIGHT PHOTOGRAPH-NOT TO

FIFTH DAY.

MONDAY, 23RD NOVEMBER 1896.

PRESENT:

THE RIGHT HONOURABLE THE EARL OF SELBORNE, CHAIRMAN, Presiding.

The Hon. Sir DONALD SMITH, G,C.M.G. The Hon. Sir SAUL SAMUEL, K, C.M.G,, (.B. The Hon. D. GILLIES.

The Hon. A. G. JONES, P.C Mr. G. H. MURRAY, C.B.

Mr. W. II. MERCER, Secretary.

Dr. ALEXANDER MUIRHEAD, called in; and Examined.

Chairman.

939. I THINK you are an inventor of electrical apparatus?—Yes.

940. And you are the inventor of the duplex method?—Yes, jointly with my late brother, John Muirhead, and Mr. Herbert Taylor.

941. Would you kindly describe to the Com mittee what the duplex method is? What the purpose of it is?

942. Yes. The purpose of it is to enable transmission to take place from the two opposite ends of a cable simultaneously. The best result obtained from our system is about 90 per cent. increase on simplex.

943. An increased speed of 90 per cent.? --On simplex working.

944. On the double? Yes. Is that sufficient, or do you wish me to go into a further expla- nation of how duplex transmission is effected?

945. And the result is you have now the re- putation of getting a greater speed in practice over a submarine cable than anyone else?-1 believe so,

946. Given the same average operator, will he always get the same results out of a given length of cable composed of a certain weight of copper and gutta-percha? -iiven the same average operator, yes.

947. No matter who has manufactured the cable?--The speed depends on the electrical dimensions of the cable.

948. Yes. My point is this--- Any manufacturer will work up to the electrical specification given to him; it is not a question of choosing a manufacturer.

949. No. But you may have a cable that has to span a certain distance? Certainly.

950. And the cable manufacturer may have to make the core of that cable of a certain weight of copper, and a certain weight of gutta-percha? -Certainly.

951. That weight of copper, and that weight of gutta-percha, may be duly put into the cable? -Certainly.

952. But is there nothing at all in the way in which the core is put together Yes. The way in which the strand of copper wires is made up makes some difference in the capacity of the core, and therefore in the speed. By stating in the specification of the cable the conductivity of the copper, and the capacity per knot of the core, also the weight of the copper per knot and of

Chairman-continued.

the gutta-percha per knot, any manufacturer working to the specification would produce the same results as regards speed of transmission.

953. He will?—Yes.

954. With the same results ?--With the same results; yes.

955. There is nothing, in fact, to choose be- tween the manner in which the gutta-percha is ufactured by the different best manufac-

turers-Not to my knowledge. The lower the specific inductive capacity of the gutta- percha used, or the lower the electrostatic capacity per knot of the core specified, the faster will be the cable.

956. You are aware of the project which this Committee is appointed to consider?—Yes.

957. Supposing a cable is laid, and that the route selected is via Fanning Island, you are aware a span from Vancouver to Fanning Island, including 10 per cent. for slack, would be,

roughly, 3,560 nautical miles ?—Yes.

958 Given that span, what speed in letters per minute would be got out of a cable with a core of 500 lbs. of copper and 320 lbs. of gutta- percha per nautical mile ?--In order to make my answer perfectly definite, I must assume that a certain, I may call it, shaped signal, is taken such as that--illustrating) as a standard signal or a safe signal for the clerks to read; any sigual less definite, to be considered as unsafe. This first slip shows the signals that one would get through such a length of cable as you men- tioned at the rate of 43 letters per minute. You will observe that every variation of the current received is very well shown. An experienced clerk does not require to have such clear signals as these all that he requires is a sufficient indication of the number and direction of the battery contacts that were made at the I have met clerks who could signalling end. decipher the signals obtained at the rate of 90 letters per minute through the retardation of In answer to your question this proposed cable. I should say that, according to my experience, 76 letters per minute would be the limit of speed with average clerks.

959. That is with suitable operators, with 500 lbs. of copper and 320 lbs, of gutta-percha for 3,600 nautical miles, we could read 76 letters?- Yes; 76 letters per minute on simplex, or 70

23 November 1896.]

Dr. MUIRHEAD.

Chairman continued. letters per minute each way, say, using the duplex method of transmission,

960. Does that mean that at one and the same time 70 lettera per minute might be going each way?—Yes, 70 letters per minute each way. There is a slight reduction of simplex speed pro- duced by the apparatus introduced to effect duplex working.

961. Now will you kindly give us the same results for a core of 533 lbs. of copper and 355 lbs. of gutta-percha ?—About 10 per cent. more speed than on the 500 and 320 core,

962. And could you give me the same for a core of 552 lbs. of copper and 368 Ibs, of gutta percha ? -Fourteen per cent. more than the first one.

963. Then would it be correct to say that for the first core named you could get an approximate speed of 70 letters a minute each way; for the second core a speed of 77 letters per minute each way; and for the third a speed of 80 letters each way-Yes. That is by using the automatic curb method of transmission. By the ordinary method of transmission the speed would be much lower.

964. These auswers depend on the duplex method of transmission-The duplex method with automatic curb transmission.

965-6. These answers of yours depend on the duplex and automatic curb methods of transmis- sion? Yes. On such a long section as that in question one would advise the adoption of those methods of transmission.

967. They are both being brought into general use now?-Yes. The Muirhead duplex method of working cables has been in use on the Eastern and other lines 21 years now, and it has been established by me on all the Atlantic cables be- tween England, Ireland, and America. My system of automatic curb transmission has only just been adopted by the Eastern Companies and the Brazilian Telegraph Company. The idea of automatic curb transmission on cables is

an old one. It was originated by Lord Kelvin, in conjunction with the late l'rofessor Fleeming Jenkin and the late Cromwell Varley, 20 years ago. All I have done is to simplify the ap- paratus, and make it practicable.

968. But may I take it as practically for granted that any important new cable starting would be pretty certain to adopt that method? Yes, certainly on a long span such as that men- tioned.

Mr. Murray.

969. May I ask whether these are the results of actual experiment?—The slips of signals on that sheet of paper r* are the results of trials in my laboratory on artificial cable of the same retarda- tion as the proposed cable of 500 lbs. copper and 320 lbs. gutta-percha.

970. In your laboratory-Yes. And they are verified by experience on cables of about the same retardation.

971. But how long was the conductor on which you experimented ?-In one instance the results obtained on artificial cable of my construction in the laboratory were verified on a cable 2,350 knots in length with a very much smaller core than that of this proposed Pacific cable, but having the same retardation, nearly.

* See page 244.

Chairman,

[Continued.

972. Sir Saul Sannel would like to know what influence on the duplex system the length of a span or the depth of water has ?-On duplex transmission?

973. Yes. None whatever.

974. Can you tell the Committee the rates of speed now obtained over some of the longest sections of cable in existence-On the two Western Union Telegraph Company's cables from Penzance to Canso (Nova Scotia), which are about 2,500 knots in length, the greatest speed obtained is 120 letters per minute. These cables have a retardation of about seven-twelfths

of that of the proposed 500 and 320 cable (3,600 knuts in length).

975. The result on the section of this cable you have mentioned, the present speed, is 120 letters per minute?—Yes, the greatest speed 1 have obtained is 120 letters per minute.

976. Actual traffic each way? Yes. 977. Can you give the Committee another instance-On one of about 8 per cent. less electrical retardation than this proposed Pacific cable, a speed of 80 letters per minute has been obtained.

978. Which cable is that?-One length 2,350 knots of very small core, 200 lbs. copper and 180 lbs. gutta-percha.

979. Do you think that for practical purposes

a speed of 70 to 80 letters per minute each way is sufficient for an important cable?—It depends upon the amount of traffic to be carried and the capital cost. It would also depend upon the cost of repairs and the renewals. I do not think I can answer that question.

980, I should like to put it in another way. Suppose this span of 3,600 nautical miles, would it be safe to have a cable with heavier core than the one we have mentioned ?Do you mean for yielding a return on the capital?

981. No; I mean from the point of view of laying the cable or raising the cable for repairs. Would it be advisable or even safe to have one of heavier weight than, say, 552 lbs. of copper and 368 lbs. of gutta-percha per nautical mile?—— I have no experience of the laying of cables at all. I should say, seeing that the new Com- mercial cable of 1894 laid by Messrs. Siemens has been successful now for two years, there would be no difficulty with the heavier core you mention.

982. Can you tell the Committee how far the speed obtainable from a given cable depends upon the composition of the cable, and how far upon the operating apparatus ?—A great deal depends upon the apparatus employed, on the operators, and on the mode of transmission adopted. As regards the cable nothing but the purely electrical dimensions regulate the speed. Having decided upon the minimum speed of working of a proposed cable, the best advice I can give is to put as much copper as possible in the conductor of it (within limits, of course).

983. Suppose a cable to be laid to Australia vid Vancouver, Fanning Island, Fiji, and Nor- folk Island, can you give the Committee any idea how long it would approximately take for a message to go from London to Australia ?--It would depend on the speed of the longest section. 984. Fanning Island?-The speed on the longest section would determine that.

14

1

72