51
310
PUBLIC RECORD OFFICE
Reference :-
༴| ། །
C.O. 885
6
PUBLIC RECORD OFFICE, LONDON
ALLY WITHOUT PERMISSION OF THE BE REPRODUCED PHOTOGRAPHIC-
COPYRIGHT PHOTOGRAPH-NOT TO
THIRD DAY.
TUESDAY, 17TH NOVEMBER 1896.
PRESENT:
THE RIGHT HONOURABLE THE EARL OF SELBORNE, CHAIRMAN, Presiding.
The Hon. Sir DONALD SHITH, G.C.M.G. The Hon. Sir SAUL SAMUEL, K.C.M.G., C.B. The Hon. A. G. JONES, P.C.
The Hon. D. GILLIES. Mr. G. H. MURRAY, C.B.
Mr. W. H. MERCER, Secretary.
Mr. ALEXANDER SIEMENS, re-called; and further Examined.
Chairman.
635. MR. SIEMENS, in the composition of the cable which you have specified, there would, I imagine, be some variation in a 12-word, a 15- word, or an 18-word cable?-Yes, but simply with regard to the amount of copper and gutta- percha per mile; otherwise the construction of the cable is practically the same.
636. I understand that in a cable there are three parts?-Yes, the copper conductor, the insulating material, and the mechanical surround- inge-how shall I say it which protect the cable against mechanical injury.
637. And which is the dielectric ?-The die- lectric is the insulating material.
638. And which of those three together consti- tute the core? The core is constituted by the copper conductor and by the dielectric, which is generally gutta-percha.
639. What is the conductor; what is it com- posed of?-Only copper. It is a strand of copper; and I want also to show a variation in the arrangement. It has always been usual to make seven equal copper wires into a strand, but we have found that it is a material advantage to make a big solid wire in the centre, and have it I think surrounded by a number of small ones. I can show the difference. The United States cable has got it already. I am afraid I have not got a seven-strand one, but you can see it here even in this cable (producing specimen of cable). There is a thick wire in the centre, and small wires around. You get more copper in the same space; therefore you can use for the same weight of copper a less weight of gutta-percha.
640. And how do you test the conductivity of the conductor? That is done generally in the usual electrical method by the Wheatstone bridge method generally. You directly measure at the ends of the conductor.
041. And you test that, both per knot per pound, and also per knot per
mile? Well, in practice, of course, as you get the conductor made
up in the strand, you test the strand and you calculate from that the knot pound. That
Chairman-continued.
comes by itself; I mean the one follows from the other.
642. Now, would you describe what the die- lectric exactly consists of?-The dielectric for submarine cables is usually taken to be gutta. percha, and as good gutta-percha and as pure gutta-percha as you can get, nothing mixed with it at all; and I should say that perhaps we differ in several respects from the ordinary routine of cable making. In putting on the gutta-percha we do not use Chatterton's compound. You will find in ordinary specifications that it is specified that the strand of copper wires is to be coated with Chatterton's compound; that then
A
layer of gutta-percha is put on; then another layer of Chatterton's compound, then another layer of gutta-percha, and so on, according to the requirements; but we put on several layers of gutta-perchs, without any compound between, and without any compound on the strand of the wire, and we believe that that is a very material improvement.
643. And how do you test the specific resist- ance of the dielectric ?-There are well-known electrical methods; you do not want me to describe them technically?
644. No; but I understand that you test it per cube knot?-Oh, yes.
645. By the methods that are well known?- Oh, yes.
646. Now, how is the whole core tested when it is made ?-If I may begin by saying the diffi- culties which you experience in manufacturing the core are mostly created either by air bubbles which get in while the hot and plastic gutta- percha is being pressed round the wire, or by small foreign particles which may get into the gutta-percha; I mean, however much you may be careful, a little speck of dirt or anything which falls into the gutta-percha and is not seen or perceived and removed will get into the core, and the so-called faults, that is, defects in in- sulation, are principally caused by these two things, by the air bubbles and by foreign
17 November 1896.]
Chairman-continued.
Mr. SIEMENS,
matters. If you simply test by the or- dinary testing methods, even with 500 volts in the battery, it will not tell you where there is an air bubble, because air is a very good insulator, even a better insulator than gutta-percha, and so naturally if you test for insulation the air bubble will not be detected by your tests. And with the foreign matter, if the foreign substance does not happen to touch the copper conductor and reach right through the gutta-percha to the out- side the ordinary testing will not show you those things either; but when the cable is manufactured and laid and gets into a great depth where there is a heavy pressure of water, then the water will at once burst the air bubbles and thereby perhaps injure the insulation. And also these foreign substances, being hygroscopic mostly, being little Huffs and things, they will set up a small leak, and the curious action of an electric current if it passes through gutta-percha like that is that it eats itself a larger way, and a fault which once commences in ♫ cable is sure to lead sooner or later to a complete breakdown in the insulation at that spot. Therefore it was always the custom to specify that a cable must be maintained for 30 days by the contractor after the completion of the laying. That was just to give the pressure of water time to press the air bubbles and to find out any small leakages which might exist, and that is also, I take it, the sense of the require ment in the specification that the contractor should maintain the cables for three years, and for all cables which are laid or rather tested according to the ordinary methods that is a perfectly correct stipulation and a necessary one. Now it struck our people a long time ago that it was rather a pity to wait until the cable was laid before the air bubbles were burst; therefore we have had a big cylinder constructed by Messrs Whit- worth, and connected it with accumulators and pressure pumps, in which we can put drume having about two miles of core on. They are put into this vessel, and we have, up till now, put a pressure of four tons per square inch on to it, which repre- sents, roughly, 3,000 fathoms. I can very easily explain the connection if you remember that the ordinary pressure of the atmosphere is 15 pounds per square inch, and a column of water of about 32 feet is also balanced by the atmosphere; that is to say, every 32 feet of water gives you a pressure of 15 pounds, or, for matters of calcula- tion, say that two feet of water gives one pound pressure; that would be 15 and 30; that gives you a little margin. Well, in 3,000 fathoms, which is 18,000 feet, you want therefore 9,000 pounds pressure, which is, roughly speaking, four tons (four tons exactly is 8,960), so that if we put a pressure of four tone, or as we did when we were putting a pressure of four tons per square inch on the core, we really expose it to the condition which it is in at the bottom of the sea in 3,000 fathoms depth, and in that way if any air bubbles occur, they have been detected by this method, and then it is an easy matter in that of the manufacture to repair the core
stage and put the place in as good a condition as any other. That gets rid of the one difficulty; but the other difficulty remains, that is, a little foreign matter which does not reach quite
Chairman-continued.
[Continued.
through. Well, we found that we can deal with that by exposing every drum of core for half-an- hour to a pressure of 7,500 volta electric current, and that is done after the core has been pres- sured. The effect of that is that all the places where air bubbles have been, and where they are crushed in so that the thickness of the insulation is smaller than it ought to be, and wherever there is foreign matter which almost invariably con- ducts better than gutta-percha, that all these places are burnt through by this high tension electricity. Again we can repair the core and in that way secure a freedom from electrical faults which you cannot attain under the ordinary modes of test- ing. The effect of that has been, for instance, in that 1894 Atlantic cable, which was the first cable in which we applied the alternate current tests, that the insulation is practically exactly the same as it was two years and three months, or whatever it is, ngo; it has not altered at all. And I wish you would ask some people who know direct about it how the Anglo-American cable of 1894 is, which was made at the same time and presumably with as much care, but under the ordinary methods. I have been told that it has broken down through an
elec- trical fault; but as people, of course, are always willing to tell me bad tales about our competitors would much rather that you heard that from somebody who really knows.
647. What is the name of your cable?—The Commercial Cable.
I
Mr. Murray.
648. The Anglo is the heavier of the two, is it not? The Anglo is the heavier of the two.
they could support.
Chairman.
649. Now what do the outer coverings consist of? The outer coverings we are making now with tanned jute first, then the sheathing wires, and then tarred jute and compound. We have had for a long number of years a series of tests going on at our works. We have taken Russian hemp, and we have taken manilla, and we have taken jute and exposed it to different conditions. We have taken some hanks of each and
few taken a
strands and tested them, at first when We bought them, what weight Then we divided them and buried some in the ground. up We put others in water; we hung others on a roof exposed to the weather. We hung other And then we have from hanks inside the caves. time to time again tested the strength of these single strands. Well, the effect of this testing has been that although other materials may be stronger in the first instance, it is jute alone that keeps its strength under the most varied con- ditions. It would be too much to go into the exact details, but we have found that jute lasts the beat, and after a few years is stronger than hemp and stronger than manilla yarn, although in the first instance it is not so strong; therefore we always now take jute, and for the sheathing wires we now take, for the deep- sea portion, tempered steel, the very best quality
G 2
52