ON THE CONSTRUCTION AND MECHANICAL PROPERTIES OF SUBMARINE TELEGRAPH CABLES. By WILLIAM FAIRBAIRN, C.E., LL.D., F.R.S. TWENTY-FOUR years have now elapsed since Professor Wheatstone suggested to the Select Committee of the House of Commons on Railways, the construction of a submarine telegraph between Dover and Calais. Since that time 11,000 miles of cable have been laid, only a little more than one-fourth of which can be said to be in a working conditon; amongst the unsuccessful attempts being the Atlantic cable, measuring 2,200 miles; the Red Sea and India Telegraph, of 3,499 miles, and sundry shorter ones, measuring collectively about 2,300 miles. To account for these misfortunes is a work of some difficulty, owing to the many causes which may affect the integrity of the insulation, or the continuity of the conducting wires. The 8,000 miles of failure have not been, however, wholly lost. They have been the means of accumulating a vast amount of experience, and have suggested remedies for the inevitable difficulties which have to be encountered, now as before, both in the manufacture and in the paying-out of deep-sea cables. There are two descriptions of cables required for marine construction one for shallow water, where, owing to the liability of injury from ships' anchors, or the abrasion against rocks or gravel, it is necessary for the insulated wire to be surrounded with an extra strong covering of wire and hemp saturated with pitch; and the other for deep-sea purposes, in which case, as the cable when once laid is supposed to lie perfectly quiescent at the bottom of the ocean, no more strength nor protection is needed than will shield the wire and its insulating coating from injury during the paying-out. Respecting the shallow-water cables, in which category we class the line between Dover and Cape Grinez, laid in 1851; the line from Dover to Ostend, laid in 1853; the one from England and Hanover, 280 miles long, laid in 1858; one between Folkestone and Boulogne, laid in 1859; and one between England and Denmark, 350 miles long, also laid in 1859, all the above are the property of the Submarine Telegraph Company. In addition to these, there are several others which may come into the same class, such as the lines between England and Holland, and the Channel Islands cable, laid between this country and Alderney, Guernsey, and Jersey, in August, 1858. Amongst the most important of the deep-sea cables is that of the Atlantic Telegraph Company. This company obtained an act of incorporation in 1854, which conferred, amongst other privileges, the exclusive right of landing cables on the coast of Newfoundland, or the adjacent islands, for a term of fifty years. The company also obtained a grant of 14,000l. per annum from the British Government, and a similar one from the American Government, so long as the line was in working order. Upon these guarantees and privileges the company was formed, and # the cable was manufactured, one half by Messrs. Glass and Elliott, of Greenwich, and the other half by Messrs. Newall and Co., of Newcastleon-Tyne. As one article has already been devoted to the history of this ill-fated cable, we will not further allude to it, than to say that the failure of this enterprise may be attributed to the want of care and proper supervision in the manufacture, and, to use the words of the commission, "practical men ought to have known that the cable was defective, and to have been aware of the locality of the defects before it was laid." We might multiply instances of several other similar failures, such as the Red Sea and India, the Spezzia and Corsica, and the Bona and Cagliari cables, all of which are now useless. In deep-sea lines there are three points which require careful consideration, and which appear essential to success, namely the tensile strength and conducting power of the cable, perfect insulation, and machinery calculated to pass the cable with safety from the ship into the sea. If this latter can be properly effected, we may venture to assert that a well-insulated cable, when once laid, may be retained for a series of years in satisfactory working order. In the forthcoming Atlantic telegraph, every possible precaution has been taken to have a sound and suitable cable in the first instance, and Messrs. Glass and Elliott have not only conformed to the recommendations of the scientific committee, but they have chartered the Great Eastern steamship for the exclusive purpose of laying the cable, commencing probably at Newfoundland, and continuing the process of paying-out, as we hope, without break or interruption, till it is safely landed at Valentia. As the construction of the cable is equally important with the skill with which it is laid at the bottom of the Atlantic, it may be interesting to compare the present cable with those previously laid down, and to show with what precaution the directors of the company have undertaken this important and precarious task. In all the cables we have specified, the same general principles prevail, viz. : 1. The central conductor is a copper wire, or strand of wires. 2. The insulating covering is gutta-percha. 3. The external protection, when used, consists of hemp or other fibrous material, impregnated with pitch or some other resinous substance, nearly in all cases covered with iron or steel, more in the form of an ordinary rope. 4. The cables so prepared have been paid-out over the stern of ordinary vessels, with a pressure-break to regulate the delivery according to the speed of the vessel, which has averaged from four to six knots per hour. In all cases copper has been chosen for the conducting wire, its durability and its high conducting power rendering it peculiarly applicable for the purpose. In the first telegraphs, the conductor generally consisted of a No. 16 copper wire. This size gave abundant area, and the resistances, even when in lengths of several miles, were The Atlantic Cable and its Teachings,' QUARTERLY JOURNAL OF SCIENCE, No. 1, p. 44. not found to interfere seriously with the working. The conducting power of copper wire was taken to be directly as the area; there were, however, no precise data for determining à priori the size of wire requisite for any given length of circuit and speed of transmission. The wire was joined by being carefully lapped and soldered at the joint, and wrapped with smaller binding-wire, which was also soldered with silver solder. In spite of the utmost care in the construction of these joints, some were always imperfect, owing to their liability to fracture, and a break at any single joint destroyed the value of the whole cable. Moreover, the defects in the copper, owing to want of homogeneity, and the presence of foreign matter, frequently rendered the wire so weak that it ultimately parted after being covered, breaking the circuit, or stretched out and reduced the diameter to an inconvenient extent. It was also found that, if the covered wire was excessively stretched, and then allowed to contract, the copper wire, being incapable of regaining its original dimensions, knuckled through the elastic coating. To remedy these defects, instead of a single copper wire bundles of smaller ones, of similar area, were adopted, the joints being so distributed that the fracture, or defect, of a single wire, does not destroy the whole cable. One serious objection to this form of conductor is that, if a single wire breaks, the sharp end is liable to penetrate through the gutta-percha, and establish a communication with the outer conductor. Such a defect is not easily detected, and it can only be guarded against by close examination of the strand itself, and by the constant testing of the coating during the manufacture. In the form of a strand the bulk of the conductor is also greater, and more gutta-percha will therefore be required to cover it. It will, moreover, not be perfectly solid, but will allow water, if it happen to penetrate to any part of the wire, to pass along as in a tube. This latter objection the Gutta-percha Company propose to remove by coating the central wire of the strand with Chatterton's Compound, and then bedding the six centre wires in it in the process of twisting. The compound squeezed out between the wires unites firmly with the insulating material, and the whole becomes so solid that a few inches of this cable will prevent the percolation of water at a pressure of 600 pounds per square inch. Mr. Daft proposes to obtain the same object by bedding copper wires coated with brass in vulcanized india-rubber. Mr. Clark obtains solidity by making the conductor in the shape of a solid wire, divided into three or four sections longitudinally, fitting closely to each other. Mr. Newall unites the several wires of a strand with solder. Dr. Matthiesson, Professor Thompson, and other experimentalists, have shown that the quality of the copper exercises a material influence on the conducting power of the wire, and it is very important that copper, as pure as can be obtained in commerce, should be used. The following table, extracted from the commissioners' report, shows the relative value, or conducting powers, of certain commercial coppers : TABLE, showing the Conducting Power of certain Commercial Coppers. Traces of silver. No suboxide of copper. Traces of iron, silver (03 per cent.), and suboxide of copper. Traces of iron and suboxide of copper. Traces of iron, nickel, antimony, suboxide of copper, &c. Traces of lead, iron, nickel, suboxide of copper, &c. Traces of lead, iron, nickel, anti mony, suboxide of copper, &c. Traces of lead, suboxide of cop- From the above table, it would appear that the difference of conducting power in the different kinds of copper is caused by the impurities contained in the specimens experimented upon. The Rio Tinto copper, in so far as regards its conducting power, being no better than iron. It has been found that there are no alloys of copper which have a better conducting power than the metal itself; but, as perfectly pure copper is not to be obtained, we have only to reiterate that copper, as pure as can be possibly procured, is the only metal which should be used for the conducting wire of a submarine cable. Insulation.-As copper seems to stand out prominently as the most fitting conductor, so does caoutchouc, or india-rubber, appear almost specially intended for the purpose of insulation. Its qualities, in this respect, are of the highest order. It is tough, highly elastic, of less specific gravity than water, easily manipulated, extremely durable under water, nearly impervious to moisture, except superficially, and not excessively costly; and on its first introduction it appeared as if nothing further could be desired. One of the first and most important requirements in any insulating substance is that it should offer facilities for making the numerous joints required, either in the first construction of the line or for its repair when laid down. For this purpose, also, india-rubber appeared well adapted: if after being cut the fresh surfaces are immediately brought into contact, almost perfect reunion takes place; and if they are warmed and slightly moistened with naphtha (in which india-rubber is soluble), they are hermetically sealed. The covering was effected by first coating the copper wire with cotton and shellac varnish, and then winding a thin strip of masticated india-rubber spirally round the wire, each turn overlapping the last. Several coatings were thus put on, the union of the surfaces being secured by means of naphtha. An almost perfect insulation was the first result, the problem on which so much time and money had been expended seemed to be definitely solved, and the new material came into rapid use. A short time, however, showed the fallacy of these hopes. India-rubber, like all other gum-resins of a similar character, slowly burns or oxidizes in the air, even in darkness; but when exposed openly to the weather and to sunlight this oxidization goes on with alarming rapidity; wires hung out of doors soon become useless; the india-rubber assumed a thick gummy or semi-fluid character, and soon fell away from the wire. The joint, even when made with naphtha, was found not to be durable, and after a short time, even in unexposed situations, the coating was found loose upon the wire. Attempts were made to preserve it by enclosing it in grooved boards, and thus protecting it from the air, but in dry situations this was found to be of but little avail; and although in wet tunnels it was found to add to the durability, it was ultimately obliged to be abandoned there also. Gutta-percha was soon proposed as a remedy for these evils. When pure, and at moderate temperatures, it is a remarkably good insulator, and, moreover, is capable of being kneaded and drawn solidly on the wire through dies, thus avoiding the infinite number of joints required when india-rubber is used. From an analysis by Professor WŴ. A. Miller, it appears that pure gutta-percha is a hydro-carbon, consisting of In commerce, however, it is mixed with resin, vegetable fibre, moisture, &c.; the latter being mechanically diffused through the mass, influencing its pliability and toughness. Commercial gutta percha will remain unchanged for months in the air, provided light be excluded, and the temperature be not very high; and it will remain unaltered for years in water, especially if coated with Stockholm tar, and kept in the dark. It is, however, rapidly destroyed by alternated exposure to a moist and dry atmosphere, especially if the sun's rays have access to it. Professor Miller found that all the deteriorated portions had absorbed oxygen. |