The Atlantic Telegraph (1865)

by Sir William Howard Russell

Before the prospectus was issued, every attention was paid that the characteristics of the Cable should be suited to its work; that it should not be too dense, lest its weight should render it unmanageable in the sea—nor too light, lest it should be at the mercy of the currents as it went down. It was decided that it should weigh a ton per mile, should be just so much heavier than the water which it displaced in sinking, and of such structure as could be easily coiled and yet be a rigid line, while its centre should be composed of wire capable of conveying electrical symbols through an extent of more than 2000 miles, and should retain complete insulation when immersed in the ocean. It was a subject of close and anxious inquiry how to obtain a Cable of this form and character. No fewer than sixty-two different kinds of rope were tested before one was determined on.

In the Cable finally adopted, the central conducting wire was a strand made up of seven wires of the purest copper, of the gauge known in the trade as No. 22. The strand itself was about the sixteenth of an inch in diameter, and was formed of one straightly drawn wire, with six others twisted round it; this was accomplished by the central wire being dragged from a drum through a hole in a horizontal table, while the table itself revolved rapidly, under the impulse of steam, carrying near its circumference six reels or drums each armed with copper wire. Every drum revolved upon its own horizontal axis, and so delivered its wire as it turned. This twisted form of conducting wire was first adopted for the rope laid across the Gulf of St. Lawrence in 1856, and was employed with a view to the reduction to the lowest possible amount of the chance of continuity being destroyed in the circuit. It seemed improbable in the highest degree that a fracture could be accidentally produced at precisely the same spot in more than one of the wires of this twisted strand. All the seven wires might be broken at different parts of the strand, even some hundreds of times, and yet its capacity for the transmission of the electric current not destroyed, or reduced in any inconvenient degree. The copper used in the formation of these wires was assayed from time to time during the manufacture to insure absolute homogeneity and purity. The strand itself, when subjected to strain, stretched 20 per cent. of its length without giving way, and indeed without having its conducting power much modified or impaired.

The copper strand of the Cable was rolled up on drums as it was completed, and was then taken from the drums to receive a coating of three separate layers of refined gutta percha; these brought its diameter up to about three-eighths of an inch. The coating of gutta percha was made unusually thick, for the sake of diminishing the influence of induction, and in order that the insulation might be rendered as perfect as possible. This latter object was also furthered by the several layers of the insulating material being laid on in succession; so that if there were accidentally any flaw in the one coat, the imperfection was sure to be removed when the next deposit was added. To prove the efficacy of the proceeding, a great number of holes were made near together in the first coating of a fragment of the wire, and the second coat was then applied in the usual way. The insulation of the strand was found to be perfect under these circumstances, and continued so even when the core was subjected to hydraulic pressure, amounting to five tons on the square inch. The gutta percha which was employed for the coating of the conducting strand, was prepared with the utmost possible care. Lumps of the crude substance were first rasped down by a revolving toothed cylinder, placed within a hollow case, the whole piece of apparatus somewhat resembling the agricultural turnip machine in its mode of action. The raspings were then passed between rollers, macerated in hot water, and well churned. They were next washed in cold water, and driven at a boiling-water temperature, by hydraulic power, through wire-gauze sieves, attached to the bottom of wide vertical pipes. The gutta percha came out from the sieves in plastic masses of exceeding purity and fineness, and those masses were then squeezed and kneaded for hours by screws, revolving in hollow cylinders, called masticators; this was done to get the water out, and to render the substance of the gutta percha sound and homogeneous everywhere. At each turn of the screw, the plastic mass protruded itself through an opening left for feeding in the upper part of the masticator, and was then drawn back as the screw rolled on. When the mechanical texture of the refined mass was perfected by masticating and kneading, it was placed in horizontal cylinders, heated by steam, and squeezed through them by screw pistons, driven down by the machinery very slowly, and with resistless force. The gutta percha emerged, under this pressure, through a die, which received the termination of both cylinders, and which at the same time had the strand of copper wire moving along through its centre. The strands were drawn by revolving drums between the cylinders, and through the die. They entered the die naked bright copper wire, and issued from it thick, dull-looking cords, a complete coating of gutta percha having been attached to them as they traversed the die. Six strands were coated together, ranging along side by side at the first covering. Then a series of three lengths of the strand received the second coat together. The third coat was communicated to a solitary strand. The strand and its triple coating of gutta percha were together designated “the core.”

The copper strand was formed and coated with gutta percha in two mile lengths. Each of these lengths, when completed, was immersed in water, and then carefully tested to prove that its continuity and insulation were both perfect. The continuity was ascertained by passing a voltaic current of low power through the strand from a battery of a single pair of plates, and causing it to record a signal after issuing from the wire. A different and very remarkable plan was adopted to determine the amount of insulation. One pole of a voltaic battery, consisting of 500 pairs of plates, was connected with the earth; the other pole was united to a wire which coiled round the needle of a very sensitive horizontal galvanometer, and then ran on into the insulated strand of the core, the end of which was turned up into the air, and left without any conducting communication. If the insulation was perfect, the earth would form one pole of the battery, and the end of the insulated strand the other pole, and the circuit be quite open and uninterrupted; consequently no current would pass, and the needle of the galvanometer would not be deflected in the slightest degree. If on the other hand there was any imperfection, or permeability in the sheath of gutta percha, a portion of the electricity would force its way from the strand through the faulty places and surrounding water to the earth, a current would be set up, and the needle of the galvanometer deflected; the deflection being in proportion to the current which passed, and therefore its degree would become a measure of the amount of imperfection.

When about fifty of the two-mile lengths of core were ready, these were placed in the water of the canal which ran past the gutta percha works, and were joined up by their ends into one continuous strand of 100 miles, the joints being covered with gutta percha. The hundred-mile length was then put through a careful scrutiny in the same way that the smaller portions were tried,—and next it was halved, quartered, and separated into groups of twenty, ten, and finally two miles, and each of these were again separately examined, and tested in comparison with similar lengths previously approved.

Whenever separate lengths of the gutta percha covered core were to be joined together, the gutta percha was scraped away for a short distance from the ends, and these were made to overlap. A piece of copper wire was then attached by firm brazing, an inch or two beyond the joint on one side, tightly bound round until it reached to the same extent on the other side, and then was there firmly brazed on again. A second binding was next rolled over the first in the same fashion, and extended a little way beyond it, and finally several layers of gutta percha were carefully laid over, and all round the joint by the agency of hot irons. If the core on each side of the joint was dragged opposite ways until the joint yielded, the outer investment of the wire unrolled spirally as the ends were pulled asunder, and so the conducting continuity of the strand was maintained, although the mechanical continuity of the strand itself was broken.

The two-mile coils of completed and proved core were wound on large drums with projecting flanges on each side, the rims of which were shod with iron tires, so that they might be rolled about as broad wheels, and made to perform their own locomotive offices as far as possible. When the core was in position on these channelled drums, the circumference of the drum was closed in carefully by a sheet of gutta percha, which thus constituted its core-filled channel a sort of cylindrical box or packing case. In this snug nest each completed coil of core was wheeled and dragged away to be transferred to the manufactory, either at Birkenhead or Greenwich.

The core-filled drums, having arrived at the factory of the Cable, the drums were mounted by axles, and kept ready so that one extremity of the length of core might be attached to the Cable as it was spun out, when the drum previously in use had been exhausted. During the unrolling of the core from the drum, it was wound tightly round by a serving of hemp, saturated with a composition made chiefly of pitch and tar, the winding being effected by revolving bobbins as the core was drawn along. This hempen serving constituted a bed for the external coat of metallic wires, and prevented the insulating sheath of gutta percha from being injured by pressure during the final stage of the construction. Each new length of core was attached to the Cable by precisely the same operation as that used at the gutta percha works in joining the two-mile coils for testing; shortly before an old drum was exhausted, its remainder was rapidly pulled off and placed in the joiner’s hands, so that it might be made continuous with the core on a new drum, before the outgoing Cable began to draw upon it.