Escher.gif (426 bytes)

History of the Atlantic Cable & Undersea Communications
from the first submarine cable of 1850 to the worldwide fiber optic network

1898 - Submarine Telegraph Lines
by Ewing Matheson

Introduction: By 1898, forty years after the first Atlantic cable, the submarine cable industry was well established. In that year Ewing Matheson published the third edition of his Aid Book to Engineering Enterprise, a 900-page guide to all kinds of engineering works.

In the book’s preface, Matheson writes: “... the author endeavours to classify and link together the technical and economic conditions which determine the success or failure of Public Works and Engineering ventures, and in doing this, he enumerates those preliminary particulars on which designs and estimates of cost can alone be framed”.

While Matheson’s article on submarine telegraphy contains much that is generally known, its breakdown of the costs associated with laying cables at the end of the nineteenth century is a valuable resource. For reference, multiply the costs by a factor of about 80 to find the equivalent value a century later.

-- Bill Burns

Submarine telegraph lines are pre-eminent amongst the modern wonders of the world. The first submerged cable in salt-water was successfully laid from Dover to Calais in 1851, and after three attempts, all more or less failures, in 1857, 1858, and 1865, to cross the Atlantic, submarine telegraph communication between Ireland and America was achieved in 1866, and has since been maintained. When telegraph lines are laid under water, conditions very different from those pertaining to land lines have to be satisfied. The mere insulation and protection from the water could be effected without the elaborate cable-structure which is generally adopted. Thus the submarine line which was laid successfully in 1855, during the Crimean War, between Varna and Balaclava, a distance of 300 miles, consisted of a single copper wire, No. 16 S.W.G. (= 1.62 mm.), covered only with gutta-percha, the outside diameter being about 5/16 in. This cable, though not made in the perfect manner which has since been attained, worked successfully for about nine months.

But other exigencies have to be met. Just as the wire of a land line has to be made strong enough to withstand the heavy strains of suspension and of the wind, so has the submarine cable to satisfy conditions also entirely outside those of the electrician. A submarine cable must be made strong enough to sustain its own weight and the other strains to which it is subject while it is being laid or picked up; and it is obvious that the greater the depth of water the greater will be the length, and therefore the weight of the bight. But if the tensile strength be increased by giving more or thicker sheathing-wires, the weight is again increased, and so the two opposing circumstances grow together. The effective weight of a cable is of course less when submerged than when above water, and cables can be constructed with few or light sheathing-wires, which weigh very little when submerged. The real danger lies in the strains from the vessel from which the cable is being laid or picked up; and even when great care is exercised, it is sometimes impossible, in a rough sea or in a gale of wind, to prevent considerable tension on the cable as it is passing over the stern or when being picked up. Indeed, unless sufficient precaution be taken, the vessel is in effect sometimes anchored to the cable, with consequences that may be anticipated.

Submerged cables are subjected also to the risk of damage from ships’ anchors, and this risk is of course greatest in shallow waters and much frequented seas; therefore cables which are laid in such situations have to be made much stronger than those for deep ocean service. Thus, while the weight of a deep-sea cable will range from 1½ to 3 tons per mile, a weight of from 4 to 6 tons per mile is considered necessary for such situations as the English Channel: the greater weight being reached when the cable contains more than one conductor. As the cable approaches the shore the danger from anchors is increased. Moreover, the cable is then exposed to the surf, and is liable to be chafed on rocks or shingle; therefore the shore ends of cables are made exceptionally strong, a weight even of from 15 to 20 tons per mile being occasionally attained. In waters where the boring shells (teredo navalis) or worms are exceptionally destructive, sheathing-wires are insufficient, and the cable must be enveloped in copper strips if it is to be long preserved.

So many cables have been laid, and so much experience has been gained under what would appear to be every possible combination of circumstances, that all the incidents of manufacture, laying, picking-up, and repairing, are known to those concerned in such enterprises. Consequently, instead of the very great uncertainty which at first existed, telegraph contractors will undertake the laying of a cable, confident that, whatever be the accidents of the operation and the cost, they will finally succeed. But the amount of money invested is so great, and the interests at stake are so momentous, that too much care cannot be exercised in the design and manufacture of the cable in the first instance. Not only must the cable be capable of transmitting messages, but its electric condition must be such as to do so with the greatest speed, so that by taking the maximum number of words it can earn as much as possible. Every material used in its construction has to be carefully selected with regard to the necessary qualities; and the continuity and insulation of the conductor have to be tested through every stage of its manufacture.

Each conductor is generally composed of several copper wires twisted together into one strand, instead of one solid wire as in a land line, as by this means is neutralised the effect of fracture in any one wire, which, though of comparatively small consequence in a land line, which can be easily repaired, would be serious in a submarine cable, necessitating in the case of deep-sea cables a costly repairing expedition. In regard to insulation, the manufacture of submarine cables had by 1889 been brought to such perfection, that no immediate improvements were looked for. But the more recent cables of the Atlantic, Indian, Australian, and other lines have in this respect surpassed all that was expected of them.

Cables differ in their size and exact arrangement, but usually each conductor is enveloped in gutta-percha. By the accumulation of very varied experience with submerged cables, great improvements have been made in the details of construction. Not only have the conductors to be electrically insulated, but the materials for this purpose have to. be separated by other coatings to prevent certain deleterious chemical action; the inroads of the teredo worm in inland seas have to be prevented (this risk does not arise in deep-sea cables) by a wrapping of brass tape, and then over a coating of tan ed yarn the cable is sheathed with steel wires. In order to prevent corrosion, which rapidly destroyed some of the earlier cables, each of the outer wires is coated, and the complete cable finally enveloped in tarred yarn and a bituminous compound. Long distance ocean cables generally contain one conductor, but cables with as many as seven conductors are used in some cases.

The longer the cable, the greater is the resistance it offers to the electric current, and the ordinary ink-recording or needle instruments in use on land-lines are not sensitive enough for signalling through long cables. For the Atlantic cables two kinds of instruments are in use, viz., the “mirror” instrument and the syphon ink-recording instrument. In the former, a small mirror, suspended by a cocoon fibre, is deflected by the current, and the rays of a lamp placed opposite to the mirror are reflected by it on to a scale, the combination of the movements of the bright spot on the scale denoting the letters. The syphon-writer is a very delicate instrument, in which a fine glass tube, acting like a pen, traces the signal on a moving paper strip or riband. By these means the effect of the current is utilised to the utmost, and the necessary battery-power kept as low as possible. Although the time actually occupied in transmitting each separate signal is, even in the longest cable, very short, still the signals become less defined and separable as the length increases. To meet this, a greater interval must be left between each signal, and therefore in long cables the number of words per minute is much less than with short lines. The improvements since 1870 have been so great that while at that date 50 words per minute could be transmitted from London to Dublin, 300 words or more could be sent in 1887. On long ocean cables the number is very much less. These circumstances must obviously be taken into consideration when choosing the number of conductors and kind of cable to transmit expected traffic. But although it may prove cheaper to lay one cable with two or more conductor-wires than two cables each with one or few wires, it must also be remembered that a duplicate cable tends largely to ensure a regular revenue, which an accident to a single cable would stop entirely. For long ocean lines more than one conductor renders the cable too heavy.

For laying across rivers, bays or estuaries, not much special preparation is necessary. In shallow waters a cable can be best laid by men wading and carrying it from hand to hand: where this is impossible, vessels or large boats can be employed for the purpose. In inland seas a sailing-vessel towed by a steamer can be temporarily equipped as a telegraph-ship, but there is always more risk in such cases than where a proper steamer is used. A towed vessel does not steer well, especially in rough weather, and cannot be stopped quickly in case of fault. Cables have, however, been successfully laid for considerable distances by such means when no proper telegraph-steamer was available. In the case, however, of an ordinary iron steamer used temporarily for such service, the great cost of altering it to make tanks for the cable and for other necessary arrangements is a reason in favour of employing a wooden vessel, which can be altered more cheaply.

But although there are contractors and companies who, for a stipulated price, will undertake to make a suitable cable, and to lay it safely, taking all the risk of accident and failure, the price charged for such a service is based upon the very heavy expenses incident to it, and also upon the probabilities of loss as estimated by previous experience. The machinery and equipments of vessels, specially constructed for laying and picking-up deep-sea cables, have been brought to great perfection; and as a result of dearly-bought experience, such vessels and machinery can be employed with a certainty of ultimate success. But although a considerable number of engineers have now acquired experience in the laying of cables, the number of properly-equipped vessels is small, the cost of maintaining them very high, and when in actual work the salaries, wages, and insurance special to the enterprise add enormously to the expenses—never very small—which are always incident to the voyages of steamers. And when a cable has to be laid in a distant sea, the time occupied and the distance from supplies add greatly to the cost of an undertaking, which, as has been shown, is in any case an expensive one. The contractors for such works naturally require in return a high rate of profit. The expenses of a fully-equipped telegraph-steamship, including coal, stores, wages, salaries, insurance, and interest on capital, will range from £5,000 to £8,000 per month, according to the size of the vessel and the exact nature of the service. And to this expense must be added not only the profit of the contractors on the capital embarked in the construction and maintenance of such a vessel, but, if they have undertaken, as is generally the case, the responsibility of laying the cable successfully, an additional sum, which may be considered as a premium of assurance against loss.

Apart from the liability to rust, which, as previously described, can be prevented, cables will last for 25 years. Experience like this greatly encourages telegraph enterprise, as it proves that a cable properly made and laid has life long enough to amply repay its cost.

The cost of telegraph-cables, of course, varies according to their size and kind, but it is also subject to fluctuations, not usual in most other manufactures. Although the cost of all articles varies with the value of the materials of which they are composed, it is not often that so many circumstances have to be taken into account as with submarine cables. Copper, hemp, gutta-percha, and iron are the materials used for cables, and the current prices for each of these are determined by causes differing in each case. It may be said, generally, that the cost of cables has diminished, partly because of the skill and experience that have been acquired in all the processes of manufacture, partly because of the fall in the price of copper, and partly because of the greater competition which now exists. The cost of submarine cables (1896), exclusive of the cost of laying, ranged from £100 to £500 per mile according to weight, as given below.

A submarine cable of the most recent kind, with one conductor composed of seven No. 22 gauge copper wires, twisted into one strand, insulated and protected as just described, the cable having an outside diameter of ¾ in., and weighing about 3 tons per nautical mile, would cost from £100 to £120 per mile, according to the prices current at the time for the various materials used. The shore ends and intermediate portions would have conductors and coatings similar to the deep-sea portion, but with sheathing-wires adapted to the nature of the sea-bed, exposure to anchors, and other circumstances; the weight per mile being increased to between 6 tons and 18 tons according to the diameter. The extra price per mile (in addition to that for the deep-sea portion) would be from £25 to £40 per ton of extra weight.

A submarine cable with three conductors, each made and insulated in its own coatings, as in the cable just described, would have an outside diameter of from 1¼ to 1½ in., and a weight of from 7 to 10 tons per mile, depending on the sheathing-wires, the strength of which would be determined by the risks to which the cable was to be exposed. The deep-sea portion, which would be used for ocean routes and unexposed seas, would cost from £200 to £300 per mile, while the shore ends and intermediate portion would be heavier and more costly, as described for the single-conductor cable. The average cost of cables as laid throughout the world may be taken at £300 per mile; but losses and initiatory expenses have increased the amount sunk, so that the capital actually spent by the cable companies equals about £350 per mile of cable.

For any but a short and easy line, a submarine telegraph is a very speculative undertaking, and one in which, to justify the investment of capital, the compensating revenue must be tolerably certain. Different investors prefer different kinds of ventures, and hence that of making and laying a cable is kept separate from that of the ownership of the cable when laid. The telegraph company which employs the constructors to lay as well as make a cable, have obviously to pay for their services and risks, although they have the satisfaction and protection of knowing beforehand how much they will have to pay. But even where a submarine telegraph has been successfully laid, it is still in an entirely different position from a land line. A land telegraph, if well constructed, need cost but little for maintenance, and the proprietors may be contented with moderate traffic at the commencement if there is a probability of increase; but a submarine line should pay handsomely from the first if it is to satisfy the expectations which alone will justify the investment of capital. Although experience has removed many of the difficulties and risks of repairing and maintaining cables, yet, in order to put such an undertaking on a stable basis, a duplicate line should be laid. It is evident, if the above considerations be true, that a large proportion of the annual revenue—especially when there is only one cable—must be reserved, either for maintenance or as a sinking-fund, before the revenue can justly be applied to dividend. To reduce as far as possible the risks peculiar to such enterprises, special “Trusts” have been formed, by which the capital of investors is divided over several different cable undertakings.

Submarine cables of moderate length for connecting two parts of the same country, may naturally form part of one general national telegraph system, of which land lines may form the other and perhaps greater part. Where, however, the Government or other owners of the land lines are unable or unwilling to lay a submarine cable, the undertaking affords a legitimate opportunity for private enterprise. In the case of international cables, there are but few instances of the Government of a country doing more than afford permission or encouragement to private capitalists. Assistance or encouragement may be given in any of the various ways previously referred to, or privileges in regard to the connection with existing or contemplated land lines may be granted. In some cases an absolute monopoly has been conceded, which excludes all other cables from being landed; but such a concession, as being against public policy, is rare. The co-operation of numerous capitalists, through the medium of joint-stock companies, is the almost invariable method of conducting these enterprises, which are too speculative to justify large individual outlay.

The following information is that which is necessary to enable a telegraph engineer to decide on the most suitable kind of cable; to allow contractors to estimate the cost, and capitalists to calculate how far the enterprise will be remunerative

1. A Map, or chart, showing the places between which the cable is to be laid, and showing any existing cables which must be avoided.

2. The Depth of water will probably be approximately known from the soundings marked on existing charts; but it is necessary to take special soundings, not only to ascertain the depth of water, but to discover the nature of the sea-bed. Abrupt alterations in depth are to be avoided, as also is a bed of rocks, or boulders. The regular tracks of passing vessels which may approach or intersect the line of the cable, should be indicated, and any other particulars furnished by which the liability to damage from ships’ anchors may be estimated. From the information thus obtained, the route which offers the most favourable conditions can be chosen. Such conditions are a level or undulating bed, composed of mud, ooze, or sand.

3. Places favourable for Landing should be surveyed and described. A sheltered bay, a cove, or estuary, with a sandy beach, and out of the way of anchoring vessels, is the sort of place to be chosen. It is desirable to choose such a place, even if not in the direct line, for if an increased length of cable or land line is rendered necessary, the expense will be amply repaid by the safety afforded to the cable.

4. Information should be furnished as to the Seasons of the year most favourable for laying the cable, and as to the currents, tides, and winds which usually prevail.

5. The Land lines with which communication must be made should be indicated on the map, and the terms on which the transmission of cable-messages and exchange of traffic over such lines will be allowed.

6. The amount of Traffic which is to be provided for.

Text courtesy of the Library of the University of Wisconsin, via Google Book Search

Aid Book to Engineering Enterprise
by Ewing Matheson, M. Inst. C.E.
Third Edition, Revised
London: E. & F.N. Spon, Ltd, 125 Strand.
New York: Spon & Chamberlain, 12 Cortlandt Street.


Last revised: 18 September, 2013

Return to Atlantic Cable main page

Search all pages on the Atlantic Cable site:

Research Material Needed

The Atlantic Cable website is non-commercial, and its mission is to make available on line as much information as possible.

You can help - if you have cable material, old or new, please contact me. Cable samples, instruments, documents, brochures, souvenir books, photographs, family stories, all are valuable to researchers and historians.

If you have any cable-related items that you could photograph, copy, scan, loan, or sell, please email me: [email protected]

—Bill Burns, publisher and webmaster: