Escher.gif (426 bytes)

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

Atlantic Telegraph Instruments and Working of the Cable
by William Thomson

Introduction: William Thomson, knighted for his work on the Atlantic cable and subsequently created Baron Kelvin of Largs, wrote a 23-page Electric Telegraph article for the 1860 edition of Encyclopedia Britannica, a comprehensive survey of the history of telegraphy to that date. These extracts on the instruments and working of the 1858 Atlantic Telegraph are taken from that article.

--Bill Burns

Electric Telegraph

Encyclopædia Britannica, 1860, Volume XXI, Article “Electric Telegraph

by William Thomson

Section VIII: Telegraphs Using Only Two Degrees of Current

No better Morse telegraphic printing through any line, long or short, can be shown than the messages received at Valentia [1] from Newfoundland, a specimen of which is represented in the annexed fac-simile on a scale of one-fourth lineal dimensions.

[1] The Morse alphabet, as generally used through Europe, bears modifications which have been made chiefly by German telegraphists. It is on the whole better adapted to the German than the English language; and English telegraphers would do well to make several changes on important letters, which could readily be done without confusion. This has been done by the Atlantic operators in the case of the letters M and O, which have been interchanged with much advantage in consequence of the greater frequency of the letter O in the English language. Thus the Atlantic Morse alphabet is as follows:

These messages were recorded in the following manner: The receiving clerk watched the image of a lamp reflected upon a horizontal paper scale, from a mirror galvanometer in circuit between the cable and the earth. Every time he saw the image begin to move towards the right, he pressed a Morse key placed beside him, and every time he saw it begin to come back towards the left, he lifted his hand and released the key. This action made and broke the circuit for a local battery and Bain's recorder, and thus produced marks and blanks consecutively on the riband of paper, of lengths proportional to the times during which the spot of light was moving to the right and to the left respectively. The receiving clerk, in fact, acted the part of relay, with what perfection the Valentia telegrams show, as the reader may judge from the specimen of them we present.

Service No. 222. Valencia, Aug 21, 1858


This system of receiving and recording messages by human relay has great advantages over every other method hitherto tried or proposed for working through long submarine lines. In the first place, the motions directly produced by the signal currents are not, as when a mechanical [1] relayrecords them, limited by stops. They may therefore be observed through a wide range on either side of zero, and the signals may be clearly disentangled from influences of inductive embarrassment and of earth currents, which would lay the mechanical relay over to one side or the other, and either prevent it from turning and giving any indication at all of the signals intended, or alter the proportions of blanks and marks so as to render it impossible to decipher the result.

[1] That is to say, any inanimate electro-magnetic or galvanometric relay.

The human relay, employed in the manner which has been described, to record the motions of a spot of light on a scale, is much more sensitive and more reliable than any mechanical relay can be in the circumstances. A degree of current which cannot turn the most delicate of inanimate relays gives an ample motion to the spot of light. Even when the common relay turns, its contact often fails, or is uncertain, unless it is driven by superabundant power—more than it would be advisable ever, to use through an ocean telegraph. The human relay has no corresponding liability to failure. It dispenses with the local battery and chemicals required for working Morse or Bain recorders in the ordinary way; and the simple pressure of the hand, direct on the lever of a Morse instrument, or on a proper appliance for carrying a pencil or other convenient style, allowing it when pressed to bear properly on a riband of paper running through regulated wheel-work, gives a most accurate and sure record of Morse signalling through a submarine line at any rate up to ten or twelve words a minute.

If, however, the inductive embarrassment is so slight as to allow higher speeds than this to be regularly used, it is probable that the eye, mind, and hand of the human relay might not readily act in combination with the requisite rapidity, and that a common mechanical relay would be necessary or preferable.

The mirror galvanometer, planned by the writer of this article for receiving signals through the Atlantic cable (and available, with advantage in point [2] of speed and the smallness of battery power which it requires, for every submarine telegraph in which, through inductive embarrassment, the ordinary relay method becomes limited to any speed less than ten or twelve words a minute), consists of a very light mirror, with a small bar magnet of file steel cemented to its back, supported, within a helix of insulated wire, by silk fibre allowing it freedom to turn about an axis in its own plane and perpendicular to the magnet attached to it. The apparatus is complete with a lamp and scale attached to a frame bearing also the galvanometer; and a lens bounding one side of the chamber in which the mirror is supported, of such a focal length that rays from any point of the flame shall, after entering through the lens, being reflected by the mirror, and repassing through the lens outwards, be brought to a focus on the scale.

[2] If, for instance, the ordinary relay cannot give more than six words a minute, messages may be perfectly recorded by a mirror galvanometer at the rate of ten words a minute. When there is still more inductive embarrassment, limiting the mechanical relay to four words or less per minute, a double speed may with ease be obtained by the new method.

In instruments of this kind, designed for land-use, the mirror is generally hung by a single silk fibre, about 1/8th of an inch long, in the hollowed end of a plug of metal fitting half-way into the cylindrical hollow of the coil, of which the diameter is rather less than half-an-inch. The diameter of the mirror is about 2/5 ths of an inch, and the length of the magnet the same; the weight of mirror and magnet together being from a grain to a grain and a half. Such an instrument, with 500 yards of No. 40 copper wire (weighing 0.5 grain per foot) in its coil, and with only the earth's magnetic force to direct the needle to its position of equilibrium, will at Glasgow give a deflection moving the spot of light over about 200 divisions of  1/40 i each, on a scale 24 i from the mirror, when under the influence of an electromotive force equal to 1/100 of that of a single cell of Daniell's maintained between the two ends of the coil.

The conductor of the Atlantic cable, which weighed from 110 to 120 grains per foot, was about 220 times as massive as that of the galvanometer coil; and therefore, if of equally good copper, 2,400 statute miles of it would have resisted 38 times as much as the 500 yards of the galvanometer. Hence, 0.38 of the electromotive force of a single cell would have given 200 divisions of deflection through the whole cable, if perfectly insulated and of good copper. But two or three divisions of deflection are quite sufficient for reading signals from, and therefore one cell of Daniell’s would have given more than ample signal-currents through the actual copper of the Atlantic cable, if ordinarily well insulated.

The quickness of the galvanometer indications under the directive force we are now supposing - that of the horizontal component of terrestrial magnetism at Glasgow - is nearly sufficient for signalling at the rate of 40 dots per minute [3] (which would give about 2½ words per minute); as may be judged from the circumstance, that the natural rate of vibration of the needle and mirror, when deflected and left to oscillate under the terrestrial influence alone, was about 84 vibrations (that is, 42 motions in one direction, and 42 returns). In this condition of the galvanometer, however, the ordinary earth-currents through the cable (when connected with plates of the same metal, sunk in the earth or sea on the two sides of the Atlantic), would far exceed the signal-current, and might often be as much as 20 or 30 cells would produce. There would be no difficulty in compensating their effect by steel magnets properly placed beside the galvanometer, in directions perpendicular to the plane of the mirror; or in directly balancing them with an opposing electromotive force, by means of an instrument for dividing the electromotive force of a battery into any desired number of parts, which the author constructed and applied for the purpose at Valentia. Sometimes, however, in certain terrestrial and atmospheric conditions (especially when aurora borealis is seen), the changes in the earth-currents are so rapid, that they would seem like signals, and render the reading of a message for the time difficult or impossible. By rendering the galvanometer less sensitive, which is best done by introducing steel magnets to add to the earth's directing influence, and using a proportionately more powerful sending battery, this disturbing influence is diminished, and may be rendered practically harmless, or as nearly as possible so.

[3] The rate from Newfoundland to Valentia was raised from 20 dots per minute, for which the sending mechanism was prepared, up to from 40 to 42. according to instructions telegraphed from Valencia, when the experience of the mirror galvanometer as a receiving instrument was bad.

It is convenient in other respects also to have the needle more powerfully directed than by the earth alone; and the action will probably be found in all respects satisfactory, if about nine times the earth’s horizontal force at Glasgow is applied to direct the needle (as is easily done by means of a single bar, or two or three bars, of hardened and magnetized steel, a few inches long). The oscillations of the needle will be then three times more rapid (or about 252 per minute), and a battery of 5 cells will be amply sufficient for signalling. If from 10 to 20 cells be used, there will be superabundant power for Morse signalling at the highest attainable rate, with the proper compensations introduced for inductive embarrassment which the mathematical theory has pointed out. If, as is to be hoped will soon be the case, the Red Sea cable is brought into working order, it is probable that 10 words a minute or upwards will be obtained as a regular rate of Morse signalling through one of its sections of from 500 to 700 nautical miles, by using for sending a battery of 10 Daniell’s, with a key constructed to produce the theoretical compensations, and, for receiving, a mirror galvanometer, adjusted with a still much higher directive force of steel magnets, recording by “human relay” on a riband of paper to be marked by hand, without local battery, either with a Morse style or pencil (or with Siemens’ ink-marking apparatus, constructed on the whole in accordance with a principle brought into use by John in Vienna, and Digney frères in Paris, which has been provided already for the Red Sea Company to be worked by relay and local battery).

Fig. 11

The “marine galvanometer” constructed by the writer for use on board the Agamemnon and Niagara, differs from the land mirror galvanometer only in the use of still higher directive force on the needle, and in the mode of suspension of the mirror and needle, which was, by means of a fine platinum wire, or, as has since been found better, a stout bundle of 20 or 30 silk fibres, firmly stretched between two fixed points of support. This instrument, as now constructed with some improvements, is represented in the accompanying drawings. The first shows the back of the mirror with attached needle, (a piece of thin, flat file, ground smooth,) the filament bearing it, the mountings for holding the two ends of the filament, and the bobbin of the front half of the galvanometer coil, which bears these mountings.

The second drawing shows a perspective view of the instrument complete, with stand and lamp, taken from a photograph. The galvanometer itself is protected by a strong plate-glass shade; in the front side of which (not seen) a slit, about 1½ long by ½ broad, is cut to allow the light from the lamp to enter, and to be reflected back from the mirror to the scale without loss by passing through the plate-glass. A large compound magnet of sheet-steel, bent and tempered glass hard, is seen round the galvanometer, and N marks one set of its poles: the other set, of contrary name, n is diametrically opposite. By an adjusting screw, A, this magnet can be turned through a considerable angle about a vertical axis, to bring the spot of light to zero, or to any desired point of the scale.

Fig. 12

On the back of the back half of the galvanometer bobbin are seen rings and screws, forming a system of connections by which four parts of its coil may be used either singly, or connected in series, or in series of double arcs, or entirely in multiple arc. A similar set of connections, not seen in the view represented, are attached on the front, for similar arrangements of the front half of the coil. A set of connections G, on the side of the stand, allow the front and the back halves of the coil to be used either alone, or the two in series, or the two in double arc, and at any time one of these arrangements to be changed for another in a moment; also, to reverse the extreme galvanometer terminals in their connection with binding screws (not seen in the view) for attaching the outer electrodes.

Some of the outer connections of a testing apparatus, to be used along with the galvanometer, are seen on a box attached to the frame between the galvanometer and the lamp. This apparatus contains 18 standard resistance coils of German silver, [4] of resistances expressed in terms of 100,000,000 British absolute units by the numbers and fractions, 1024, 512, 256, 128, 64, 32, 16, 8, 4, 2, 1, ½, ¼, ⅛, 1/16, 1/32, 1/64, 1/64; also four conductors of widely different resistances, each accurately bisected (electrically), to be used any one, or in combination, one arm of one with one arm of another, to constitute part arc of a Wheatstone's balance, of which the remainder would be made up of conductor to be tested, and standards.

[4] Chosen because it changes less in its resistance, with changes of temperature, than any other metal yet known and available.

On the remote side of the box (not seen) are connections for applying these bisected conductors in the most convenient possible manner. On the top of the box is a double spring key (not shown) for applying battery electrodes, in either direction, to proper terminals for the Wheatstone's balance when formed; and D represents another kind of double spring key for reversing instantaneously the connections between the terminals of the galvanometer coil and their proper terminals in the “balance.”

A set of screws on insulated studs and plates, planned to connect the resistance coils in series, and to short-circuit any one or more of them, or to connect as many of them as may be desired, in multiple arc, between two main terminals, are marked F F.

The first mode of connection is the same as that originally reduced to practice graphic by Mr Cromwell F. Varley, and makes up the required resistance by the addition of resistances. The second, and the combination of the two plans is entirely novel, and, by allowing conducting powers to be added, gives the same sensibility at the low end of the scale of resistances, that Mr Varley’s does at the high end. It is essential that the graduation by 2 and the powers of 2 be followed, in the resistances, when the same set of coils is to be used in the two ways. If they were merely to be used in series, the numbers 1, 2, 3, 5, 10, 12, 13, 15, &c, preferred by Mr Varley, would be somewhat more convenient.

Beyond the testing apparatus is seen the top of the lamp-glass, the scale, and the slit behind which the flame of the lamp is placed. A common flat-wicked paraffin lamp is used, with the edge of its flame towards the galvanometer. In performing accurate measurements, the slit in front of the flame is made very fine, by a screw acting on two plates of metal, so as to give a fine line of light to read from on the scale. In using the instrument for receiving telegraphic signals, the slit is made so wide as to allow the full image of the flame to fall on the scale, which will cover two or three of the 1/40 i divisions.

In the marine galvanometer of this kind hitherto made, the adjusting magnets are so highly magnetised as to give about 200 times the Glasgow terrestrial horizontal directive power. The rapidity of the natural oscillations of the needle is so great as to show merely a spread out band of light, on the scale, when a current is suddenly made, or broken through the galvanometer coil; but dynamical principles show that it roust be about V200 (or 14 1/7) times what has been mentioned above as the natural rate for a similar needle under terrestrial influence at Glasgow alone; and must therefore be about 20 vibrations per second.

A single cell applied to the galvanometer coil, arranged in series, gives a deflection of about 75 divisions, on a scale 18 t from the mirror. Care having been taken to balance the needle and mirror, so as to have the centre of gravity precisely in the line of the supporting filament, the indications of the instrument remain steady and undisturbed during the roughest usage it can be exposed to on board ship. The greatest amount of tremulous motion, in the table or support on which it is placed, does not cause vibrations in the spot of light, and the whole may be inclined to an angle of 45° in any direction, without sensibly disturbing the position of the spot of light on the scale.

By diminishing somewhat the magnetism of the adjusting magnets, the instrument may be brought to any requisite degree of sensibility suitable for receiving ordinary telegraphic signals, in the manner explained above, for which it is perfectly adapted; but for the highest degree of sensibility which has been stated, its multiple filament would be unsuitable.

The most complete possible mode of recording telegraphic signals would be to show a curve representing precisely the varying strength of the current received at every moment. An instrument executing this plan has been constructed by the writer, and found to succeed in circumstances essentially more trying to its capabilities than those of actual work through a long submarine line. In this instrument the record consists of minute perforations in a broad riband of paper, made by a very rapid succession of sparks from a Ruhmkorff’s induction coil, worked by a local battery at the receiving-station. A vertical piece of platinum wire attached to a light horizontal glass index, borne by a finely but firmly suspended magnetic needle, conducts the perforating current from a vessel of water, into which its lower end dips, to the riband of paper, which is carried near its upper end in a direction perpendicular to its line of motion. This instrument will disentangle signals from inductive and other embarrassments, in cases in which even the method by mirror galvanometer and human relay fails. But it requires more power to work it, at a sufficient speed, than the mirror galvanometer. It is probable, however, that it may be improved so as to require no more than has been mentioned above as sufficient for the mirror; and, if so, it even may come to be preferred to every other method hitherto tried or proposed for receiving messages through long submarine lines.

Section XI: Atlantic Telegraph

When a great experiment is made, the materials employed are generally lost, and knowledge is too often the only form of power acquired as the result. In the year 1857, as much iron as would make a cube of 20 feet side, was drawn into wire long enough to extend from the earth to the moon, and bind several times round each globe. This wire was made into 126 lengths of 2,500 miles, and spun into 18 strands of 7 wires each. A single strand of 7 copper wires of the same length, weighing in all 110 grains per foot, was three times coated with gutta-percha, to an entire outer thickness of 0.4 of an inch; and this was “served” outside with 240 tons of tarred yarn, and then laid over with the 18 strands of iron wire in long contiguous spirals, and passed through a bath of melted pitch. In August of that year, about ⅛th of the entire length of this compound rope was laid from the Irish, shore westwards, and lost by a breakage at the stern of the Niagara. The remainder was conveyed in the two ships to Devonport, and stored for the winter in Keyham dockyard. A length of 55 miles of the portion which had been lost was lifted in tolerably good condition a few months later. During the ensuing winter and spring, about 900 miles more of similar cable were manufactured; and in the months of April and May 1858, the whole length of 3,000 British statute miles was shipped on board HMS Agamemnon and the US steam frigate Niagara.

After an experimental cruise in the Bay of Biscay, to test the appliances for laying the cable by actual trials in water 2,500 fathoms deep, and some slight alteration of the machinery made in consequence on returning to Plymouth, the two ships, accompanied by HMS Valorous and Gorgon, paddle-steamers—the former tender to the Agamemnon, the latter to the Niagara—set out for the middle of the Atlantic on the 10th June. After three unsuccessful, but not discouraging attempts, in which between 400 and 500 statute miles of cable were lost, the ships returned from the different points they had reached, to rendezvous in Queenstown harbour, where, on the arrival of the Agamemnon on the 12th July, the whole squadron were again together, and remained long enough to take in coal and make other requisite preparations for a final attempt. On the 17th July they again put to sea westwards. On the 29th they met at the mid-ocean rendezvous, joined the ends of the cable between the two ships bearing it, and commenced laying it in 2,400 fathoms water, the Niagara continuing westwards, and the Agamemnon returning to the east. This time no accident stopped the continuous paying out; and on the 5th of August the two ships cut the cable, and left the ends on shore on the two sides of the Atlantic.

The possibility of laying an electric cable across 2,000 miles of ocean, in depths of from 1,800 to 2,500 fathoms— seriously doubted by nearly all practical engineers, and considered a perfectly chimerical project by some of the most eminent—was thus triumphantly demonstrated. The risk of failure in future attempts was brought almost within the limits of a common “sea risk;” the weather having been by no means favourable, especially on the Agamemnon's side, where, during three days of the six, strong breezes from several quarters were experienced, and at one time a fresh gale of head-wind. The telegraphic operations performed between Valencia and Trinity Bay during the remainder of the month of August will render the year 1858 ever memorable in the history of the world. The world’s news was read on the same day in the capitals of Europe and. America. Question and answer passed freely, and friendly conversation was held, between the operators on the two sides of the Atlantic. The Queen of England and the President of the United States interchanged congratulatory messages, and assurances of mutual good-will on the part of the two great nations under their authority. One short message saved thousands of pounds of money, and an inestimable amount of anxiety, by giving timely notice of an accident, which disabled one of the Transatlantic steamers off the American coast, bound for England. Another—nearly the last utterance of the failing cable—countermanded two British regiments under orders to embark, and prevented them from leaving the American colonies on a bootless voyage across the Atlantic.

The last words of the Atlantic telegraph were read at Valencia on the 20th October 1858 “two hundred and forty t-k (? two)—Daniell’s now in circuit.” The full message—as was afterwards learned in the old, and, alas! at this moment, the only way of receiving intelligence from the other side of the Atlantic—was “two hundred and forty trays [1] and seventy-two liquid Daniell’s now in circuit.” This prodigious power, one thousand times as much as would have given perfect signals to the mirror galvanometer in use as receiving instrument at Valencia through the same cable, if ordinarily well insulated, proved insufficient for continuing telegraphic work, and it became certain that only mending the insulation in one or more faulty places could restore communication. Before the process of laying was complete, indications of very defective insulation had been given by the readings of the “marine galvanometer” recorded on board each ship. After the ends were landed, the insulation became farther deteriorated; every attempt to establish communication by means of the regular telegraphic instruments prepared for the use of the company proved a failure; and it was only by the introduction of the mirror receiving instrument on each side of the Atlantic that an interchange of intelligence was effected.

[1] A form of battery introduced by the writer of this article for nautical and land telegraphic use, which has the good electric qualities of the Daniell's, without the disadvantages inseparable from the use of porous cells. It consisted of copper trays, each strewn with crystals of sulphate of copper covering its bottom, and filled with sawdust moistened with sulphuric acid and water, on the top of which an amalgamated zinc plate was laid. These trays, thus charged, were piled one with its outside copper bottom on the top of the zinc of another, in columns of from five to twenty, after the fashion of an old voltaic pile. The writer has since made many trials of different forms and has been led to plan, with the assistance of Mr F. Jenkin, an improved “Sawdust Daniell’s,” from the use of which, in various applications, be anticipates much advantage.

As soon as messages began to come from Newfoundland, they were read with very great ease at Valencia on the new system. At Newfoundland, on the other hand, three days passed, during which messages, continually being sent from Valencia, were not read or even recognised to be signals at all; and it was only by the occasional introduction of the mirror instrument into circuit, in accordance with instructions given at Devonport to special operators sent out in charge of it, that the first words were read on the other side of the Atlantic. A “detector,” or common telegraph galvanometer, of a kind then much used by British practical electricians, was next tried, and it was found possible to read by it, although with great difficulty (the signal deflections scarcely amounting to half a degree), and only at an excessively slow rate (half a word per minute or less); but when, as was often the case, these attempts failed altogether, the mirror was had recourse to, to see whether any message was coming or not. Matters were conducted in this unsatisfactory way at the Newfoundland station for about a week after the first words had been read, until the mirror was permanently introduced into circuit and regularly used as receiving instrument, in accordance with an order transmitted through the cable from Valencia on the 21st of August. From that time forward the messages were read with about equal ease at the two ends; but the days of the first Atlantic Telegraph were numbered. On the 1st of September it conveyed the two military messages. On the following day it conveyed one congratulatory message for a public meeting in New York, addressed to Mr Cyrus Field, to whose untiring energy it in a great measure owed its existence; and it failed to convey a second similar message on the same day. From that time till its death-struggle, on the 20th of October, it was silent.

And now that splendid combination of matter lies on the bottom of the Atlantic, its Newfoundland end irrecoverably lost. A few miles of it—possibly 200 miles—may be lifted from Valencia, and used for some minor telegraphic work, perhaps cut up into target telegraphs; but its value can scarcely, if at all, exceed the expense of lifting it. £375,000 have been spent in the great work; and the world, if not the subscribers, will reap the profit, when Europe and America become united by a cable that will not fail.

Return to the Atlantic Cables index page

Last revised: 21 September, 2011

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:

—Bill Burns, publisher and webmaster: