TOC | Previous Section: Sir W. H. Preece's Method | Next Section: G. Marconi's Method
A History of Wireless Telegraphy (2nd edition, revised), J. J. Fahie, 1901, pages 161-176:


    Mr Smith's researches in wireless telegraphy date back to 1883. His first suggestions, of the induction order, were contained in a paper on Voltaic-Electric Induction, which he read before the Institution of Electrical Engineers on November 8 of that year. These have already been noticed in our account of Edison's invention (supra).
    Somewhat late; early in 1885, Mr Smith turned his attention to conduction methods, and worked out a plan which, in a modified form, has been in actual operation for the last three years.
    The rationale of the system is described by Mr Smith as follows:--
    "Messages have been sent and correctly received through a submarine cable two thousand miles in length, the earth being the return half of the circuit, by the aid of the electricity generated by means of an ordinary gun-cap containing one drop of water; and small though the current emanating from such a source naturally was, yet I believe it not only polarised the molecules of the copper conductor, but also in the same manner affected the whole earth through which it dispersed on its way from the outside of the gun-cap to its return, through the cable, to the water it contained. I further believe that the time will come, perhaps sooner than may be expected, when it will be possible to detect even such small currents in any part of the world in the same way that it is now possible to do in comparatively small sections of it.
    "For researches of this description it is necessary to employ as sensitive an instrument as it is possible to obtain, to pick up, so to speak, such minute currents. Now, there is that wonderful instrument the telephone. I say wonderful advisedly, for as far as I know it is not to be equalled for the simplicity of its mechanical construction and the ease with which it can be manipulated, and yet is so peculiarly sensitive. I have used it in most of my experiments as the receiving instrument, although of course there are other well-known instruments that could be employed, as all depends upon the potential of the current to be detected. The sending arrangement was either an ordinary Morse key so manipulated for a short or long time as to give the necessary sounds in the telephone to represent dots and dashes, or a double key and two pieces of mechanism giving dissimilar sounds were employed with good results. I gave much time and thought to the subject, the results of each experiment giving me much encouragement to proceed.
    "Of the many experiments made I select the following, as I think it will clearly illustrate my system for telegraphing to a distant point not in metallic connection with the sending station. A wooden bathing-hut on a sandy beach made a good shore station, from which were laid two insulated copper wires 115 fathoms in length. The ends of the wires, scraped clean, were twisted round anchors, their position being marked by buoys about 100 fathoms apart, and in about 6 fathoms of water. Midway between the two a boat was anchored with a copper plate hanging fore and aft about 10 fathoms apart, and consequently about 45 fathoms from either end of the anchored shore wires. This boat represented the sea station, and, owing to the state of the sea, a very wet and lively one it proved; therefore, taking this fact into consideration, together with the crude nature of the experiment, it was remarkable with what distinctness and ease messages were passed. The last message sent from shore was, 'Thanks: that will do; pick up anchors and return.' To this the reply came from the boat, 'Understand,' and they then proceeded to carry out instructions. The boat employed was a wooden one, but it would have been much better for my purpose had it been of metal, for then I should have used it instead of one of the collecting plates, as the larger the surface of these plates the better the results obtained.". 95
    This method was secured by patent, June 7, 1887, from the specification of which (No. 8159) I take the following particulars: At the present time wherever electric telegraph communication is established between the shore and a lighthouse, either floating or on a rock, at a distance from the shore, it is effected through an insulated conductor or cable. Much difficulty is, however, experienced owing to the rapid wearing of the cable, so that it is liable to break whenever a storm comes on, and when, consequently, it is most required to be in working order. By this invention 'communication can be effected between the sending station and the distant point without the necessity of metallic connection between them. Fig. 25
    A in the drawing (fig. 25) is a two-conductor cable led from a signal-station B on shore towards the rock C. At a distance from the rock one of the conductors is led to a metallic plate D submerged on one side of the rock, and at such a distance from it as to be in water deep enough for it not to be affected by waves. The other conductor is led to another metallic plate E similarly submerged at a distance from the opposite side of the rock. F F are two submerged metallic plates, each opposite to the plates D and E respectively. G G are insulated conductors leading from the plates F F to a telephone of low resistance in the lighthouse H.
    To communicate from the shore, an interrupter or reverser I and battery K are connected to the shore ends of the two-wire cable. The telephone in the lighthouse circuit then responds to the rapid makes and breaks or reversals of the current, so that signalling can readily be carried on by the Morse alphabet. If a vibrating interrupter or reverser be used, a short or long sound in the telephone can be obtained by a contact key held down for short or long intervals.
    A more convenient way is to use two finger-keys, one of which by a series of teeth on its stem produces a few breaks or reversals of the current, whilst the other key when depressed produces a greater number of breaks or reversals.
    For communicating from the lighthouse to the shore a battery and make-and-break apparatus are coupled to the insulated conductors on the rock, and a telephone to the shore ends.
    In the same way communication could be carried on from the shore to a vessel at a distance from it, if the vessel were in the vicinity of two submerged plates or anchors, each having an insulated conductor passing from it to the shore, and if two metallic plates were let go from the vessel so that these plates might be at a distance apart from one another. The position of the two submerged plates might be indicated by buoys. In this way communication might be effected between passing vessels and the shore, or between the shore and a moored lightship or signal-station.
    A similar result might be obtained with a single insulated conductor from the shore by the use of an induction apparatus, the ends of the secondary coil being connected by insulated conductors to the submerged plates.
    An important modification of this method was subsequently effected by Messrs Willoughby S. Smith & W. P. Granville, 96 based on the following reasoning Fig. 26
    In fig. 26 represents an insulated conductor of any desired length with ends to earth E E as shown. C is a rock island on which is extended another insulated wire C D, with its ends also connected to earth. Now, if a current is caused to flow in A B, indications of it will be shown on a galvanorneter in the circuit C D. This is Preece's arrangement at Lavernock and Flat Holm.
    Now, if we rotate the line A B round A until it assumes the position indicated in fig. 27, we have Messrs Smith & Granville's arrangement, where, owing to the proximity of B to D signalling is practicable with a small battery power. Thus, where the distance from B to D was 60 yards, one Leclanché cell was found to be ample. As a permanent current in A B causes a permanent deflection on the galvanometer in C D this deflection cannot be produced otherwise than by conduction.
    Again, let A B (fig. 28) represent an insulated conductor having its ends submerged in water (the distance between A and being immaterial). Now cause a current to flow continuously, and it will be found that the water at each end of the conductor is charged either positively or negatively (according to the direction of the current) in equipotential spheroids, diminishing in intensity as the distance from either A or B is increased. To prove this, provide a second circuit, connected with a galvanometer, and with its two ends dipping into the water. Now, it will be found that a current flows in the C D circuit as long as the current in A B is flowing; the current in C D diminishes as C and D are moved farther away from n, and also diminishes to zero if the points C D are turned until they both lie in the same equipotential curve as shown by the dotted line.
    It must be well understood that although, for the sake of clearness, the equipotential curves are shown as planes, yet in a body of water they are more or less spheres extending symmetrically around the submerged ends of the conductor, and therefore it is evident from the foregoing that the position of C D, in relation to B, must be considered not only horizontally but vertically. 97 Fig. 27
    Early in 1892 the Trinity Board placed the Needles Lighthouse at the disposal of the Telegraph Construction and Maintenance Company, so that they might prove the practicability of the method here described. The Needles Lighthouse was chosen on account of its easy access from London.
    In May 1892 an ordinary submarine cable was laid from Alum Bay to within 60 yards of the lighthouse rock, where it terminated, with its conductor attached to a specially constructed copper mushroom anchor. An earth plate close to the pier allowed a circuit to be formed through the water. On the rock itself two strong copper conductors were placed, one on either side, so that they remained immersed in the sea at low water, thus allowing another circuit to be formed through the water in the vicinity of the rock.
    The telephone was first tried as the receiving instrument, with a rapid vibrator and Morse key in the sending circuit. This arrangement was afterwards abandoned, as it was not nearly so satisfactory as a mirror-speaking galvanometer, and the men, being accustomed to flag work, preferred to watch a light rather than listen to a telephone. The speaking galvanometer used is a specially constructed one, and does not easily get out of order, so that, everything being once arranged, the men had only to keep the lamp in order.
    Messrs Smith & Granville devised a novel and strong form of apparatus for a "call," and by its means any number of bells could be rung, thus securing attention. The instruments both on rock and shore were identical, and, in actual work, two to three Leclanché cells were ample. Fig. 28
    By the means above described, communication was obtained through the gap of water 60 yards in length. This by no means is the limit, for it will be apparent that the gap distance is determined by the volume of water in the immediate neighbourhood of the rock, as well as by the sensitiveness of the receiving instrument and the magnitude of the sending current.
    This method is well suited for coast defences. For instance, if a cable is laid from the shore out to sea, with its end anchored in a known position, then it would be easy for any ship, knowing the position of the submerged end, to communicate with shore by simply lowering (within one or two hundred yards of the anchored end) an insulated wire having the end of its conductor attached to a small mass of metal to serve as "earth," the circuit being completed through the hull of the ship and the sea. 98
    As this method has been in practical use at the Fastnet Lighthouse for the last three years, the following account of the installation, which has been kindly supplied by Mr W. S. Smith, will be of interest:--
    "The difficulty of maintaining electrical communication with outlying rock lighthouses is so great that it has become necessary to forego the advantages naturally attendant upon the use of a submarine cable laid in the ordinary way continuously from the shore to the lighthouse, inasmuch as that portion of the cable which is carried up from the sea-bed to the rock is rapidly worn or chafed through by the combined action of storm and tide. By the use of the Willoughby Smith & Granville system of communication this difficulty is avoided, for the end of the cable is not landed on the rock at all, but terminates in close proximity thereto and in fairly deep undisturbed water. This system, first suggested in 1887 and practically demonstrated at the Needles Lighthouse in 1892, has--on the recommendation of the Royal Commission on Lighthouse and Lightship Communication--been applied to the Fastnet, one of the most exposed and inaccessible rock lighthouses of the United Kingdom.
    "The Fastnet Rock, situated off the extreme S.W. corner of Ireland, is 80 feet in height and 360 feet in length, with a maximum width of 150 feet, and is by this system placed in electrical communication with the town of Crookhaven, eight miles distant.
    "The shore end of the main cable, which is of ordinary construction, is landed at a small bay called Galley Cove, about one mile to the west of the Crookhaven Post Office, to which it is connected by means of a subterranean cable of similar construction having a copper conductor weighing 107 lb. covered with 150 lb. of gutta-percha per nautical mile. The distant or sea end of the main cable terminates seven miles from shore, in 11 fathoms of water, at a spot about 100 feet from the Fastnet Rock; and the end is securely fastened to a copper mushroom-shaped anchor weighing about 5 cwt., which has the double duty of serving electrically as an 'earth' for the conductor, and mechanically as a secure anchor for the cable end.
    "The iron sheathing of the last 100 feet of the main cable is dispensed with, so as to prevent the possibility of any electrical disturbance being caused by the iron coming in contact with the copper of the mushroom; and, as a substitute, the conductor has been thickly covered with india-rubber, then sheathed with large copper wires, and again covered with india-rubber--the whole being further protected by massive rings of toughened glass.
    "To complete the main cable circuit, a short earth line, about 200 yards in length, is laid from the post office into the haven. Fig. 29
    "By reference to the diagram (fig. 29) it will be seen that if a battery be placed at the post office, or anywhere in the main cable circuit, the sea becomes electrically charged--the charge being at a maximum in the immediate vicinity of the mushroom, and also at the haven 'earth.' Under these conditions, if one end of a second circuit is inserted in the water anywhere near the submerged mushroom--for instance, on the north side of the Fastnet--it partakes, more or less, of the charge; and if the other end of this second circuit is also connected to the water, but at a point more remote from the mushroom--for instance, at the south side of the Fastnet--then a current will flow in the second circuit, due to the difference in the degree of charge at the two ends; and accordingly a galvanometer or other sensitive instrument placed in the Fastnet circuit is affected whenever the post office battery is inserted in the main cable circuit, or, vice versâ, a battery placed in the Fastnet circuit will affect a galvanometer at the post office.
    "In practice ten large-size Leclanché cells are used on the rock, the sending current being about 1.5 ampères, and in this case the current received on shore is equal to about ·15 of a milliampère. The received current being small, instruments of a fair degree of sensitiveness are required, and such instruments, when used in connection with cables having both ends direct to earth, are liable to be adversely affected by what are known as 'earth' and 'polarisation' currents, consequently special means have been devised to prevent this.
    "The receiving instrument is a D'Arsonval reflecting galvanometer, which has been modified to meet the requirements by mounting the apparatus on a vertical pivot, so that by means of a handle the galvanometer can be rotated through a portion of a circle--thus enabling the zero of the instrument to be rapidly corrected. This facility of adjustment is necessary on account of the varying 'earth' and 'polarisation' currents above mentioned.
    "An entirely novel and substantial 'call' apparatus has also been designed, which automatically adapts itself to any variation in the earth or polarisation current. It consists essentially of two coils moving in a magnetic field, and these coils are mounted one at each end of a balanced arm suspended at its centre and free to rotate horizontally within fixed limits. The normal position of the arm is midway between two fixed limiting stops. Any current circulating in the coils causes the whole suspended system to rotate until the arm is brought into contact with one or other of the stops--the direction of rotation depending upon the direction of the current. A local circuit is thus closed, which releases a clockwork train connected to a torsion head carrying the suspending wire, and thus a counterbalancing twist or torsion is put into the wire, and this torsion slowly increases until the arm leaves the stop and again assumes its free position. If, however, the current is reversed within a period of say five or ten seconds, then the clockwork closes a second circuit and the electric bell is operated. By this arrangement, whilst the relay automatically adjusts itself for all variations of current, the call-bell will only respond to definite reversals of small period and not to the more sluggish movements of earth-currents. It is evident that one or more bells can be placed in any part of the building. The receiving galvanometer and the 'call' relay have worked very satisfactorily, and any man of average intelligence can readily be taught in two or three weeks to work the whole system.
    "To enable the two short cables that connect the lighthouse instruments with the water to successfully withstand the heavy seas that at times sweep entirely over the Fastnet, it has been found necessary to cut a deep 'chase' or groove down the north and south faces of the rock from summit to near the water's edge, and to bed the cables therein by means of Portland cement. And since the conductors must make connection with the water at all states of sea and tide, two slanting holes 2½ inches in diameter have been drilled through the solid rock from a little above low-water mark to over 20 feet below. Stout copper rods connected with the short cables are fitted into these holes, and serve to maintain connection with the water even in the roughest weather, and yet are absolutely protected from damage."
    Mr Granville supplies some interesting particulars as to the difficulties of their installation at the Fastnet. 99 "The rock," he says, "is always surrounded with a belt of foam, and no landing can be made except by means of a jib 58 feet long--not at all a pleasant proceeding. Now, here is a case where the Government desired to effect communication telegraphically, but, as had been proved by very costly experiments, it was impossible to maintain a continuous cable, the cable being repeatedly broken in the immediate vicinity f the rock. This, therefore, is a case where some system of wireless telegraphy is absolutely necessary, but neither of the systems described would answer here. 100 Dr Lodge advises us to eschew iron, and to avoid all conducting masses. But the tower and all the buildings are built of boiler-plate, and that which is not of iron is of bronze. In fact, the rock itself is the only bit of non-conducting, and therefore nonabsorbing, substance for miles around. It is very clear in a case of this sort--and this is a typical case--that it is absolutely impracticable to employ here Dr Lodge's method. Now we bear in regard to the method used--and successfully used--at Lavernock, that a certain base is required, of perhaps half a mile, a quarter of a mile, or a mile in length; and that base must bear some proportion to the distance to be bridged. But where can you get any such base on the rock? You could barely get a base of 20 yards, so that method utterly fails. Then we come to the case suggested by Mr Evershed, of a coil which would be submerged round the rock. Well, where would the coil be after the first summer's breeze, let alone after a winter gale? Why, probably thrown up, entangled, on the rock. A few years ago, during a severe gale, the glass of the lantern, 150 feet above sea-level, was smashed in; and at the top of the rock, 80 feet above the sea-level, the men dare not, during a winter's gale, leave the shelter of the hut for a moment, for, as they said,--and I can well believe it,--they would be swept off like flies. This is a practical point, and therefore one I am glad to bring to the notice of the Institution; and, I repeat, if wireless telegraphy is to be of use, it must be of use for these exceptional cases."
    Strange as it may seem, we have been using, on occasion, wireless telegraphy of this form for very many years without recoguising the fact. Every time in ordinary telegraphy that we "work through a break," as telegraphists say, we are doing it. An early instance of the kind is described in the old 'Electrician,' January 9 and 23, 1863. Many years ago, in Persia, the author has often worked with the ordinary Morse apparatus through breaks where the wire has been broken in one or more places, with the ends lying many yards apart on damp ground, or buried in snow-drifts. As the result of his experiences in such cases the following departmental order was issued by the Director, Persian Telegraphs, as far back as November 2, 1881: "In cases of total interruption of all wires, it is believed that communication may in most cases be kept up by means of telephones. Please issue following instructions: Fifteen minutes after the disappearance of the corresponding station, join all three wires to one instrument at the commutator. Disconnect the relay wire from the key of said instrument, and in its stead connect one side of telephone, other side of which is put to earth. Now call corresponding station slowly by key, listening at telephone for reply after each call. Should no reply be received, or should signals be too weak, try each wire separately, and combined with another, until an arrangement is arrived at which will give the best signals." The Cardew sounder or buzzer has in recent years been added, and with very good results. It will thus be seen that Mr Willoughby Smith's plan is really an old friend in a new guise.
    In 1896 Mr A. C. Brown, of whose work in wireless telegraphy we have already spoken (supra), revived the early proposals of Gauss, Lindsay, Highton, and Dering, re the use of bare wire, or badly insulated cables, in connection with interrupters and telephones. He also applied his method to cases where the continuity of the cable is broken. "Providing the ends remain anywhere in proximity under the water, communication can usually be kept up, the telephone receivers used in this way being so exceedingly sensitive that they will respond to the very minute traces of current picked up by the broken end on the receiving side from that which is spreading out through the water in all directions from the broken end on the sending side." (See Mr Brown's patent specification, No. 30,123, of December 31, 1896.)
    Recently he has been successful in bridging over in this way a gap in one of the Atlantic cables; but in this he has done nothing more than the present writer did in 1881, and Mr Willoughby Smith in 1887.

    95 'Electrician,' November 2, 1888.
    96 See their patent specification, No. 10,706, of June 4, 1892.
    97 This fact, Mr Smith thinks, fully explains Preece's launch experiments (supra). For instance, when the launch towing the half-mile of cable parallel to the wire on the mainland was close to the shore, the cable, although allowed to sink, could only do so to a very limited extent, and therefore still remained in a favourable position for picking up the earth-currents from A B (fig. 28); but when one mile from the shore, and in deep water, the cable was able to assume somewhat of a vertical position with the two ends brought more or less into the same equipotential sphere, it naturally resulted in a diminution or cessation of the current in the C D or launch circuit, and hence the absence of signals.
    98 'Electrician,' September 29, 1893. See also the 'Times,' November 24, 1892.
    99 'Jour. Inst. Elec. Engs.,' No. 137, p. 941.
    100 I.e., those advocated by Professor Lodge and Mr. Sydney Evershed. See 'Jour. Inst. Elec. Engs.,' No. 137, pp. 799, 852.
TOC | Previous Section: Sir W. H. Preece's Method | Next Section: G. Marconi's Method