TOC | Previous Section: W. F. Melhuish--1890 | Next Section: Professor Erich Rathenau--1894
A History of Wireless Telegraphy (2nd edition, revised), J. J. Fahie, 1901, pages 119-130:


    Early in 1892 Mr Charles A. Stevenson of the Northern Lighthouse Board, Edinburgh, threw out the suggestion that telegraphic communication could be established between ships at sea and between ship and shore by means of coils. 71 He tried many experiments in the course of that year, the results of which he reported to the Royal Society of Edinburgh on January 30, 1893. In this paper 72 he describes two methods of communicating between the shore and a ship, each of which supposes a cable to be submerged along the coast, and to be earthed in the sea--presumably (for the account is not clear) through an induction coil or transformer.
    In the first method the ship has a wire, with a telephone in circuit, stretched from bow to stern, and terminating in coils which dip into the water, and which may or may not be insulated. When the ship approaches or crosses the cable at a right angle, or nearly so, the currents set up in the latter by a magneto-electric machine at the shore end are rendered audible in the telephone on board. If the coils be in the line of the cable, as they will be when the ship is over it lengthways, or approaches it broadside on, no sound is heard in the telephone, thus indicating the position of the vessel with respect to the known direction of the cable. An insulated wire, 400 feet long, was laid through a small lake (Isle of May) of brackish water, 15 feet deep. Alternations of current were set up in this wire by the bobbins of three-fifths of a De Meritens' magneto-electric machine (yielding 80 volts at its terminals). A small boat, having a wire with a telephone in circuit stretched from bow to stern and terminating in coils dipping into the water 10 feet apart, was rowed about in the vicinity of the submerged wire, and it was found that the currents in this wire were distinctly audible in the telephone up to a distance of over 300 feet.
    The second method described by Mr Stevenson consisted in dropping into the sea from the deck of a ship a large electro-magnet (3 feet long, with 2000 turns of one-eighth inch copper wire) with a telephone in circuit. The interruptions of the current from six dry cells through a wire 200 feet long could be heard in the telephone at a distance of 40 feet in air, while with twelve dry cells the effect was audible through 60 feet of salt water. Indeed, he says, there seemed to be little difference whether the medium was air, or fresh or salt water.
    The first described method was practically tried in America, early in 1895, by Professor Lucien Blake, and was favourably spoken of in his report to the American Lighthouse Board, September 1895. A lightship is moored in 65 feet of water and four-and-a-half miles off Sandy Hook. Out from the shore an armoured cable was laid, terminating in a transformer, the core of the cable being earthed on the armour through the high-resistance coil of the transformer, while the terminals of the lower-resistance coil were earthed in the following manner. Three insulated copper wires, each one quarter mile in length, were laid parallel on the sea bottom and 300 feet apart. At one end they were connected together and joined to one terminal of the transformer, while the other and distant ends were earthed by means of pieces of wire netting about 20 feet square. The other terminal of the transformer was earthed by a similar piece of netting. At the centre of this "grid" arrangement the ship's moorings were fixed. The connections on board were as follows: The two hawse pipes were connected in the pipe by a copper bar, and extra plates were put between the metallic sheathing and the hawse pipes so as to ensure good sea connection. From the copper bar an insulated wire was carried to a telephone in the aftercabin, thence a wire from the other telephone terminal was carried aft and connected to a tail-piece of flexible conductor over the stern and dipping into the water.
    When intermittent currents were sent into the cable from the shore, there was set up in the area under the ship "an unequal electrical distribution, such that the potentials were of sufficient difference at the two ends of the ship to operate the telephone on board. Experiments showed that sufficient difference existed between bow and rudder sheathing, and even between bow and stern sheathing, to operate the telephone, but the effect was greatest with bow sheathing and stern tail-rope."
    In another experiment on board the lighthouse tender Gardinia, the telephone circuit terminated in two plates, 7 feet by 3 feet, submerged from bow and stern, a distance of 113 feet. Here too, "sufficient difference of potential existed between the plates to make conversation with the shore possible while the tender was steaming about in the neighbourhood."
    Mr Stevenson speaks of this as an electro-static effect, but as I understand it, and certainly as it has been tried by Prof. Blake, the method seems to belong more to the conductive order, and to be identical with that of Messrs Smith & Granville, to be presently described (infra).
    Mr Stevenson calls his second method "electro-magnetic," in contradistinction to the first or "electro-static" one, and with certain dispositions of the submerged cable it might be available for communicating between the shore and a lightship through a few fathoms of water. It is, however, interesting to us as being a step forward in the evolution of Mr Stevenson's ideas from conductive to inductive methods.
    In a further paper, read before the Royal Society of Edinburgh, March 19, 1894, he describes his experiments with insulated coils of wire, or more correctly spirals and says that a trial of his new method on a large scale had recently been made with a view of ultimately employing it for effecting communication between Muckle Flugga, in the Shetlands, and the mainland.
    As regards the efficacy of the principle, the inductive effect of one spiral on another at a distance has long been known; but hitherto, even with a very strong battery, it was impossible to bridge a greater distance than 100 yards, which for practical purposes was, of course, useless.
    It is evident that if two coils are placed vertically so that their axes are coincident, their planes being parallel, or if they be placed so that their planes are in the same plane, they will be in good positions for electric currents sent in one to be apparent by induction in the other. For a small diameter, and where the electrical energy is small, the first position is suitable; but where the energy is great and the diameter of coils great--in fact, when it is wished to carry the induction to many times the diameter of the coils--then it will be found that it is better to let the two coils be in the same plane, as it becomes impracticable to erect coils of large diameter with their planes vertical, but it is easy to lay them on their sides.
    Mr Stevenson made a large number of laboratory experiments on the interaction of coils, with the view of calculating the number of wires, the diameter of coils, the number of ampères, and the resistance of the coils that would be necessary to communicate with Muckle Flugga; and, after a careful investigation, it was evident the gap of 800 yards could, with certainty, be bridged by a current of one ampère with nine turns of post-office wire in each coil, the coils being 200 yards in diameter, and with two good telephones on the hearing coil.
    Two coils, on telegraph-poles and insulators, were erected at Murrayfield, one coil being on the farm of Damhead and the other on the farm at Saughton, and as nearly as was possible on a similar scale, and the coils of similar shape to what was wished at Muckle Flugga. On erecting the coils, communication was found impossible, owing to the induction currents from the lines from Edinburgh to Glasgow, the messages in those lines being quite easily read, although the coils were entirely insulated and were not earthed. The phonophore which the North British Railway Company have on their lines kept up nearly a constant musical sound, which entirely prevented observations. On getting the phonophore stopped, it was found that 100 dry cells, with 1·2 ohms resistance each and 1·4 volts, gave good results, the observations being read with great ease in the secondary by means of two telephones. The cells were reduced in number down to fifteen, and messages could still easily be sent, the resistance of the primary being 24 ohms and the secondary no less than 260 ohms. If the circuit had been of good iron, with soldered joints and well earthed, the resistance would have been only 60 ohms. The induced current generated in the secondary would therefore be in the ratio of 480 [? 520] to 210; or, allowing for the resistance in the two telephones, we get practically only half the current we would have got if the line bad been a permanent in place of a temporary one.
    A trial was made of the parallel-wire system: 73 with 20 cells the sound was not heard, and with 100 cells it was heard as a mere scratch in comparison with the sound with the coil system with only 15 cells. A trial was made with the phonophore: the coils worked with 10 cells with perfect ease, and a message was received with only 5 cells. Speech by means of Deckert's transmitter was just possible, but it is believed that if the hearing circuit had been of less resistance it would have been easy to hear.
    "It is difficult," says Mr Stevenson, "to understand how this system of coils, in opposition to the parallel-wire system, has not been recognised as the best; for assume that, with the arrangement we had, we heard equally with 100 cells by both systems, both having the same base (200 yards), then, by simply doubling the number of turns of wire on the primary and using thick wire, the effect would have been practically doubled, whereas by the parallel-wire system there is nothing for it but to increase the battery power. The difficulty of the current is thus removed by using a number of turns of wire. It must always be borne in mind that the effect is the result of simply increasing the diameter, keeping current and resistance the same. The larger the diameter the better. What is wanted is to get induction at a great distance from a certain given base with a small battery power, and the laboratory experiments and the trials in the field show that the way to overcome the difficulty of the current is by using a number of turns of wire. The secret of success is to apportion the resistance of primary and secondary, and the number of turns on each, to a practical battery power."
    1. Coil System.--At 870 yards from centre to centre of coils, averaging each 200 yards diameter, with nine turns of wire, it was found that with a phonophore messages were sent with five dry cells, the resistance in primary being 30 ohms and the resistance of secondary 260 ohms, the current being 0·23 ampère, which, with nine turns, gives 2 ampère turns.
    2. With a file as a make and break, it worked with 10 cells, giving 0·4 ampère or 3·6 ampère turns.
    3. Parallel-Wire System.--With a file as a make and break, and with parallel lines earthed, it was heard with 100 cells, giving 1·1 ampère. Fig. 14
    The primary coil circuit was entirely metallic in the Murrayfield trials, as it would have to be if erected at Muckle Flugga; but the secondary coil was earthed. When, however, the secondary was also made a complete insulated metallic circuit, with eight turns of wire, there seemed to be little difference in the result.
    The calculation of the diameter necessary to hear at a given distance is simple, from the fact that the hearing distance is proportional to the square root of the diameter of one of the coils, or directly as the diameter of the two coils, so that, with any given number of ampères and number of turns, to hear double the distance requires double the diameter of coils, and so on. 74
    In concluding his paper, Mr Stevenson says:--
    "It has been attempted to be shown that the coil system is not only theoretically but practically the-best; and I trust that we will soon hear of the Admiralty, &c., experimenting with it, and ultimately putting it in practice. Meantime my brother has recommended the Commissioners of Northern Lighthouses to erect the coil system at Muckle Flugga, and the Commissioners have approved; and I hope soon to hear of the erection of this novel system of communication at the most northern point of the British Isles, as well as on our warships to assist in their manoeuvring, by the establishment of instantaneous communication unaffected by wind or weather.
    "The application of the coil system to communication with light vessels is obvious-viz., to moor the vessel in the ordinary way, and lay out from the shore a cable, and circle the area over which the lightship moorings will permit her to travel by a coil of the cable of the required diameter, which will be twice the length of her chain cable. On board the vessel there will be another coil of a number of turns of thick wire. Ten cells on the lightship and ten on the shore will be sufficient for the installation." 75
    In a recent communication 76 Mr Stevenson gives some additional particulars. Referring to his proposed installation at the North Unst lighthouse, on Muckle Flugga, he tells us a gap of half a mile had to be bridged. The Commissioners of Northern Lighthouses, being impressed with the experiments shown them on a small scale--through stone and mortar--and on a larger scale at Murrayfield, decided on installing the system on Muckle Flugga; but, subsequently, financial difficulties arose, and the project was allowed to drop.
    "It is well to remember," he says, "that in the Murrayfield trials a small number of cells was purposely used. Theory and formulæ give one the impression at first sight that a single outstretched wire is always best--the simple fact of getting a greater effect at a distance as a coiled wire is uncoiled and made straight supporting this impression; but formulæ, if they are to be practical, ought to take into account a limited area and workable amounts of resistance, current, &c., and then the fact is disclosed that the coiling of wires (whether condensers be used with them or not) becomes an advantage for most work which the engineer will be called upon to deal with.
    "It is not necessary, as has been stated, that the coils should be identical in size and shape. Far from it; each case must be treated for size and configuration by itself. For instance, in the case of Muckle Flugga, my design was for a line two miles in length on the mainland, with a coil at the end enclosing a larger area than the one on the rock, which latter was opened out to the maximum possible. Again, in the case of Sule Skerry and the Flannan Islands, on the north-west of Scotland, where telegraphy by induction would be of great value, it would be impossible to make the coils of large diameter, but the coil on the mainland should be of large dimensions; indeed a single long wire with the ends earthed would be, perhaps, the best arrangement.
    "For guarding a dangerous coast, a similar wire of many miles in length would be suitable for communicating warning signals to vessels on board of which were detectors, with coils necessarily of small dimensions. There are two ways of doing this, both of which I have tried. First, by means of a submarine cable along the line of coast. In this case the currents set up in the cable have to bridge only the sheet of water to the vessel, say twenty fathoms; or, if an electro-magnet be let down from the ship, only four or five fathoms. But here the cost and maintenance of a cable would be a weighty objection. The other way is to erect a pole line on shore, either along the coast or in the form of a coil on a peninsula. The main difference from the first plan is that the currents would have to be stronger to bridge the distance of several miles instead of a few fathoms; but the cost in comparison with a cable would be very small. I have tried this system with two miles of pole line and a coil about a quarter of a mile distant with perfect and never-failing success.
    "I have made numerous trials of the coil versus parallel-wire system since 1891, and I have found--and other observers seem also to have found--that it is not practical to work the latter more than three or four times the length of base; whereas by coils I have found it possible to work many times their diameter. Thus in 1892, at the Isle of May lighthouse, I signalled to a distance 360 times the diameter of an electro-magnet coil with currents from a de Meritens' magneto-electric machine. Again, at Murrayfield, I signalled four times the base with five dry cells; and I have in Edinburgh a coil with iron core 17 inches diameter, which with one cell can easily signal through a space twenty-five times its diameter."

    71 'Engineer,' March 24, 1892.
    72 On Induction through Air and Water at Great Distances without the use of Parallel Wires.
    73 I.e.,, Preece's method, to be presently described. See et seq., infra.
    74 Professor Lodge has recently shown that the law of distance is not the square root of diameter, but the two-thirds power, with a given primary current; and so doubling the circumference of each coil will permit signalling over more than double the distance, if other things can be kept the same. For such magnification, however, the thickness of the wire must be magnified likewise, or else more power will be consumed in the enlarged coil. 'Jour. Inst. Elec. Engs.,' No. 137, p. 803. Possibly Mr Stevenson did not take into account the increase in resistance owing to the increased length of wire, so that for practical purposes his formula may be sufficiently accurate.
    75 On May 28, 1892, Mr Sydney Evershed patented a similar method of using coils in connection with his very delicate receiving instrument or relay. The plan was actually tried in August 1896 on the North Sand Head (Goodwin) lightship. One extremity of the cable was coiled in a ring on the bottom of the sea, embracing the whole area over which the lightship swept while swinging to the tide, and the other end was connected with the shore. The ship was surrounded above the water-line with another coil. The two coils were separated by a mean distance of about 200 fathoms, but communication was found to be impracticable. The screening effect of the sea water and the effect of the iron hull of the ship absorbed practically all the energy of the currents in the coiled cable, and the effects on board, though perceptible, were very trifling--too minute for signalling. See Evershed's paper on Telegraphy by Magnetic Induction, 'Jour. Inst. Elec. Engs.,' No. 137, p. 852; also Stevenson on Telegraphy without Wires, 'Nature,' December 31, 1896.
    76 'Jour. Inst. Elec. Engs.,' No. 137, p. 951; also No. 139, p. 307.
TOC | Previous Section: W. F. Melhuish--1890 | Next Section: Professor Erich Rathenau--1894