TOC | Previous Section: Professor Erich Rathenau--1894 | Next Section: Willoughby Smith's Method
A History of Wireless Telegraphy (2nd edition, revised), J. J. Fahie, 1901, pages 135-160:




The invention all admired; and each how he
 To be the inventor missed--so easy seemed
 Once found, which yet unfound most would have thought


SIR WM. PREECE, lately the distinguished engineer-in-chief of our postal telegraphs, has made the subject of wireless telegraphy a special study for many years, his first experiment dating back to 1882. 79 From that year up to the present he has experimented largely in all parts of the country, and has given us the results in numerous papers--so numerous, in fact, that they offer a veritable embarras des richesses to the historian. In what follows I can only attempt a résumé, and that a condensed one; but to the reader greatly interested in the subject I would advise a careful study of all the papers, a list of which I append:--
  1. Recent Progress in Telephony: British Association Report, 1882.
  2. On Electric Induction between Wires and Wires: British Association Report, 1886.
  3. On Induction between Wires and Wires: British Association Report, 1887.
  4. On the Transmission of Electric Signals through Space: Chicago Electrical Congress, 1893.
  5. Electric Signalling without Wires: Journal of the Society of Arts, February 23, 1894.
  6. Signalling through Space: British Association Report, 1894.
  7. Telegraphy without Wires: Toynbee Hall, December 12, 1896.
  8. Signalling through Space without Wires: Royal Institution, June 4, 1897.
  9. Ætheric Telegraphy: Institution of Electrical Engineers, December 22, 1898.
  10. Ætheric Telegraphy: Society of Arts, May 3, 1899. 80
    In his first-quoted paper of 1882, speaking of disturbances on telephone lines, Sir William says: "The discovery of the telephone has made us acquainted with many strange phenomena. It has enabled us, amongst other things, to establish beyond a doubt the fact that electric currents actually traverse the earth's crust. The theory that the earth acts as a great reservoir for electricity may be placed in the physicist's waste-paper basket, with phlogiston, the materiality of light, and other old-time hypotheses. Telephones have been fixed upon a wire passing from the ground floor to the top of a large building (the gas-pipes being used in place of a return wire), and Morse signals, sent from a telegraph office 250 yards distant, have been distinctly read. There are several cases on record of telephone circuits miles away from any telegraph wires, but in a line with the earth terminals, picking up telegraphic signals; and when an electric-light system uses the earth, it is stoppage to all telephonic communication in its neighbourhood. Thus, communication on the Manchester telephones was not long ago broken down from this cause; while in London the effect was at one time so strong as not only to destroy all correspondence, but to ring the telephone-call bells. A telephone system, using the earth in place of return wires, acts, in fact, as a shunt to the earth, picking up the currents that are passing in proportion to the relative resistances of the earth and the wire." 81 Fig. 17
    He then describes the experiment which he had recently (March 1882) made of telegraphing across the Solent, from Southampton to Newport in the Isle of Wight, without connecting wires. "The Isle of Wight," he says, "is a busy and important place, and the cable across at Hurst Castle is of consequence. For some cause the cable broke down, and it became of great importance to know if by any means we could communicate across, so I thought it a timely opportunity to test the ideas that had been promulgated by Prof. Trowbridge. I put a plate of copper, about 6 feet square, in the sea at the end of the pier at Ryde (fig. 17). A wire (overhead) passed from there to Newport, and thence to the sea at Sconce Point, where I placed another copper plate. Opposite, at Hurst Castle, was a similar plate, connected with a wire which ran through Southampton to Portsmouth, and terminated in another plate in the sea at Southsea Pier. We have here a complete circuit, if we include the water, starting from Southampton to Southsea Pier, 28 miles; across the sea, 6 miles; Ryde through Newport to Sconce Point, 20 miles; across the water again, l¼ mile; and Hurst Castle back to Southampton, 24 miles.
    "We first connected Gower-Bell loud-speaking telephones in the circuit, but we found conversation was impossible. Then we tried, at Southampton and Newport, what are called buzzers (Theiler's Sounders)--little instruments that make and break the current very rapidly with a buzzing sound, and for every vibration send a current into the circuit. With a buzzer, a Morse key, and 30 Leclanché cells at Southampton, it was quite possible to hear the Morse signals in a telephone at Newport, and vice versâ. Next day the cable was repaired, so that further experiment was unnecessary." 82
    Preece, however, kept the subject in view, and in 1884 he began a systematic investigation, theoretically and experimentally, of the laws and principles involved--an investigation which he has hardly yet completed. In his papers read at the International Electrical Congress, Chicago, August 23, 1893, and at the Society of Arts, London, February 23, 1894, he gives a résumé of his experiments from 1884 to date.
    He begins the latter paper by asking the same momentous question which a lady once put to Faraday, What is electricity? Faraday, with true philosophic caution, replied (I quote from memory) "Had you asked me forty years ago, I think I would have answered the question; but now, the more I know about electricity, the less prepared am I to tell you what it is." Sir William is not quite so epigrammatic, nor nearly so cautious; but, then, we have learned a great deal since Faraday's time. "Few," he says, "venture to reply boldly to this question--first, because they do not know; secondly, because they do not agree with their neighbours, even if they think they know; thirdly, because their neighbours do not agree among themselves, even as to what to apply the term. 83 The physicist applies it to one thing, the engineer to another. The former regards his electricity as a form of ether, the latter as a form of energy. I cannot grasp the concept of the physicist, but electricity as a form of energy is to me a concrete fact. The electricity of the engineer is something that is generated and supplied, transformed and utilised, economised and wasted, meted out and paid for. It produces motion of matter, heat, light, chemical decomposition, and sound; while these effects are reversible, and sound, chemical decomposition, light, heat, and motion reproduce those effects which are called electricity."
    In experiments of this kind it is necessary to point out that if we have two parallel conductors, separated from each other by a finite space, and each forming part of a separate and distinct circuit, either wholly metallic or partly completed by the earth, and called respectively the primary and the secondary circuits, we may obtain currents in the latter either by conduction or by induction; and we may classify them into those due to:--
    1. Earth-currents or leakages.
    2. Electro-static induction currents.
    3. Electro-magnetic induction currents.
It is very important to eliminate (1), which is a case of conduction, from (2) and (3), which are cases of induction, pure and simple.

1. Earth-currents or Leakages.

    When a linear conductor dips at each end into the earth, and voltage is impressed upon it by any means, the resulting return current would probably flow through the earth in a straight line between these two points if the conductibility of the earth were perfect; but as the earth, per se, is a very poor conductor (and probably is so only because it is moist), lines of current-flow spread out symmetrically in a way that recalls the figure of a magnetic field. These diffused currents are evident at great distances, and can be easily traced by means of exploring earth-plates or rods. The primary current is best produced by alternating currents of such a frequency as to excite a distinct musical note in a telephone, and if these currents rise and fall periodically and automatically, they produce an unmistakable wail, which, if made and broken by a Morse key into short and long periods, can be made to represent the dots and dashes of the Morse alphabet. The secondary circuit, which contains the receiving telephone, is completed in the case of an earth area by driving two rods into the ground, and in the case of water by dipping plates therein, 5 to 10 yards apart.
    It is therefore necessary to be able to distinguish these earth-currents from those due to induction, as they are apt to give false effects, and to lead to erroneous conclusions. This is easily done, if the instrument be sensitive enough, by making the primary current continuous when the earth-current also becomes continuous, whereas the induction currents will be momentary, and will only be observed at the beginning and end of the primary or inducing current.

2. Electro-static Induction Currents.
Fig. 18
Fig. 19

    When a body, A, is electrified by any means and isolated in a dielectric, as air, it establishes an electric field about it; and if in this field a similar body, B, be placed, it also is electrified by induction. If B be placed in connection with the earth, or with a condenser, or with any very large body, a charge of the same sign as A is conveyed away, and it (B) remains electrified in the opposite sense to A. A and B are now seats of electric force or stress. The dielectric between them is displaced or, as we say, polarised--that is, it is in a state of electric strain, and remains so as long as A remains charged; but if A be discharged or have its charge reversed or varied, then similar changes occur in B, and in the dielectric separating them. A may be an extended wire forming part of a complete primary circuit, and its charge may be due to a battery or other source of electricity; then, in the equally extended secondary wire B (fig. 18), the displaced charge in flowing to earth establishes a momentary current whose direction and duration depend on the current in A, and on its rate of variation.
    The strained (polarised) state of the dielectric, and the charges on A and B, remain quiescent so long as the current flows steadily; but when it ceases we have again, and in both circuits, momentary currents, as shown by the arrows (fig. 19), which flow until equilibrium is restored.
    The secondary currents due to discharge, like those due to charge, flow in opposite directions at each end, and there is always some intermediate zero point.
    It is thus easy in long circuits, by observing their direction, to differentiate currents of induction due to electro-static displacement from those due to electro-magnetic disturbance.
    The effects of electro-static induction do not play an important part in the inquiry immediately before us, but they are of great consequence in questions of speed of signalling in submarine cables and long, well-insulated land-lines, and in clearness of speech in long-distance telephony. 84

3. Electro-magnetic Induction Currents.

    Magnetic force is that which produces, or tends to produce, polarisation in magnetisable matter (as iron, nickel, cobalt), and electro-magnetic disturbance or stress in non-magnetisable matter and the ether. An electric current in a conductor is a seat of magnetic force, and establishes in its neighbourhood a magnetic field. The lines of force in this field are equivalent to circles in a plane perpendicular to the direction of the current, which circles, during the rise of the current, flow outwards or expand, and, during the fall of the current, flow inwards or contract, much like the waves on the surface of smooth water when a pebble is thrown in, but moving with the speed of light Thus any linear conductor placed in the field of another parallel conductor carrying a current is cut at right angles to itself by these lines of force--in one direction as the current rises, and in the opposite direction as the current falls. This outward and inward projection of magnetic force through such linear conductor excites electric force in that conductor, and if it form part of a circuit an electric current is set up in that circuit.
    So far for the theory of the subject. Now for its experimental elucidation. Besides those cases of interference mentioned, others were of frequent occurrence in the experience of the postal-telegraph officials, the most striking being that known as the Gray's Inn Road case. In 1884 it was there noticed that messages sent in the ordinary way through insulated wires, buried in iron pipes along the road, could be read upon telephone circuits erected on poles on the house-tops 80 feet high. To cure the evil the telegraph wires had to be taken up and removed to a more distant route. 85
    In 1885 Preece arranged an exhaustive series of experiments in the neighbourhood of Newcastle, which were ably carried out by Mr A. W. Heaviside, to determine whether these disturbances were due to electro-magnetic induction, and were independent of earth conduction; and also to find out how far the distance between the wires could be extended before this influence ceased to be evident. Insulated squares of wire, each side being 440 yards long, were laid out horizontally on the ground one quarter of a mile apart, and distinct speech by telephones was carried on between them; while when removed 1000 yards apart inductive effects were still appreciable.
    With the parallel lines of telegraph, ten and a quarter miles apart, between Durham and Darlington, the ordinary working currents in the one were clearly perceptible in a telephone on the other. Even indications were obtained in this way between Newcastle and Gretna, on the east and west coasts, forty miles apart; but here the observations were doubtless vitiated by conduction or leakage through the large network of telegraph wires between those two places. 86
    The district between Gloucester and Bristol, along the banks of the Severn, was next (1886) selected, where for a length of fourteen miles, and an average distance apart of four and a half miles, no intermediate disturbing lines existed. Complete metallic circuits were employed, the return wires passing far inland, in the one case through Monmouth, and in the other through Stroud. In one wire currents of about ·5 ampère were rapidly made and broken by mechanical means, producing on a telephone a continuous note which could be broken up by a Morse key into dots and dashes, as in Cardew's vibrator. Weak disturbances were detected in the secondary circuit, showing that here the range of audibility with the apparatus in use was just overstepped. The unexpected fact was also shown in these experiments that, whether the circuits were entirely metallic or earthed at the ends, the results were the same. 87
    Similar trials were made on lines along the valley of the Mersey. A new trunk line of copper wires that was being erected between London and the coast of North Wales was then experimented upon, and some interesting results were obtained in the district between Shrewsbury and Much Wenlock, and between Worcester and Bewdley.
    In the autumn of the same year (1886) some admirable results were obtained by Mr Gavey, another of Preece's able assistants, near Porthcawl, in South Wales--a wide expanse of sand well covered by the tide, thus giving the opportunity of observing the effects in water as well as in air. Two horizontal squares of insulated wire, 300 yards each side, were laid side by side at various distances apart up to 300 yards, and the inductive effects of one on the other noted. Then one coil was suspended on poles 15 feet above the other, which was covered with water at high tide. No difference was observable in the strength of the induced signals, whether the intervening space was air or water or a combination of both, although subsequent experience (1893) showed that with a space of 15 feet the effect in air was distinctly better than through water.
    The conclusion drawn from all these experiments was that the magnetic field extends uninterruptedly through the earth, as it does through the air; and that if the secondary circuit had been in a coal-pit the effect would be the same. In fact, Mr Arthur Heaviside succeeded in 1887 in communicating between the surface and the galleries of Broomhill Colliery, 350 feet deep. He arranged a circuit in a triangular form along the galleries about two and a quarter miles in total length, and at the surface a similar circuit of equal size over and parallel to the underground line. Telephonic speech was easily carried on by induction from circuit to circuit. 88
    As the result of all these experiments and innumerable laboratory investigations, Preece deduced the following formulæ. The first shows the strength of current C2 induced in the secondary circuit by a given current C1 in the primary circuit--
equation 89
where R equals the resistance of the secondary circuit, D the distance apart of the two circuits, L the length of the inductive system, and I the inductance of the system. The value of I, obtained by experiment on two parallel squares of wire, 1200 yards round and 5 yards apart, was found to be ·003.
    The second equation gives approximately the maximum distance X which should separate any two wires of length L, C1 being the primary current and R the resistance of the secondary circuit.
The constant 1·9016 was obtained by experimenting on two parallel wires, each one mile long, when the primary circuit, being excited by one ampère, the limit of audibility in the secondary was reached at 1·9016 miles. This formula shows the desirability of using copper wires of the largest size practicable, so as to reduce the value of R. Other very important elements of success are (1) the rate at which the primary currents rise and fall, the faster the better, and (2) the reduction to a minimum of such retarding causes as capacity and self-induction.
    Having thus threshed out the laws and conditions of electro-magnetic disturbances, and determined the distance at which they could be usefully applied, it only remained for Sir William to put his conclusions to a practical test. Accordingly, when the Royal Commission on electric communication between the shore and lighthouses and lightships was appointed in June 1892, he made his proposals to the Government, and on receiving sanction forthwith proceeded to carry them out. Fig. 20
    The Bristol Channel proved a very convenient locality to test the practicability of communicating across distances of three and five miles without any intermediate conductors. Two islands, the Flat Holm and the Steep Holm, lie off Penarth and Lavernock Point, near Cardiff, the former having a lighthouse upon it (fig. 20). On the shore two thick copper wires combined in one circuit were suspended on poles for a distance of 1267 yards, the circuit being completed by the earth. On the sands at low-water mark, 600 yards from this primary circuit and parallel to it, two gutta-percha covered copper wires and one bare copper wire were laid down, their ends being buried in the ground by means of bars driven in the sand.
    One of the gutta-percha wires was lashed to an iron wire to represent a cable. These wires were periodically covered by the tide, which rises here at spring to 33 feet. On the Flat Holm, 3·3 miles away, another gutta-percha covered copper wire was laid for a length of 600 yards.
    There was also a small steam launch having on board several lengths of gutta-percha covered wire. One end of such a wire, half a mile long, was attached to a small buoy, which acted as a kind of float to the end, keeping the wire suspended near the surface of the water as it was paid out while the launch slowly steamed ahead against the tide. Such a wire was paid out and picked up in several positions between the primary circuit and the islands.
    The apparatus used on shore was a 2-h.p. portable Marshall's engine, working a Pyke and Harris's alternator, sending 192 complete alternations per second of any desirable strength up to a maximum of 15 ampères. These alternating currents were broken up into Morse signals by a suitable key. The signals received on the secondary circuits were read on a pair of telephones--the same instruments being used for all the experiments.
    The object of the experiments was not only to test the practicability of signalling between the shore and the lighthouse, but to differentiate the effects due to earth conduction from those due to electro-magnetic induction, and to determine the effects in water. It was possible to trace without any difficulty the region where they ceased to be perceptible as earth-currents and where they commenced to be solely due to electro-magnetic waves. This was found by allowing the paid-out cable, suspended near the surface of the water, to sink. Near the shore no difference was perceptible, whether the cable was near the surface or lying on the bottom, but a point was reached, just over a mile away, where all sounds ceased as the cable sank, but were received again when the cable came to the surface. The total absence of sound in the submerged cable was rather surprising, and led to the conclusion either that the electro-magnetic waves of energy are dissipated in the sea-water, which is a conductor, or else that they are reflected away from the surface of the water, like rays of light. 90
    Experiments on the Conway Estuary, showing the relative transparency of air and water to these electro-magnetic waves, tend to support the latter deduction; for if much waste of energy took place in the water, the difference would be more marked. As it is, there seems to be ample evidence that the electro-magnetic waves are transmitted to considerable distances through water, though how far remains to be found. 91
    There was no difficulty in communicating between the shore and Flat Holm, 3·3 miles. The attempt to speak between Lavernock and Steep Holm, 5·35 miles, was not so successful: though signals were perceptible, conversation was impossible. There was distinct evidence of sound, but it was impossible to differentiate the sounds into Morse signals. If either line had been longer, or the primary currents stronger, signalling would probably have been possible.
    In 1894 Preece carried out some satisfactory experiments near Frodsham on the estuary of the Dee, which was found to be a more convenient locality than the Conway sands. Here, as at Conway and other places, squares and rectangles were formed of insulated wires, and numerous measurements were made (with reflecting galvanometers and telephones) of the effects due to varying currents in the primaries, and at varying distances between them and the secondaries.
    In Scotland also some very successful trials were made. There happens to be a very convenient and accessible loch in the Highlands--Loch Ness--forming part of the route of the Caledonian Canal between Inverness and Banavie, having a line of telegraph on each side of it. Five miles on each side of this loch were taken, and so arranged that any fractional length of telegraph wire on either side could be taken for trial. Ordinary, and not special, apparatus was employed. Sending messages, as before, by Morse signals and speaking by telephone across a space of one and a quarter mile was found practical, and, in fact, easy; indeed, the sounds were so loud that they were found sufficient to form a call for attention. Fig. 21
    The following apparatus was in use on each side of the loch: A set of batteries consisting of 100 dry cells, giving a maximum voltage of 140; a rapidly revolving rheotome, which broke up the current into a musical note; a Morse key, by which these musical notes could be transformed into Morse signals; resistance coils and ampère-meters to vary the primary current; two Bell telephones joined in multiple arc to act as receivers. For the transmission of actual speech simple granular carbon microphones, known as Deckert's, were used as transmitters, and a current of two ampères was maintained through these and two Bell telephones in circuit with the line wire.
    Any lingering fear that earth conduction had principally to do with these results was removed by making the earth's terminals on the primary circuit at one end at Inverness nine miles away, and at the other end in two directions in a parallel glen about six miles away.
    One very interesting fact observed at Loch Ness was that there was one particular frequency in the primary circuit that gave a decided maximum effect upon the telephones in the secondary circuit. This confirms the presence of resonance, and is, of itself, a fact sufficient to prove the effects as being due to the transformation of electro-magnetic waves into electric currents. 92
    During the same year (1894) experiments were carried out between the island of Arran and Kintyre across Kilbrannan Sound. Two parallel lines on opposite sides, and four miles apart, were taken (fig. 21); and, in addition, two gutta-percha covered wires were laid along each coast, at a height of 500 feet above sea-level and five miles apart horizontally.
    Incidentally some extremely interesting effects of electro-magnetic resonance were observed during the experiments in Arran. A metallic circuit was formed partly of the insulated wire 500 feet above the sea-level and partly of an ordinary line wire, the rectangle being two miles long and 500 feet high. Wires on neighbouring poles, at right angles to the shorter side of the rectangle, although disconnected at both ends, took up the vibrations, and it was possible to read all that was signalled on a telephone placed midway in the disconnected circuit by the surgings thus set up.
    The general conclusions arrived at as the result of these numerous and long-continued experiments may be briefly summed up as follows: 93-- Fig. 22
    The earth acts simply as a conductor, and per se it is a very poor conductor, deriving its conducting property principally, and often solely, from the moisture it contains. On the other hand, the resistance of the "earth" between the two earth plates of a good circuit is practically nothing. Hence it follows that the mass of earth which forms the return portion of a circuit must be very great, for we know by Ohm's law that the resistance of a circuit increases with its specific resistance and length, and diminishes with its sectional area. Now, if the material forming the "earth" portion of the circuit were, like the sea, homogeneous, the current-flow between the earth plates would follow innumerable but definite stream lines, which, if traced and plotted out, would form a hemispheroid. These lines of current have been traced and measured. A horizontal plan on the surface of the earth is of the form illustrated in fig. 22, while a vertical section through the earth is of the form shown in fig. 23.
    With earth plates 1200 yards apart these currents have been found on the surface at a distance of half a mile behind each plate; and, in a line joining the two transversely, they are evident at a similar distance at right angles to this line.
    Now this hemispheroidal mass could be replaced electrically by a resultant conductor (R, fig. 23) of a definite form and position; and, in considering the inductive action between two circuits having earth returns, it is necessary to estimate the position of this imaginary conductor. This was the object of the experiments at Frodsham. Fig. 23
    If the material of the earth be variable and dry the hemispheroid must become very much deformed and the section very irregular: the lines of current-flow must spread out farther, but the principle is the same, and there must be a resultant return. The general result of the experiments at Frodsham indicates that the depth of the resultant earth was 300 feet, while those at Conway are comparable with a depth of 350 feet. In the case of Frodsham the primary coil had a length of 300 feet, while at Conway the length was 1320 feet. At Loch Ness, and between Arran and Kintyre, where the parallel lines varied from two to four miles, the calculated depth was found to be about 900 feet. The depth of this resultant must, therefore, increase with the distance separating the earth plates, and this renders it possible to communicate by induction from parallel wires over much longer distances than would otherwise be possible.
    The first and obvious mode of communicating across space is by means of coils of wire opposed to each other in the way familiar to us through the researches of Henry and Faraday. All the methods here described consisted in opposing two similar coils of wire having many turns, the one coil forming the primary circuit and the other coil the secondary circuit.
    Vibratory or alternating currents of considerable frequency were sent through the primary circuit, and the induced secondary currents were detected by the sound or note they made on a telephone fixed in the secondary circuit.
    The distance to which the effective field formed by a coil extends increases with the diameter of the coil more than with the number of turns of wire upon it. A single wire stretched across the surface of the earth, forming part of a circuit completed by the earth, is a single coil, of which the lower part is formed by the resultant earth return, and the distance to which its influence extends depends upon the height of the wire above the ground and the depth of this resultant earth.
    In establishing communication by means of induction, there are three dispositions of circuit available--viz., (a) single parallel wires to earth at each extremity; (b) parallel coils of one or more turns; (c) coils of one or more turns placed horizontally and in the same plane.
    The best practical results are obtained with the first arrangement, more especially if the conformation of the earth admits of the wires being carried to a considerable height above the sea, whilst the earth plates are at the sea-level. By adopting this course the size of the coil is practically enlarged, and even if it be necessary to increase the distance between the parallel wires in order to get a larger coil, the result is still. more beneficial. In a single-wire circuit we have the full effect of electro-static and electro-magnetic induction, as well as the benefit of any earth conduction, but in closed coils we have only the electro-magnetic effects to utilise.
    In one experiment two wires of a definite length were first made up into two coils forming metallic circuits, then uncoiled and joined up as straight lines opposed to each other, with the circuit completed by earth. The effects, and the distance between which they were observable, were very many times greater with the latter than with the former arrangement.
    The general law regulating the distance to which we can speak by induction has not been rigorously determined, and it is hardly possible that it can be done, owing to the many disturbing elements, geological as well as electrical. In practice we have to deal with two complete circuits of unknown shape, and in different planes. We have obtained some remarkably concordant and accurate results in one place; but, on the other hand, we have met with equally discordant results in another place. Still, from the approximate formula before given, we deduce the important fact that for parallel lines the limiting distance increases directly as the square of the length, which shows that if we can make the length of the two lines long enough it would be easy to communicate across a river or a channel. Of course, as previously pointed out, the formula does not take into account the effects of the reverse magnetic waves generated by the return current through the earth, and at present no data exist on which a satisfactory calculation can be based; but, for example, there is little doubt that two wires, ten miles long, would signal through a distance of ten miles with ease.
    "Although," says Sir William in conclusion, "communication across space has thus been proved to be practical in certain conditions, those conditions do not exist in the cases of isolated lighthouses and light-ships, cases which it was specially desired to provide for. The length of the secondary must be considerable, and, for good effects, at least equal to the distance separating the two conductors. Moreover, the apparatus to be used on each circuit is cumbrous and costly, and it may be more economical to lay an ordinary submarine cable.
    "Still, communication is possible even between England and France, across the Channel, and it may happen that between islands where the channels are rough and rugged, the bottom rocky, and the tides fierce, the system may be financially possible. It is, however, in time of war that it may become useful. It is possible to communicate with a beleaguered city either from the sea or on the land, or between armies separated by rivers, or even by enemies.
    "As these waves are transmitted by the ether, they are independent of day or night, of fog, or snow, or rain, and therefore, if by any means a lighthouse can flash its indicating signals by electro-magnetic disturbances through space, ships could find out their positions in spite of darkness and of weather. Fog would lose one of its terrors, and electricity become a great life-saving agency."
    At the Society of Arts (February 23, 1894), Sir William gave rein to his imagination, and, looking beyond these mundane utilities, concluded his address with the following magnificent peroration:--
    "Although this short paper is confined to a description of a simple practical system of communicating across terrestrial space, one cannot help speculating as to what may occur through planetary space. Strange mysterious sounds are heard on all long telephone lines when the earth is used as a return, especially in the calm stillness of night. Earth-currents are found in telegraph circuits and the aurora borealis lights up our northern sky when the sun's photosphere is disturbed by spots. The sun's surface must at such times be violently disturbed by electrical storms, and if oscillations are set up and radiated through space, in sympathy with those required to affect telephones, it is not a wild dream to say that we may hear on this earth a thunderstorm in the sun.
    "If any of the planets be populated with beings like ourselves, having the gift of language and the knowledge to adapt the great forces of nature to their wants, then, if they could oscillate immense stores of electrical energy to and fro in telegraphic order, it would be possible for us to hold commune by telephone with the people of Mars."
    The first application of Preece's system to the ordinary needs of the postal-telegraph service was made on March 30, 1895, when the cable between the Isle of Mull and Oban, in Scotland, broke down. As there was no ship available at the moment for effecting repairs, communication was established by laying a gutta-percha-covered copper wire, one and a half mile long, along the ground from Morven, on the Argylishire coast, while on Mull the ordinary telegraph (iron) wire connecting Craignure with Aros was used, the mean distance separating the two base lines being about two miles. No difficulty was experienced in keeping up communication, and many public messages were transmitted for a week until the cable was repaired. In all about 160 messages were thus exchanged, including a press telegram of 120 words.
Fig. 24
    The diagram (fig. 24) shows the apparatus and connections, as regards which it is only necessary to say that a is a rheotome, or make-and-break wheel, driven so as to produce about 260 interruptions of the current per second, which give a pleasant note in the telephone, and are easily read when broken up by the key into Morse dots and dashes; b is a battery of 100 Leclanché cells, of the so-called dry and portable type; c is a switch to start and stop the rheotome as required; and d is a telephone to act as receiver.
    Since March 1898 this system has been permanently established for signalling between Lavernock Point and the Flat Holm, and has been handed over to the War Office. Permanent lines of heavy copper wire have been erected parallel to each other, one being on the Flat Holm and the other on the mainland.
    The heavy and cumbrous Pyke and Harris alternator of the earlier experiment over the same line (p. 149, ante) has been replaced by 50 Leclanché cells. The frequency has been raised to 400 makes and breaks per second, thus greatly increasing the strength of the induced currents. By the use of heavy copper base lines the resistances have been made as low as practicable. There is no measurable capacity, self-induction is eliminated, and there is no impedance. Hence the signals are perfect, and the rate of working is only dependent on the skill of the operator. It is said that as many as 40 words per minute have been transmitted without the necessity for a single repetition--a speed which few telegraphists can achieve, and still fewer can keep up.
    A little later Mr Sydney Evershed's relays were added to work a call-bell, which was the only thing wanted to make the system complete and practical94

    79 Indeed, it so happens that one of the first experiments he ever made in electricity was on this very subject in 1854. See supra.
    80 This list does not pretend to be complete. Doubtless there are other papers, which have escaped my notice.
    81 For early notices of the same kind, see supra.
    82 Captain (now Colonel) Hippisley, R. E., who conducted these trials thought that the presence of the broken cable across the Solent somewhat vitiated the results, as its heavy iron sheathing may have aided in conducting the current.
    83 "Substantialists" call it a kind of matter. Others view it as a form of energy. Others, again, reject both these views. Prof. Lodge considers it a form, or rather a mode of manifestation, of the ether. Prof. Nikola Tesla demurs to this view, but sees no objection to calling electricity ether associated with matter, or bound ether. High authorities cannot even yet agree whether we have one electricity or two opposite electricities.--Sir W. Crookes, 'Fortnightly Review,' February 1892.
    84 For an interesting investigation of electro-static phenomena on telephone circuits, see Mr Carty's papers in the 'Electrician,' December 6, 1889, and April 10, 1891.
    85 The following are more recent cases of the same kind. Currents working the City and South London Electric Railway affect recording galvanometers at the Greenwich Observatory, four and a half miles distant; and even a diagram of the train service could be made out by tapping any part of the metropolitan area.
    Some ten years ago one of the dynamos at the Ferranti electric. light station at Deptford by some accident got connected to earth, with the result that the whole of the railway telegraphs in the signal-boxes of the railways in South London were temporarily put out of order and rendered inoperative, while the currents flowing in the earth were perceived in the telegraph instruments so far northwards as Leicester and so far south as Paris.
    86 British Association Report, 1886.
    87 These experiments were repeated with more experience and greater success in 1889.
    88 Subsequent experiments showed that the conclusion arrived at for earth and air was only partially true for water. Telephonic speech was carried on in Dover Harbour through 36 feet of water, but no practical signals could be obtained through 400 feet at North Sand Head, Goodwin Sands, showing that the effect must diminish in water with some high power of the distance.
    89 This formula does not allow for the reverse effect of the return current through the earth, as to which no data exist at present.
    90 See note, infra.
    91 See note 88, supra.
    92 This is still a moot question, many competent authorities, as Lodge, Rathenau, W. S. Smith, and Stevenson, being of opinion that the effect is partly inductive and partly conductive. See Dr Lodge's contention, 'Jour. Inst. Elec. Engs.,' No. 137, p. 814.
    93 British Association Report, 1894, Section G.
    94 During the summer of 1899 Sir William began a new series of experiments on wireless telephony at the Meuai Straits, the results of which he communicated recently to the British Association (Bradford, September 8, 1900). After referring to his Loch Ness experiments (ante), where telephonic signals were found possible across an average space of 1·3 miles with parallel base lines of 4 miles each, Sir William states that his new experiments fully bore out this fact, and determined the further fact that maximum effects are obtained when the parallel wires are terminated by earth-plates in the sea itself--showing that the inductive effects through the air are enhanced by conductive effects through the water, and that, consequently, shorter base lines are permissible. Ordinary telephone transmitters and receivers were used.
    This new method has been successfully applied to establishing communication between the Skerries and Cemlyn, Anglesey, across 2·8 miles average distance, and between Rathlin Island and the Irish coast, about 4 miles across.
TOC | Previous Section: Professor Erich Rathenau--1894 | Next Section: Willoughby Smith's Method