difficult operation. It is a very different matter when the fault is a mere flaw in an otherwise perfect cable, but this is the principle. Tests are taken hourly during the manufacture of a cable, to determine that it is electrically sound throughout, and that the insulation is complete. To ascertain the latter point, that is to say, that there is no leakage of electricity through the guttapercha covering, the wire is first suddenly charged from a battery, and as suddenly discharged through a galvanometer. A certain deflection of the galvanometer needle is thereby produced, and noted as deflection No. 1. It is then charged a second time, and left for a few minutes, at the end of which it is again discharged through the galvanometer, and deflection No. 2 is noted; the difference between deflection No. 1 and deflection No. 2 corresponds to the amount lost by leakage during the time the wire remained charged.

It may easily be imagined that the tests are never more anxiously performed than when the cable, duly completed, and coiled in great tanks on board the telegraph ship, is being slowly paid out into the deep. Every minute signals are passed from the shore to the ship, through the gradually increasing length of submerged line, through all the thousands of coils which lie piled, tier above tier, in the cable tanks, down to the testing cabin. There the electrician on duty sits, with his mirror galvanometer, watching the reading-screen, where the movements of a small spot of light give him tidings of the shore he is leaving. The whole interest of a great and costly expedition is thus centred in the little quiet testing-room.

The operations, both of the ordinary working of submarine telegraphs and the testing operations, are much complicated by earth-currents. Very often, especially at the time when the phenomena of the Aurora Borealis are prevalent, strong currents pass over the lines, entering by one of the earth connections, and leaving it by the other. They are never constant for long together, and they change direction so rapidly as seriously to affect the delicate instruments in use on submarine lines. They are most violent during magnetic storms, which seem in some mysterious manner to be dependent upon the Aurora, which, in its turn, is a manifestation of some ultimate cause of which nothing positive is known. Observation shows that earthcurrents are frequent at the time of earthquakes; the Journal of the Society of Telegraph Engineers' contains constant notes of these coincidences, observed by the ever-watchful members of the Society. Many of the most distinguished electricians think that the sun is the ultimate cause of this, as of all other forms of terrestrial magnetism, and it certainly seems probable that such

is the case. It is well known that the period of maximum and minimum of spots on the sun extends over a cycle of eleven years, and during that time the Aurora becomes proportionately more or less intense. 'In 1850,' says Mr. Latimer Clarke, 'two simultaneous observations of the sun were made by observers many miles apart, when both saw a body flash into the sun and cause a disturbance of the sun's chromosphere; and it was subsequently found that at that moment almost all the magnets of the world were disturbed by this sudden movement.'* fact that the cause and effect were apparently simultaneous incidentally confirms the inference that the velocity of the transmission of magnetism is the same as the velocity of the transmission of light.

The

In

Sometimes the earth-currents are of enormous power. 1871, a break having occurred in the Atlantic Cable, Mr. Graves was able to devote a great deal of time to the investigation and tabulation of the earth-currents which appeared on the broken line. At one time during a great magnetic storm, which was felt all over the world, Mr. Graves, who was observing at Valentia, saw currents of such strength, that 'a distinct arc of flame burned between the key and the earth connection.' The power necessary to produce this he estimated at not less than 2000 cells of Daniell's battery. The late Admiral FitzRoy found the indication of coming storms predicted with singular fidelity by magnetic disturbances of the earth. He could sometimes see the approach of a storm days before the barometer and thermometer indicated anything of the kind. The ordinary observations of the telegraph operators confirm this. They can actually feel a storm coming across the Atlantic for days beforehand, by the increasing vagaries of their troublesome visitor, the earthcurrents. It is not improbable that observations may eventually be found susceptible of such generalisation as to afford really reliable weather forecasts.

One of the latest telegraphic marvels is the arrangement by which it is made possible to send two separate messages along the same wire in opposite directions at the same time. This is known as Duplex telegraphy; and perhaps nothing connected with the practical working of telegraphs has excited more wonder, and been found more difficult to understand. The first question which is naturally asked is, How can the currents pass one another in the line-wire? and, if they do pass, how is it that they do not interfere one with another? If, however, we have been fortunate enough to make clear the principle of the

* 'Journal of the Society of Telegraph Engineers,' iv. 121.

differential

differential galvanometer, the difficulty will vanish. In sending an ordinary message, the current passes from the battery at the sending end through the instrument at that end, along the line, through the instrument at the receiving end, and so through the earth back to the battery from which it started. In doing so it of course moves the instrument at the sending end, as well as the distant one; for it passes through them both. Now suppose that the operators at both ends were to dispute for the possession of the circuit, and send opposite currents simultaneously through the wire, the result would be hopelessly to confuse the signals, and make reading impossible; but if the operators were to set to work to unravel the apparent confusion, they would soon find that when station A and station B, in the course of their confused struggle for the possession of the line, happened to send a current in the same direction, the needle acted upon would strike the stops with double force, while, if they sent in opposite directions, the needle would hardly move at all. Thus each operator would be able to perceive that the signals of the other station were visible on his instrument, and were only prevented from being legible by the confusion introduced by the current he himself was sending along the wire. It would doubtless occur to them that if by any means each could so arrange that neither station's own or outgoing currents should affect his own needle, leaving the dial free to show only the effect produced by the incoming current, the difficulty of reading would vanish. The question then arises, How can each instrument be so connected that neither sender shall move his own needle, and yet so that the coils shall always remain in circuit?

Now, in a differential galvanometer, if two equal currents are simultaneously sent through the two coils in opposite directions, the result is that the needle stands still. Apply that principle to the line in such a way that the current, when either end makes a signal, shall at that end divide itself into two, and the two halves pass round the sending instrument in opposite directions. This will only happen when the two half-currents are exactly equal, which will only be the case if the two circuits they have to travel are equal. To effect this, one half-current must pass along the line-wire to earth, and the other half be sent to earth through a resistance exactly equal to the linewire.

When the balance is established, neither sender, when he signals, will move his own instrument, which will be left free to record signals from the opposite side; but the operator at the other end will be able to read them, for each station

will see the current sent by the other, though neither can see his own.*

It will be seen that the two currents do not pass one another, as has been imagined, but that, when both stations signal at the same time, the current sent by either station acts upon the distant instrument, by determining whether the currents sent by that station shall pass through the line or the resistance-coils.

On land-lines suspended in the air the resistance of the signalling-wire to the current is easily ascertained, and is easily imitated on the second or artificial circuit; but in submarine lines there is not only the resistance to be taken into account, but the retarding capacity of the cable. In any given cable each mile presents a certain resistance, and also a certain retarding capacity. The second circuit on land-lines need only imitate the resistance, but in cables the retarding capacity must also be imitated. It is not enough that the whole of the second circuit should be equal to the whole of the cable, but that each separate part of it should be equal to each corresponding part of the cable. The latest plan, which seems at length to have made duplex working in submarine wires practically possible, is that adopted by Mr. Muirhead. He forms his second circuit by sheets of paper, prepared with paraffin, as an insulator, having on one side a strip of tinfoil wound to and fro, to represent the resistance, and on the other a sheet of tinfoil to represent its

When station A signals separately, the current is divided at e, and its effect balanced in the coils a c, and b d, of the home instrument; but it passes through both coils of C in the same direction, entering at a', passing from c' to b' through the junction e', and to earth by d' and the resistance R, and therefore produces a signal. If both A and C depress their keys at the same moment, the two batteries are added to one another, as far as the line-wire is concerned.

retarding

retarding capacity. Each strip of paper may thus be made to represent precisely a given length of cable; and a given number of such sheets would exactly imitate the cable in every part of its length; so that the non-signalling half of the current sent through the artificial resistance escapes to earth under precisely similar conditions to that which passes over the line.

The result of this is that a single wire will convey signals simultaneously in two opposite directions, and that one wire will do the work of the two which have hitherto been required. On all marine lines this invention is of the greatest possible importance, because while theoretically it only doubles the carrying capacity of each cable, in practice it does a great deal more; as it does away with the loss of time consequent on arranging about the precedence of outward and homeward messages.

The method was first tried on the line between Marseilles and Bona, and it has since been brought into operation between Marseilles and Malta, between Suez and Aden, and, lastly, between Aden and Bombay. It is stated that on a recent occasion, when there was a breakdown of the Indo-European line, the duplex system became of the greatest possible use; and although there are still practical difficulties to be encountered before it can be adopted in lines where very long distances have to be accomplished without a break, there is no doubt that the theory is so well established that its universal adoption is only a question of time.

'The telegraph,' says Sir Lintorn Simmons, 'is an essential in war; war can scarcely be carried on without it.' Mr. George von Chauvin, who was secretary to the German Director-General of Telegraphs during the Franco-German war, appeared as a witness before a Committee of the House of Commons on Postal Telegraphs last year, and gave a very animated account of the way in which that service was performed by the German army during his period of service. He tells us that the telegraph was in constant use for the arrangement of the transport of ammunition; of the whole service of the commissariat; the transport of wounded soldiers and prisoners; for the regulation of traffic in the field railways, which was very heavy, and which frequently necessitated the shunting of ammunition trains to let a train of wounded soldiers go by, or stopping a train of soldiers to bring up ammunition. It was also used for the investiture of fortresses like Paris and Metz, where it would have been impossible to have an army large enough to girdle round the whole enceinte. The lines of attack round Paris extended some twenty German (above ninety English) miles; the field telegraph was used along this extended line to bring together