needed, inasmuch as the current does not pass through the sender's Morse. In the position shown, the keys, kept in place by a spring. make contact at 3 and 4, so that each station is ready to receive a message. On either key being depressed a current is sent along

G

K

B

M

M'

the line and through the other instrument; so that dots or dashes will there be printed, according to the duration of the contact made by the sending key.

If both keys be depressed at once no message can pass. Α preliminary signal is usually sent, so that the receiving clerk may set going the clock-work that moves the paper.

the

The Morse is often worked by sound, the ear soon acquiring power of reading off the message in this way.

§ 6. Relays. On long lines the current will usually be too feeble to work the printing Morse. In such cases a relay is employed. This is an arrangement consisting of a local battery in circuit with the Morse, and a key that can make or break this circuit; this key being worked by the feeble line current through the intermediency of an electro-magnet.

In the figure on the next page the feeble line current works the electro-magnet R; this acts upon the soft iron piece a, and by means of the lever makes contact at 1, 2; the local circuit is thus completed, and the Morse S is worked by the strong local current.

§ 7. Transmission through Cables, under Water. When the line connecting two places has to be laid under water, it then consists of one or more copper wires thickly coated with some insulating material; round this again is generally twisted stout iron wire in order to give the strength necessary, and this is coated

with some waterproof material. The whole is called a cable; the messages conveyed by it are sometimes called cablegrams.

Difficulties occurring in transmission by cable.-It was soon found that it was not possible to send, through a cable of any considerable length, the clearly separated dots and dashes required by the Morse receiver. The sharp tap on the key at the sending end is represented at the receiving end by a feeble current that slowly rises to a maximum and again slowly sinks to zero. With such currents we cannot use any of the instruments described

above. Had not the Thomson's reflecting galvanometer, or some similar instrument, been invented, cablegraphy would hardly have been practicable. With this instrument, however, the reading of a message is easy. The principle of the reflected beam of light gives the means of magnifying the motion to any desired extent. The needle is of very small mass and possesses but little moment of inertia about its axis of suspension; and hence its deflexion will follow, with but an inappreciable lag, the varying undulatory current. And the extreme delicacy of the instrument enables very

small currents to be perceived. Finally, the small mass of the needle, its method of suspension, and the command of it that is given by the controlling magnet, renders it possible to use a modified form of this instrument on board ship during the laying of the line.

Reading the messages.—On a land line we are able to give to the needle very rapid deflexion to right and to left of the zero mark. But on a cable the currents last so long that it is not practicable to wait for the dying away of one current before sending another. Hence the currents overlap one another, and allowance has to be made for this in reading the message. The clerk at the receiving end watches the spot of light, and observes, not its movements to right or left of the zero mark in the middle of the scale, but its jerks to right or left wherever it happens to be. In fact, a shifting zero mark is used.

Cause of retardation of signals.—The origin of the alteration of the sharp signals sent, into the slow and gradual undulations received, lies mainly in the fact that the cable forms a long condenser of very great capacity.

In the case of an overland wire, the sharp making of contact effected by striking the key is followed by the passage along the wire of what we may call a tide of electricity; this tide having a relatively abrupt front, as has the well-known 'bore' that is sometimes seen on a tidal river. The receiving instrument is affected, almost instantaneously, with the full force of the current. But in the case of the cable matters are very different. The tide of electricity rushes into the cable, but as it proceeds it is to a great extent detained to charge electrostatically the cable; this being a condenser in which the wire and the sea-water form two coatings separated by the insulating di-electric. Thus the rise to a maximum occurs very gradually, as would the rise of tide in a stream out of which ran many side channels that must be filled by the tide before it can advance up the river. (We need not say that the analogies here given are very incomplete.)

Hence at the receiving instrument the current rises very slowly to a maximum.

§ 8. Earth Currents. Condenser System of Working. It is found that currents are continually flowing from one point of the earth to another, these currents being due to thermo-electric

or to other unknown causes. When the two earth plates of a telegraphic line happen to be buried where the earth is not at one potential, a current will flow along the line wire. The greater the AV of the two places at which the plates are sunk, the stronger will be the E.M.F. driving the current along the line. During those disturbances of the earth's magnetic field which are known as magnetic storms, these earth currents become very strong. and may entirely over-ride and nullify the ordinary currents by which the messages are transmitted; such 'storms' occur, for example, during brilliant displays of the aurora borealis, and during any noticeable changes in the surface of the sun.

These earth currents give great trouble on long lines, and especially in long cables. There are two methods of meeting the difficulty.

(i.) The return wire method.-Where it is practicable, a return wire may be used instead of an 'earth.' If the clerks find the earth current to be so strong on any particular day as to give trouble, they may temporarily complete the circuit by means of any wire that does not happen to be in use on that day.

(ii.) The condenser system.-The diagram here given represents the principle of the condenser method. In this system the line is

K

G

B

not continuous, but the line from the sending station is connected with one condenser plate, while from the other plate passes away the line to the receiving station. In the simple arrangement shown in the figure, a depression of the key would send a charge into the condenser plate a. This would call up an equal and opposite charge in the plate b; and the effect upon the instrument G would

This

be the same as though the current sent had passed on. action of course ceases when the condenser is charged, but, as this latter is always of very great capacity, the current due to the electrostatic induction lasts as long as is necessary.

When the key is again released, the plate a is discharged to earth; and hence there will be a discharge of b also, and therefore a reverse current through the instrument G.

In practice there is a condenser at each end, the cable being totally insulated. There is double induction, but, by the electrostatic principles discussed in Chapter X. and earlier, the instrument at the receiving end will be influenced by currents passing in the same direction as if they had come direct from the sending station,

The condensers employed consist usually of many sheets of tin-foil separated by insulating sheets of paraffined paper; sheets of tin-foil being connected up into two alternate sets, answering to the two plates of a condenser.

§ 9. Insulation of Wires.-If we cut connection with the earth at one end of a line, and send a current into the wire at the other end, a galvanometer in the line will settle down to some steady deflexion. In fact, the 'insulated' line wire leaks to earth along the whole distance; the insulators acting as conductors of very great resistance. The longer the line, the more is the leakage and the lower is the resistance offered by the insulators collectively. It is usual to fix upon some definite resistance (in ohms) as the resistance per mile that ought to be offered by any given line of wire. If, without any special cause, such as fogs, rain, or snow, the insulation fall below this standard, then it is suspected that some accidental connection with earth has been made.

§ 10. Duplex Telegraphy.-In §§ 4 and 5 we noticed that it was not possible to send a message from each station to the other respectively at one and the same time. Could this be done, it is evident that we could get twice the use out of a line.

We can readily devise a method in which each clerk could transmit his own message through his own instrument; the problem to be solved is much harder, viz. how the message sent from A shall be indicated by the instrument at B only, while that sent from B shall be indicated at A only.

This problem has been solved in two ways, one of which we will briefly describe.