The function of these wires, it will be understood, is to form one plate of a condenser, the earth being the other plate and the air in between the "dielectric" or insulator.

In the case of ships "the earth" is represented by the hull of the vessel. It makes a particularly good "earth" since it is in perfect contact with a vast mass of salt water, and that again is in contact with a vast area of the earth's surface. Salt water is a surprisingly good conductor of electricity.

In land stations "earth" consists of a metal plate well buried in damp ground. The whole question of conduction of electricity through the earth is very perplexing. There seems to be resistance offered to the current at the point where it enters the ground, but after that none at all. Consequently the resistance between two earth plates a few yards apart and between similar ones a thousand miles apart is about the same. Though the earth is made up mainly of what, in small quantities, are very bad conductors indeed, taking the earth as a whole it is an exceedingly good conductor. That makes it all the more important that where the current enters should be made as good a conductor as possible, and the construction and location of the earth plates is therefore very carefully considered so as to get the best results.

Wires, of course, connect the antenna to the earth, thereby forming what is called an "oscillatory cir

cuit." The ordinary electric circuit is a complete path of wire or other good conductor around which the current can flow in a continuous stream. An oscillatory circuit is one which is incomplete, but the ends of which are so formed that they constitute the two "plates" of a condenser. In that way, according to theory, the circuit is completed between the two ends by a strain or distortion in the "Ether" between them. A continuous current will not flow in such a circuit, but an alternating, intermittent or oscillating current will flow in it in many respects as if there were no gap at all but a complete ring of wire.

At some convenient point in this oscillatory circuit are inserted the wireless instruments, one set for sending and the other set for receiving, either being brought into circuit at will by the simple movement of a switch.

In small installations the central feature of the sending apparatus is an Induction Coil operated by a suitable battery or by current from a dynamo. Connected with it is a suitable spark gap consisting of two or three metal balls well insulated and so arranged that the distance between them can be delicately adjusted. This is generally done by a screw arrangement with insulating handles, so that the operator can safely adjust them while the current is on.

The current from the battery or dynamo to the coil is controlled by a key similar to those used in ordinary telegraphy, the action being such that on depressing the key the current flows and the coil pours forth a torrent of sparks between the knobs of the spark-gap, but on letting the key up again the

sparks cease. Since the sparks send out etherial waves which in turn affect the distant receiving apparatus it follows that a signal is sent whenever the key is depressed. Moreover, if the key be held down a short time a short signal is sent, but if it be kept depressed for a little longer a long signal is sent, by which means intelligible messages can be transmitted over vast distances.

Certain specified wave lengths are always used in wireless telegraphy. That is to say, the waves are sent out at a certain rate so that they follow each other at a certain distance apart. In other words, it is necessary to be able to adjust the rate at which the currents will oscillate between the antenna and earth. Every oscillatory circuit possesses two properties which are characteristic of it. These two properties are known as Capacity and Inductance. It is not necessary to explain here what these terms mean precisely. It is quite sufficient just to name them and to state that the rate at which oscillations take place in such a circuit depends upon the combined effect of these two properties. Consequently, if we can arrange things so that capacity or inductance or both can be added to a circuit at will and in any quantity within limits, we can within those limits obtain any rate of oscillation which we desire and consequently send out the message-bearing waves at any interval we like; in other words, we can adjust the wave-length at will.