It was later discovered by Poulsen that if one of the arc electrodes was kept cool by making it of copper and passing water through it that the efficiency was greatly increased. A further improvement was obtained by burning the arc in an atmosphere of coal gas or hydrogen. By surrounding the arc with a powerful magnetic field, its resistance is greatly increased and the voltage raised.

FIG. 147.—Poulsen wireless telephone equipment. The condenser shunted around the are usually consists of a number of metal plates, placed above one another in a tank of insulating oil. The inductance is simply a single helix or bare wire.

By connecting a telephone transmitter to the arc in the same manner that it is connected to the speaking arc, the oscillations can be varied in accordance with the vibrations of the voice. The apparatus is connected to the aerial and the earth through the medium of a loose-coupled helix, formed by providing the helix in series with the arc and condenser, with a secondary winding.

FIG. 148.—The Majorana wireless telephone transmitter.

The ordinary carbon transmitter, in its common form, is unsuited for wireless telephonic work, on account of its inability to handle large amounts of power. Many modifications have been designed, the usual procedure being to make it on a larger scale.

One of the most interesting forms, and also probably the best, is that devised by an Italian inventor, Majorana.

Its action will be clear from the illustration. T is a tube in which water or some other liquid is allowed to flow in the direction of the arrow. The bottom of the tube is contracted so that the stream will issue in a fine jet. The tube is made of strong, rigid material, except at one point, D, where an opening is covered with a thin elastic diaphragm. This diaphragm is connected by means of a short rod to a second diaphragm, which is provided with a mouthpiece. The water normally flows out of the jet in a smooth, unbroken column, breaking into drops at about the point A. As soon as it is disturbed in any way, however, the distance from the outlet of the tube and the point where the drops commence becomes shortened. The vibrations of the voice, thrown into the mouthpiece and striking the diaphragm, are transmitted to the membrane through the medium of the little rod, and so cause corresponding changes in the pressure of the fluid in the tube. Each variation or disturbance in the pressure increases or decreases the length of the stream before it breaks into drops.

A pair of fine wires are inserted in the stream where the contractions are the strongest.