signal currents pass through the suspended coil to earth, and in doing so the coil turns to left or right, according as the currents are positive or negative.

These movements of the coil are communicated by a connecting thread to a fine glass siphon, which is constantly spurting ink upon a band of travelling paper, and hence the trace of the ink on the paper follows and

delineates the movements of the coil. In Fig. 28, s is the signal coil suspended between the poles of two powerful electro-magnets м M.

In Fig. 29, the coil s is shown connected by a

FIG. 29.

thread and a small multiplying lever u to the siphon t swung from i. The siphon is as fine as a bristle, and dips into the ink-box K with its shorter arm, while the longer arm presents its nozzle to the travelling paper. So fine is the bore of the siphon that the ink will not

Understand a b c d e f g h i j k l m n o

FIG. 29a.

run unless it is electrified, and hence an ingenious machine o, Fig. 28, termed the "mouse-mill," is provided to generate electricity by "induction," after the manner indicated in Chapter III. pp. 24-25. The stream of electricity thus supplied is led from the mill to the ink by way of the strip of paper, or other semi-conductor F, and the receiving plate D. The particles of ink thus electrified rush through the fine bore of the siphon and rain upon the paper. The latter is pulled continuously past the siphon-point by the pulley A, which is driving the "mousemill." There are many adjustments and interesting actions about the siphon recorder, but it would require a whole chapter to describe them, and we will now conclude our notice of this, the most delicate and beautiful of all telegraph instruments, with a specimen of the message it delivers.

CHAPTER V.

THE TELEPHONE AND MICROPHONE.

THE speaking telephone of Professor Graham Bell has to a large extent superseded the telegraph instruments formerly employed on private lines for special correspondence. This electrical marvel was invented

FIG. 30.

-M

in 1876, and first exhibited

at the Centennial Exhibition in Philadelphia that year. It was originally feeble in its effects, owing to the transmitter and the receiver being identical. This apparatus is illustrated in Fig. 30, where the internal construction and external appearance are both shown. It consists of a bar magnet м, contained within the stem and having a small bobbin c of silkcovered copper wire surrounding its upper pole.

Directly over this pole and bobbin is placed a circular disc D of thin sheet iron, fixed round its edges by the

wooden mouth-piece o, but free to vibrate to and fro at its middle. The ends of the coil c are connected by two wires running outside the stem to terminals T T at its lower end, and connecting wires w w run from these outside the instrument and connect it in circuit with the line wire.

Formerly two such instruments were simply joined up to the line and the earth, and no battery was employed, for on speaking into the mouth-piece of one the vibrations of the voice bent the iron disc to and fro in front of the magnetic pole, and thus, by magnetoelectric induction, set up undulatory currents of electricity in the little coil around the pole. These currents were of a strength and form dependent on the vibrations of the voice, and after travelling along the line they passed through the coil of the receiving instrument, where by a reverse action they set the iron disc into a vibration sympathetic with the vibration of the disc of the transmitting instrument. The result was that an ear placed at the mouth-piece of that instrument could hear a feeble imitation of the distant speaker's voice.

It was, however, very feeble, owing to the fact that the voice itself generated the electric current; and louder effects were obtained by Edison, who adopted the principle of causing the voice only to vary the strength of a battery current flowing in the line, just as a miller regulates the stream of water flowing in the race by working the sluice valve up and down.

He did this by the "carbon transmitter" shown in section in Fig. 31. There there is no magnetism employed, and the strength of current is regulated by passing it through a wafer of carbon, which alters its