Let us suppose that four transmitters are sending messages simultaneously from London to Edinburgh.
Wires from all four instruments are led into a circular contact-maker, divided into some hundreds of insulated segments connected in rotation with the four transmitters. Thus instrument A will be joined to segments 1, 5, 9, 13; instrument B to segments 2, 6, 10, 14; instrument C with segments 3, 7, 11, 15; and so on.
Along the top of the segments an arm, connected with the telegraph line to Edinburgh, revolves at a uniform rate. For about 1/500 of a second it unites a segment with an instrument. If there are 150 segments on the “distributor,” and the arm revolves three times a second, each instrument will be put into contact with the line rather oftener than 110 times per second. And if the top speed of fifty words a minute is being worked to, each of the fifteen signals occurring in each second will be on the average divided among seven moments of contact.
A similar apparatus at Edinburgh receives the messages. It is evident that for the system to work satisfactorily, or even to escape dire confusion, the revolving arms must run at a level speed in perfect unison with one another. When the London arm is over segment 1, the Edinburgh arm must cover the same number. The greatest difficulty in multiplex telegraphy has been to adjust the timing exactly.
Paul la Cour of Copenhagen invented for driving the arms a device called the Phonic Wheel, as its action was regulated by the vibrations of a tuning-fork. The wheel, made of soft iron, and toothed on its circumference, revolves at a short distance from the pole of a magnet. As often as a current enters the magnet the latter attracts the nearest tooth of the wheel; and if a regular series of currents pass through it the motion of the wheel will be uniform. M. la Cour produced the regularity of current impulses in the motor magnet by means of a tuning-fork, which is unable to vibrate more than a certain number of times a second, and at each vibration closed a circuit sending current into the magnet. To get two tuning-forks of the same note is an easy matter; and consequently a uniformity of rotation at both London and Edinburgh stations may be insured.
So sensitive is this “interrupter” system that as many as sixteen messages can be sent simultaneously, which means that a single wire is conveying from 500 to 800 words a minute. We can easily understand the huge saving that results from such a system; the cost of instruments, interrupter, &c., being but small in proportion to that of a number of separate conductors.
The word-sending capacity of a line may be even further increased by the use of automatic transmitters able to work much faster in signal-making than the human brain and hand. Sir Charles Wheatstone’s Automatic Transmitter has long been used in the Post-Office establishments.
The messages to be sent are first of all punched on a long tape with three parallel rows of perforations. The central row is merely for guiding the tape through the transmitting machine. The positions of the holes in the two outside rows relatively to each other determine the character of the signal to be sent. Thus, when three holes (including the central one) are abreast, a Morse “dot” is signified; when the left-hand hole is one place behind the right hand, a “dash” will be telegraphed.
In the case of a long communication the matter is divided among a number of clerks operating punching machines. Half-a-dozen operators could between them punch holes representing 250 to 300 words a minute; and the transmitter is capable of despatching as many in the same time, while it has the additional advantage of being tireless.
The action of the transmitter is based upon the reversal of the direction or nature of current. The punched tape is passed between an oscillating lever, carrying two points, and plates connected with the two poles of the battery. As soon as a hole comes under a pin the pin drops through and makes a contact.