advantage of using a thicker wire on the longer lines is thus seen. No. 4 wire is, for this reason, used on some of the longer lines.
The Earth—Earth Currents. Mention has been made of the ‘earth’ in the preceding description. This is the technical term used in relation to the fact, discovered by Steinheil in 1838, that the earth itself serves the purpose of completing the circuit, and renders the employment of a second or return wire unnecessary. The ‘earth’ may consist of a buried plate of metal connected with the battery or line wire, and of sufficient surface to afford the necessary contact, with sufficient earth to make a good connection, it being understood that a small quantity of earth does not connect well. Gas or water pipes form excellent ‘earths,’ care being taken that the connection is made with the main pipe itself, and not with a branch, where a badly made point might spoil the connection. Where similar ‘earths’ are in use, as, for instance, a copper plate at one end and an iron pipe at the other, a quasi battery is created, and minute currents pass along the line; these, however, are too weak to be of consequence. The earth, being of great size, offers no sensible resistance to the passage of the current, in the same way that a large wire offers less resistance than a thin one. While this quality of the earth is one of the most valuable aids to telegraphy (reducing so materially the cost of wire erection), it presents at times those embarrassing interruptions known as earth currents. These currents, at all times unwelcome visitors to a telegraph office, are very variable, changing rapidly from positive to negative, altering their direction with the hour of the day, and leaving one circuit to appear on another in a manner not explainable.
The Relay. There are several methods by which the transmission of signals is facilitated or accelerated. First among these may be placed the relay. In the previous description of the Morse system, we have assumed the instrument to be worked directly by the current sent along the line. On long circuits, however, direct working could only be accomplished by great battery power, owing to inevitable loss by leakage, experienced by a current before it reaches its destination. Consequently what is known as a relay is used and the sounder or recording instrument is worked by a ‘local current’ derived from a local battery at the receiving station. This is accomplished in the following way: The receiving instrument connected to the line instead of being made heavy enough to give an audible sound by which the signal may be understood, is made extremely light and delicate so that it can be operated by a weak current.
RELAY.
This relay consists of an electro-magnet wound with a large number of turns of insulated wire and an armature on a lever to which is attached a contact point which strikes another stationary contact when the lever is drawn toward the magnet, closing a circuit through these extra points in precisely the same way that the usual signaling key is worked. These contacts are connected by short wires with a regular sounder and a small battery. Whenever the lever of the relay is attracted it closes this extra or local circuit and causes the sounder to respond to every signal precisely as if it were connected directly to the main line.
Duplex-Working. The first suggestion of using a single line for more than one message was due to Moses G. Farmer of Boston, who in 1852 devised a method for this purpose and carried on successful experiments on a small scale. In 1853 Dr. Wilhelm Gintl first made use of the differential relay, and on the line between Vienna and Prague worked a method from which the modern practice of duplex telegraphy may be said to date. This system was rendered more successful in the following year by Carl Fischen of Hanover, and his invention was acquired by Siemens and Halske of Berlin, who had almost simultaneously devised a similar method. This improved system was introduced into Austria and from time to time there followed improvements devised by Stark and others. In England a duplex system was invented by Preece, while in other countries various methods from different inventors were forthcoming. In 1868 J. B. Stearns of Boston invented a system of duplex telegraphy which has since been used most successfully on many lines throughout the United States. The theory and practice of modern duplex telegraphy are somewhat complex and difficult to understand, but an outline of the principles involved may be given. In the differential method, which is largely used on land lines, there is what is known as a differential relay, that is a relay wound with two sets of coils in each of which the current flows in a different direction. Consequently when two currents of equal intensity pass through such a relay their effects are neutralized and the wire does not become magnetized. At each station a relay of this kind is provided, and there are two complete circuits, one of which includes the line wire, while the other consists of resistance coils equivalent in resistance to the line and known as the artificial line. The key and battery are common to both circuits, the points of divergence being at the relay and at the ground. If the key at one station is depressed the current will flow through both sets of coils of the relay at this station without producing any magnetizing effect, but at the distant station it will traverse but one set of coils and will accordingly cause the relay to operate the local sounder. The same thing of course holds good for the distant station. If both keys are depressed simultaneously the two batteries will combine to produce an increased strength of current flowing through one set of the differential coils of each relay, and by virtue of this preponderance of current the armature of the relay will be magnetized. A modification of this method is known as the polar duplex system and employs at each station two batteries or dynamos, so arranged that the direction of the current flowing in the line depends on whether the key is at rest or depressed. The current divides
