We come now to the submarine telegraph, which differs in many respects from the overland telegraph.  Obviously, since water and moist earth is a conductor, a wire to convey an electric current must be insulated if it is intended to lie at the bottom of the sea or buried underground.  The best materials for the purpose yet discovered are gutta-percha and india-rubber, which are both flexible and very good insulators.

The first submarine cable was laid across the Channel from Dover to Calais in 1851, and consisted of a copper strand, coated with gutta-percha, and protected from injury by an outer sheath of hemp and iron wire.  It is the general type of all the submarine cables which have been deposited since then in every part of the world.  As a rule, the armour or sheathing is made heavier for shore water than it is for the deep sea, but the electrical portion, or “core,” that is to say, the insulated conductor, is the same throughout.

The first Atlantic cable was laid in 1858 by Cyrus W. Field and a company of British capitalists, but it broke down, and it was not until 1866 that a new and successful cable was laid to replace it.  Figure 51 represents various cross-sections of an Atlantic cable deposited in 1894.

The inner star of twelve copper wires is the conductor, and the black circle round it is the gutta-percha or insulator which keeps the electricity from escaping into the water.  The core in shallow water is protected from the bites of teredoes by a brass tape, and the envelope or armour consists of hemp and iron wire preserved from corrosion by a covering of tape and a compound of mineral pitch and sand.

The circuit of a submarine line is essentially the same as that of a land line, except that the earth connection is usually the iron sheathing of the cable in lieu of an earth-plate.  On a cable, however, at least a long cable, the instruments for sending and receiving the messages are different from those employed on a land line.  A cable is virtually a Leyden jar or condenser, and the signal currents in the wire induce opposite currents in the water or earth.  As these charges hold each other the signals are retarded in their progress, and altered from sharp sudden jets to lagging undulations or waves, which tend to run together or coalesce.  The result is that the separate signal currents which enter a long cable issue from it at the other end in one continuous current, with pulsations at every signal, that is to say, in a lapsing stream, like a jet of water flowing from a constricted spout.  The receiving instrument must be sufficiently delicate to manifest every pulsation of the current.  Its indicator, in fact, must respond to every rise and fall of the current, as a float rides on the ripples of a stream.