and that in Paris, which is the only city in the world having a complete under-ground system, there are unusual facilities for the running of wires, as sewers large enough to walk about in extend even under the less important streets of the city. Moreover, it has been found that, for delicate and quick-working apparatus, such as automatic telegraphs, polarized relays, and, above all, the telephone, long underground lines are far less efficient than pole lines. There are two reasons, apart from the difficulty of securing good insulation, why these long under-ground lines are comparatively inefficient:
1. If an electric conductor be brought near to a large mass of conducting matter, as is a wire when it is taken down from a pole and buried in the earth, there appears in the current the phenomenon of retardation, by which each signal, instead of being sharp and distinct, is partly kept back, so that it overlaps and mingles with the next; the result is to limit the speed of working of the apparatus; or if, like the telephone, it be an apparatus in which the currents are necessarily extremely frequent, to confuse and destroy the signals altogether. With ordinary Morse telegraphic apparatus, this is not very troublesome on under-ground lines a hundred miles long. With delicate relays, and more especially with quick working printing telegraphs, or automatic telegraphs, such lines are very troublesome; and, with telephones, the retardation is a very troublesome matter on under-ground lines ten miles long.
2. The second difficulty is called induction, and is noticed when two or more wires are run side by side and near together, as they necessarily are in an under-ground cable.
If the signals on one wire of such a cable be sharp and quick, they cause fac-simile signals on all of the neighboring wires, and this too, though the insulation may be absolutely perfect; indeed, above a certain point, the more perfect the insulation the greater the induction. The result of this phenomenon is, that messages sent over one wire are liable to be received on all of the other wires, and, in the case of the telephone, this phenomenon is noticeable on cables one thousand feet long, and on a cable one mile long the parties on one wire can easily understand what those on the other wires are saying. For any other instrument, however, the interference only becomes annoying on much longer lines. Steady currents, like those used with electric lights, are, of course, not affected either by retardation or induction.
In our own country there is little doubt that the proper method of constructing electrical wires between cities is, to string them on poles in mid-air. A brief review of some of the European systems that have been constructed will convince us of this. Between the years 1847 and 1850 a system of cables, containing 2,648 miles of wire, was laid under-ground to connect Berlin with the other principal cities of Prussia. Gutta-percha-covered wires were drawn into lead tubes, which were then buried in trenches two feet deep. The cost of this
