5 o8 THE POPULAR SCIENCE MONTHLY. 



above figures would be very greatly increased by the cost of under- 

 ground piping and chambers to contain the cables. 



It is thus demonstrated that it is technically possible to place all 

 of the wires in a city under-ground. It is also demonstrated that the 

 cost, even when a large number of wires run side by side, is enormously 

 increased. For many purposes, as telephony or electric lighting, a 

 considerable number of wires start out from a central office together, 

 but continually bifurcate until single wires run to the houses of the 

 subscribers. The cost of one wire by itself is vastly larger than where 

 many are run together, the cost of the pipe and for laying being not 

 much greater for fifty wires than for one, and the cost of single wire 

 cables being greater per mile of wire than multiple wire cables, so 

 that the expense of putting such a system as one of our telephone 

 exchanges entirely under-ground would place the cost of the instru- 

 ments entirely out of reach of the subscribers. If telephones were re- 

 quired in every house, as are gas and water, such a system might be 

 practicable, but at present that is not likely to be the case. 



The American Bell Telephone Company has recently constructed 

 two short lines of under-ground wires in the business section of Bos- 

 ton, and these give us excellent data from which to judge of the extent 

 of technical practicability and the expense of putting all wires under- 

 ground. We have seen that in Paris the retardation and induction 

 are both obviated by the use of double and twisted wires in metallic 

 circuit. It is necessary that all of the wires be in metallic circuit, 

 for, if a metallic circuit be connected to a single-line circuit, the dis- 

 turbances are not removed. If a subscriber in one city wishes to talk 

 with a subscriber in a neighboring city, both cities must have me- 

 tallic-circuit systems and metallic circuits between the two cities. As 

 the two lines constructed in Boston are short, only about one quarter 

 of a mile each, it was deemed best to use single-line circuits, hoping 

 that the induction and retardation on so short lines would not be 

 serious. 



The system is constructed as follows : Eight wrought-iron pipes, 

 three inches in diameter, are laid side by side in two rows, about four 

 feet below the surface. At each street corner is built a brick cham- 

 ber, large enough to admit a man, and with a cover flush with the 

 street. The cables, of which several kinds are in use, run out from the 

 basement of the central office through these pipes and up the side 

 of buildings to roofs, from which they spread out to the subscribers 

 by means of ordinary overhead lines. 



Conversation over these lines is not so easily carried on as by means 

 of overhead wires, and it is frequently possible to overhear other con- 

 versation. This prohibits further extension of the single-wire system 

 under-ground, for technical reasons. The cost of the piping and cham- 

 bers is in round numbers $50,000 per mile, and these pipes are intended 

 to accommodate one thousand wires. The cost of the cables is from 



