1 66 Telephone Systems of the Continent of Europe 



a station and drops, say, twenty-one pairs at the first junc- 

 tion box ; thence it is continued by a 28-pair and a seven-pair, 

 and finally by a seven-pair alone, until all the pairs have been 

 disposed of. Cable of this kind is used for the interior wiring of 

 the central station at the Rue Gutenberg, as well as for the out- 

 side work. Capacious as the sewers are, their resources are not 

 inexhaustible, while the necessity of protecting the cables laid 

 therein in costly iron troughs renders sewer work somewhat ex- 

 pensive. These considerations have led to cables containing 

 junction lines between some of the switch-rooms being laid in 

 trenches beneath the street pavements. In one such trench, one 

 meter deep, twenty 5 2 -pair lead-covered cables are laid without 

 any protection other than a galvanised iron netting placed some 

 inches above them, designed to give warning of their existence to 

 strange workmen who may open the ground. An admirable 

 feature of these paper cables is the fact that they cannot be spoiled 

 by access of moisture. The ends are not sealed in any way, and 

 should water get in through a fault, even to the extent of short- 

 circuiting all the wires, it may be driven out and the insulation 

 raised again to its normal figure of 6,000 megohms per kilometer 

 by forcing dry air, not necessarily warm, into one end of the 

 cable, under a pressure of two kilogrammes per square millimeter. 

 This air gradually makes its way through the cable, whatever its 

 length may be (from seven to eight kilometers have actually been 

 operated on), carrying with it to the further end all the moisture 

 within it. For some hours after the application of the pressure 

 no improvement is noticeable ; then the insulation begins to go 

 up slowly, but at an ever-increasing ratio, until at the end of some 

 twenty-four hours the mending proceeds with great rapidity, so 

 that thirty hours of pressure usually suffices to restore what had 

 appeared to be a hopelessly bad cable to full working efficiency. 

 If it is not convenient to look for and remove the fault, the appli- 

 cation of pressure continuously, or for a few hours every day, will 

 keep the cable going without disturbing the subscribers. When 

 the fault is looked for, its position is first determined as nearly 

 as possible by electrical test, and the pressure then turned on. 

 Usually the workmen find the fault by the sound of air issuing 

 from it, or by simple inspection, and it may then be effectually 



