are twisted in pairs with a lay of about three inches and the pairs are cabled in reverse layers, forming a cable about 15% inches in diameter. This cable is enclosed in a lead pipe having an internal diameter of 134 inches and 8 inch walls. The pipe is not pure lead, but is alloyed with about 3 per cent. of tin, to minimize chemical action tending to eat away the lead. The resistance of the average telephone cable conductor is about 45 or 50 ohms per mile.

When the laying of the conduits is finished the cables are drawn in place. This is done by stretching a rope from manhole to manhole by any suitable process, attaching the cable to the rope, and paying out the cable from the top of the reel that stands near one of the manholes, as shown at Fig. 97j. The method of drawing in by steam or gas engine operated winch is generally followed in extensive installations, though man or horse operated winches may be used in smaller installations. Care is taken in feeding the cable into the mouth of the conduit to prevent abrasion of the sheath. This may be done by means of a funnel shaped metal guide shield in the mouth of the conduit and also carefully feeding in and directing the cable by hand as it is pulled through.

Underground Wires. Some of the first attempts to lay underground telephone wires were, as we look back upon them now, quite ludicrous. In Washington one plan that was tried, at heavy expense and with no success, was the placing of

FIG 97j.-Method of Drawing Cable in Place Through Conduits by Steam Operated Winch.

bare copper wires in a wood trough in a trench in the street, the wires being held in position and separated from each other by being drawn taut through holes in porcelain plates fixed in the trough; the trough was then poured full of tar, covered with a plank, and the trench filled. Washington's summer climate, however, proved to be a little too much for this system; a gentle slope in the street caused the tar to flow toward the lower end of the conduit, where it leaked out of the containing trough and oozed up between the paving blocks to the street surface. This allowed the wires to cross with each other, making the experiment a failure.

Rubber-insulated cables, similar to successful telegraph cables, were next tried and used to a limited extent for very short distances. As a less costly and more efficient substitute, cotton-insulated wires were formed into a cable, impregnated with paraffin, and covered with a lead sheath; but the same phenomenon that seemed for a while to be an insurmountable obstacle in the way of transatlantic telegraphy made speech difficult through a short length and impossible through a considerable length of one of these cables. The electrostatic capacity of a mile or so of the wire in such a cable, although not noticeable with the comparatively large, slowly interrupted currents used in telegraphy, was absolutely fatal to the transmission of speech, which depends upon the most minute alternating currents, of very high

frequency and very complex wave-form. If the little voice-current was not completely absorbed by the capacity of the conductor, it had its wave form so changed that it could not reproduce intelligible speech at the further end of the line.

The preservation of the wave-form is the important thing. It has been frequently pointed out that with the transmission of power over longdistance by means of alternating currents, the thing sought is to get the largest possible percentage of the developed energy delivered safely at the farther end of the line, while the telephone transmission energy loss is of small importance compared with distortion of the waveshape. The man who works on the problem of long-distance telephone transmission is probably not at all horrified by the idea of energy losses in the line of 95 or 99 per cent., provided he can prevent material change in the shape of the wave.

After the early attempts with rubber-insulated and paraffin-core cotton-insulated telephone cables, the dry-core paper-insulated cable was developed, and there is now more than 7,000,000 miles of wire in use in such cables in the plant of the associated Bell telephone companies.

By means of a wire drawn in while the duct is being laid, or by means of sectional rods pushed through the duct from one manhole to the next, a hemp or wire rope is drawn through the duct by means of which the cable is in turn dragged into place. In some localities the rat-and-ferret

This con

method is or has been in practical use. sists in sending a rat through the duct, followed by a ferret in harness, who, in his pursuit of his intended prey, draws a light cord through to the next manhole, by means of which a heavier cord or rope, and finally the telephone cable, is drawn in.

accom

New French Telephone Cable.— The panying drawing, Fig. 97k, shows the details of construction of the new telephone cable recently laid between Abbotscliffe and Cape Gris-Nez by the Telegraph Construction and Maintenance Company, Ltd., on behalf of the French Administration of Post and Telegraphs.

Copper Conquctor Iron Loading Gutta Percha

FIG. 97k.-French Submarine Cable.

It will be seen from the diagrams that the cable is of continuously loaded type, and it is therefore of special interest, because it enables a direct comparison to be made with the Pupin coilloaded cable laid between the same points to the order of the British Post Office. It is claimed for this type that it has none of the mechanical objections that are inherent in the coil-loaded type. In external appearance and in general cable