It is also very important to see that there are no sharp points on the conductors themselves, on the copper sleeve, on the edges of the lead covering or on the lead sleeve. All these should be made perfectly smooth, for points facilitate disruptive discharges. Branch joints had better be made as T-joints rather than as Y-joints, for they are better electrically and mechanically, although they occupy more room in the manholes. They are of course made in the same way as straight joints, a lead T-sleeve being used, however. For multiple arc circuits copper T-sleeves and for series circuits copper L-sleeves are used.

Telephone and telegraph cables are made of any required gauge of wire and with from 1 to 150 conductors in a cable. In jointing these the splices are never soldered, the conductors being joined either with a twist joint or with the so-called Western Union splice. Each splice is covered with a cotton or silk sleeve or a wrapping of tape, the latter being preferable, although considerably increasing the time necessary for making the joint. Great care must be taken that no ends of wire are left sticking up, for they will surely work their way through the tape and grounds, and crosses will be the result. The wires should always be joined layer to layer and each splice very tightly taped in order to get as much insulating compound around each splice as possible in the limited space. The splices should be "broken" as much as possible, so as to avoid having adjoining splices coming over each other. After the joint has been saturated with insulating compound the wires should have an outside wrapping of tape to keep them in shape, and then the sleeve is wiped on and filled. If the insulation resistance of the jointed telegraph or telephone cable is a quarter of what the cable tested in the factory, it may be considered that an exceptionally good piece of work has been done.

I have spoken more particularly of fibrous lead covered cables, as the handling of them includes practically every step of the work on any other kind of underground cable. In insulating dry core paper cables a paper sleeve is slipped over the splice, and in rubber cables the splice is wrapped with rubber tape; all other details are the same for these as for the fibrous cable.

In the laying of light and power cables every joint, as made, should be tested for insulation with a Thomson galvanometer, as the insulation must necessarily be very high, and if one joint or section of cable is any weaker than another it may be very important in the future to know it. All tests must be made after the joint has cooled, for while hot its insulation resistance will be very low.

Tests for copper resistance should also be made to determine if the splices are electrically perfect; an imperfect splice may cause considerable trouble. In telegraph and telephone cables the conductors should be of very soft copper, for in stripping the conductor of insulation it is very easy to nick the wire, and if of hard drawn copper open wires will be the result.

All work should be frequently tested for continuity with telephones, magnetos, or small portable galvanometers. It is only necessary to ground the conductors at one end and try each wire at the other end. For this sort of work a telephone receiver used with one cell of some dry battery is most convenient, and has the additional advantage of affording a means of communication while testing, and is by far the best thing for identifying and tagging conductors.

These cables should be frequently tested during the progress of the work for grounds and crosses with a Thomson instrument, and when the cable is complete, a careful series of tests of the capacity, insulation resistance, and copper resistance of each wire should be made and the exact condition of the cable determined before it is put in service, and thereafter an intelligent oversight of the condition of the circuits can thus be more readily maintained.

Where a company has extensive underground service, a regular cable gang should be in its employ, for quick and safe handling of cables demands the employment of men accustomed to the work. If the cable has been properly laid and tests show it to be in good condition before current is turned on, almost the only trouble to be anticipated will be due to mechanical injury. Disruptive discharge, puncturing the lead, may occur; but the small chance of its occurring can be greatly lessened by the use of some kind of "cable protector," which will provide for the spark an artificial path of less resistance than the dielectric of the condenser, which the cable in fact becomes.

If a fault suddenly develops on a circuit, the chances are it will be found in a manhole, and an inspection of the cable in the manhole will generally reveal the trouble without resorting to locating with a Wheatstone bridge. The cable is often cut through at the edge of the duct, or damaged by something falling on it, or by some one "walking all over it." To guard against these, the ducts should always be fitted with protectors both above and below the cable. The cables should never be left across the manholes, for they then answer the purpose of a ladder, but should be bent, around the walls of the hole and securely fastened with lead straps, that they may not be moved and the lead gradually worn through.

In telegraph cables, when one or two conductors "go," it will probably be useless to look for trouble except with instruments; but if several wires are "lost" at once it will probably be found to be caused by mechanical injury, which can be located by inspection. If it is ever necessary to loop out conductors, a joint can be readily opened and the conductors wanted picked out and connected into the branch cable and the joint again closed without disturbing the working wires. In doing this a split sleeve must be used, and the only additional precaution to be taken is in filling the sleeve to have the insulating compound not hot enough to melt the solder and open up the split in the sleeve. In cutting in service on light and power cables it is entirely practicable to do so without interruption of service on multiple arc circuits, even those of very high voltage; but they require great precaution and involve considerable risk to the jointer, and where possible the circuit to which the connection is to be made should previously be cut dead.

Where the voltage is not dangerous to human life, almost any service connection can be made without interruption of service.

I have only indicated a very few of the operations that may be found necessary, and the probable causes of troubles that may be encountered in the operating of underground circuits, believing that the different problems that arise can, with a little experience, be successfully met by any one who has a fair knowledge of the original construction of cable lines.--Electrical World.