The strain on the cable should be perfectly uniform and steady; if the power is applied by a series of jerks either the lead covering may be pulled apart or some of the conductors broken. At the reel there must always be a large enough number of men to turn it and keep the cable from rubbing on anything, and in the manhole one or more men to see that the cable feeds into the duct straight and to guide it if necessary. If the ducts are of iron and are not perfectly smooth at the ends, these should be made so with a file, and in addition a protector of some sort should be placed in the mouths of the duct, both above and below the cable. Six inches of lead pipe, split lengthwise and bent over at one end to prevent being drawn into the duct with the cable, makes a very good protector. The cable should be reeled off the drum just fast enough to prevent any of the power used in pulling the cable through the duct being utilized in unreeling it. If this latter is allowed to occur the cable will be bent too short and the lead covering buckled or broken, and also the cable may be jammed against the upper edge of the duct and perhaps cut through.
If the reel is allowed to turn faster than the cable is drawn in, the first three or four turns on the reel will slacken up, and the lead covering may either be dented or cut through by scraping on the ground. If the cable end when pulled through up to the block is not long enough to bend around the hole more than half way, the rope should be unfastened from its end, a length of rope with a well frayed out end should be run through the block, and by fastening to the cable close to the duct, with a series of half hitches, as much slack as necessary can be pulled in. If this is properly manipulated there need not be a scratch on the cable, but unless great care is taken the lead may be pressed up into ridges and the core itself damaged.
Immediately after the cable is drawn in, if the joint is not to be at once made, the open end or ends should be cut off and the cable soldered up, as most cables are very susceptible to moisture and readily absorb water even from the atmosphere. Where practicable it is always a good plan to pull the cable through as many manholes as possible without cutting the cable; for the joint is, especially in telephone or telegraph cables, the weak point. To do this the rope should be pulled through the proper duct in the next section without unfastening it from the cable; the winch should be moved to the next manhole, and pulling through then done as before. There should always be a man in every hole through which the cable is running to see that it does not bind anywhere and to keep protectors around the cable.
It is not advisable to pull more than one cable into a duct, and never advisable to pull a cable into a duct containing another cable, but if two or more cables have to go into the same duct, they should always be drawn in together. Lead covered cables and those with no lead on the outside should never be pulled into the same duct, for if they bind anywhere the soft cable will suffer where two lead covered cables would get through all right. Some manufacturers are now putting on their cables a tape or braid covering, which saves the lead many bad bruises and cuts, and is a valuable addition to a cable at very little additional expense.
Practically all electric light and power cables are either single or double conductors, and the jointing of these is comparatively a simple matter, although requiring considerable care. The lead is cut back from each end about four or five inches, and the conductors bared of insulation for two or three inches. The bare conductors should be thoroughly tinned by dipping in the metal pot or pouring the melted solder over them. A sperm candle is better than resin or acid for any part of the operations where solder is used. A lead sleeve is here slipped back over the cable, out of the way, and the ends of the conductors brought together in a copper sleeve which is then sweated to a firm joint. This part must be as good a piece of work mechanically as electrically. The bare splice is then wrapped tightly with cotton or silk tape to a thickness slightly greater than that of the insulation of the cable, and is thoroughly saturated with the insulating compound until all moisture previously absorbed by the tape is driven off.
The lead sleeve is then brought over the splice and wiped to the cable. The joint is then filled with the insulating compound poured through holes in the top of the sleeve; these holes are then closed and the joint is complete, and there is no reason why, in light and power cables, that joint should not be as perfect as any other part of the cable. When the cable ends are prepared for jointing they should be hung up in such a position that they are in the same plane, both horizontal and vertically, and firmly secured there, so that when the lead sleeve is wiped on the conductor may be in its exact center, and great care must be taken not to move the cables again until the sleeve is filled and the insulation sufficiently cooled to hold the conductor in position.
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.