wire. The currents from H and J return to the dynamo D by the - conductor as shown. In a similar manner the flow of current in any multiple-wire system may be determined, no matter how large or how small the loads on the different parts may be.

The next step is to determine the voltage at the various points, as indicated in Figs. 66 and 67. Let us first consider the case (Fig. 66) of 10 amperes being required at E, with no current used in the rest of the system. Assuming each conductor to have one ohm resistance, the drop on the + is 10 volts, and the same on the wire, so that the lamps receive only 95 volts, the pressure at the dynamo being 115 volts. There will be no drop on any of the other three wires, since no current is drawn from them. It is interesting to observe that the potential difference between the extremities of the and O wires will be 125 volts, as shown in

Fig. 66. If three other groups of lamps were added, so that 10 amperes would flow directly across from E to J, the total drop would be 20 volts as before, 10 volts drop being transferred from the to the wire, and each group would receive 110 volts instead of 95, the aggregate number of lamps being four times as great. This brings out forcibly the advantage of a perfectly balanced five-wire system over a two-wire circuit, four times as many lamps being supplied with one-quarter as much drop for each.

With groups of lamps placed at E, F, G, H, and J (Fig. 67 corresponding to Fig. 65), the pressure is far from uniform, although the system is fairly well balanced in the number of lamps, but not in their position. This potential diagram is made by drawing from the five points marked +, ✪, O, Ə, and —, lines representing the pressure in the respective conductors and portions

thereof. By comparing Figs. 65 and 67 it will be seen that the direction of these lines is easily and definitely determined, the drop or slope of each section being equal to its resistance multiplied by the current flowing in it. In this connection it should be noted that the current in each group of lamps has been assumed to be constant, but it is evident that the group at J, receiving only 105 volts, will take less current than those at G, where the pressure is 117.5 volts. This fact might be allowed for by modifying the values of the current in proportion to the voltage; but the resistance of the lamps also varies, so that it would be very difficult to calculate the current that each group would take. In practice conductors are designed to supply a given current at a certain point; and slight variations in current due to changes in resistance, working conditions, etc., are not usually considered.

This may appear to be a somewhat rough method, but is not only justifiable, but practically unavoidable. In electric railway work, for example, the current required by a car varies greatly with the speed, grade, condition of track, load on the car, etc. Hence the only practicable plan is to assume a certain average current, or a certain maximum current, in designing the generating plant, conductors, etc. The average current corresponds to the ordinary working conditions, and the maximum current to the greatest possible requirements. The same is true for electric lighting, in which variations in the resistance of lamps are far less important than the changes in the number of lamps which are continually being made.

In practice the electric-light engineer considers the initial voltage at the generators, the resistance and current in each portion of the conductors which gives the drop, the latter subtracted from the initial voltage gives the pressure at the lamps. It would be an easy matter to calculate the resistance of the lamps by dividing the pressure they receive by the current flowing through them; but as a matter of fact one rarely, if ever, does this. It is only the inexperienced student who attempts to apply Ohm's law to the circuit as a whole. The practicing engineer confines it to determining the drop on the conductors, and usually considers one portion at a time, the drop in each being equal to its resistance multiplied by the current carried by it. The same is true in power transmission and distribution, including electric railway work. In fact it is

practically impossible to predetermine the resistance of the whole circuit except in very simple cases.

The author has examined five-wire systems in successful operation on a large scale in Paris and in Manchester, England. The difficulties encountered are not serious, and are apparently not much greater than with three-wire plants. Nevertheless, the increased number of conductors does involve more complication and possibility of accident. In the future the three-wire 440-volt system will undoubtedly be selected in preference to five-wire system.

Seven-wire System. This is the next higher multiple-wire system that would be used, since it can readily be divided into two four-wire systems, or three three-wire systems, in order to supply current to individual buildings where it is not necessary to carry all seven conductors. Neither the four-wire nor the six-wire systems are capable of being conveniently divided into equal parts in this way; hence they are not to be recommended for adoption, except perhaps in some special case. The seven-wire system, with all conductors of the same size, requires 0.0972, or a little less than one-tenth as much copper as an equivalent two-wire circuit; but its complication is so great as to make it of very questionable desirability. Its design and operation would be similar to that of the three- and five-wire systems already described.

CHAPTER V.

DIRECT CURRENT TRANSFORMER SYSTEMS OF ELECTRICAL

DISTRIBUTION.

THE fact that electrical energy can be readily transformed from a higher to a lower voltage, or vice versa, constitutes one of its most important advantages, and enables it to be conveniently and economically transmitted and distributed. The most prominent example of this method is the ordinary alternating current system, in which a high pressure of a thousand volts or more, generated by the dynamos, may be carried by small wires to a considerable distance, and there transformed to a low voltage that is harmless to persons and adapted to supply lamps, etc.

The direct current can also be transmitted in a similar manner, but it requires rotary machines instead of the simple induction coils or "static" transformers that are used for the alternating

current.

Rotary transformers consist of a motor and a dynamo combined, the former being driven by the current from the main or primary circuit, and the latter generating the current for the secondary circuit, by which the lamps, etc., are supplied. It is obvious that the dynamo may be designed to produce any desired voltage without regard to that of the primary circuit. But in every case the watts -product of the volts and amperes are less in the secondary circuit by an amount corresponding to the frictional and other losses which necessarily occur. This device has been given many names, such as dynamotor, motor-dynamo, motor-transformer, rotary-transformer, motor-converter, and rotary converter. The first of these has the advantage of being a single word, but has been objected to because the order in which the two machines work is inverted. This is not a serious objection, and the term is often used, being particularly appropriate to the construction in which the motor and dynamo are incorporated as one machine. In contradistinction the name motor-dynamo may be applied when there is a combina

tion of two machines. The word transformer has been almost universally adopted for the induction coil or static transformer, so that the term converter, which was formerly applied to this device, is now free to be used for the machine in which alternating are changed into direct currents, or vice versa, in the same armature winding as shown in Fig. 71.

In dynamotors the two armature windings are placed upon the same core and are acted upon by the same field magnet, as illustrated in Fig. 68. This construction secures compactness, and also causes the armature reaction of the dynamo to practically

neutralize that of the motor, thereby avoiding sparking and other troubles. But it is open to the objection that it is somewhat difficult to insulate the two windings from each other, and absolutely prevent the high voltage of one from breaking through to the other. Therefore this arrangement is not desirable where there are great differences in potential between the primary and secondary circuits, unless special precautions are taken. Another limitation of this construction is the difficulty of acting on the two armature windings independently for purposes of regulation. Since both are wound upon the same core and are under the influence of the same field, it is hardly possible to change the speed, magnetic