recently extended use of rapidly alternating currents is another reason against the employment of iron conductors.

It will be seen, from a study of the table given on p. 22, that the choice of materials for electrical conductors of any kind is really limited to the two metals above mentioned, viz. copper and iron; so that for electric lighting work we have no alternative but to use copper.

The accompanying table concerning the various sizes of copper wire generally employed gives some instructive details. It will be seen that most of these conductors are made up of a number of comparatively small wires stranded together, the chief objects being to impart greater flexibility and to reduce the risk of complete fracture.

Now, as in overcoming the resistance of a conductor electrical power is wasted, and as it costs money to develop electrical power, it is evident that in any commercial system such waste must be kept down to a minimum. This can be done by simply reducing the resistance, that is, by increasing the size of the conducting wires; but as this also is an expensive matter, care must be taken that the addition thus made to the expenditure in conductors is not so excessive as to more than counterbalance the cost of the power continually being saved. It has been laid down. as a general rule, that, for the transmission of any given current, the size of the conductor most economical to employ is one offering such a resistance that the cost of the energy wasted per annum in heating the conductors should be equal to the interest per annum on the original outlay upon them. Knowing the average current to be transmitted, it becomes easy to find the average electrical horse-power wasted in a conductor of any given resistance; but the cost of developing a horse-power depends upon many conditions, principally local, such as the cost of fuel, attendance, rental, repairs, prime cost and efficiency of the plant. And with regard to the conductors themselves, it must be remembered that it is not merely a question of the quantity and price of the copper employed, but also of the insulation and laying.

Most main conductors are, in England, placed underground, and in many cases the cost of laying considerably exceeds the

actual cost of the wire. Further, although the resistance of a wire one inch in diameter is only one-fourth of that of a wire half an

inch in diameter, it does not cost anything like four times as much, nor even twice as much, to lay the thicker wire as it does to lay the thinner, for the labour of removing and replacing the paving, earth, &c., would be almost the same in both cases. Again, the insulation of the wire is an important and expensive item, which does not increase so rapidly as the resistance of a wire is reduced by an increase in its size; so that it does not by any means follow that a given reduction in resistance entails a proportionate increase in expense, and it becomes impossible to lay down any hard and fast rule which shall determine the exact size of conductor for any given case.

It may be noticed incidentally, that when the diameter of a round conductor is doubled, although its sectional area and therefore its conductivity is increased fourfold, its surface is only doubled. Therefore, if a current of four times the strength is passed through it, the heat developed will be four times as great (since power wasted = C2R), while the surface at which radiation takes place has only been doubled. The temperature of the thicker wire will consequently rise higher than that of the thinner one, when they carry currents in proportion to their conductivities. A wire employed for the purpose of transmitting current to lamps or motors should never be so small that the maximum current transmitted can appreciably raise its temperature; but for other special cases it may be noted that one advantage attending the use of bare conductors is the greater facility afforded by them for the radiation of heat as compared with covered conductors.

So many considerations, mostly special for every particular case, enter into the question of the best size and shape of the conductor consistent with strict economy, that we cannot discuss the matter fully here. But with regard to the reduction of resistance by the employment of high conductivity copper, it should be noticed that, as the presence of a minute quantity of foreign matter causes such a great increase in the resistance of this metal, it is always economical to use the purest copper obtainable commercially.

In systems of distribution of electrical power by means of a constant current, the question is comparatively simple, as the

current employed is not a heavy one, and has the same value at all times and in all parts of the circuit. The chief difficulty likely to arise is in providing for future extensions of the system when the potential difference which can be applied at the ends of the circuit is limited.

The more interesting and more difficult problem consists in the supply of current to lamps, or other apparatus, at a constant potential; for then the main conductors have to carry a very heavy and variable current. The matter becomes more difficult if the lamps are distributed over a wide area, or are situated at a distance from the generating station. As has been pointed out in Chap. XIII., the power wasted may in such cases be reduced to a minimum, by transmitting it in the form of a small current at high pressure, and reducing the pressure at the required point, to a suitable value. But such a system has its disadvantages. Although the cost of the copper is greatly reduced, the high potential difference employed demands very efficient and expensive insulation, the engines and dynamos must always be kept running, and when very little power is being demanded the efficiency of the transformers and of the whole system falls to a low value. For even when the secondary circuit of a parallel transformer is disconnected, some current passes through the primary, and when only one or two lamps are joined up, the power appearing in the secondary may be but a comparatively small fraction of that absorbed by the primary. When the number of transformers is large, the total power wasted becomes considerable during the times when little or no light is required.

In the other method of distribution by means of continuous currents direct from the dynamo to a number of lamps all joined up in parallel, the chief problems to be faced are the heavy loss occurring in the mains and the difficulty of regulating the supply to each lamp.

Such an arrangement is indicated by the diagram in fig. 322, where D represents a dynamo capable of maintaining a constant potential difference at its terminals; A and B, the main leads from the machine to the nearest lamp; and E, F the continuation of those leads, between which the lamps are placed. Suppose there to be 100 lamps so joined in parallel, each requiring a current

of half an ampere, and a potential difference at its extremities of 110 volts. The total current supplied by the dynamo with all the lamps in use would be 50 amperes, and this current would have to be carried by the main leads A and B. Supposing the resistance of A and B to be one-tenth of an ohm, the power wasted in overcoming this resistance would be 250 watts, and the consequent fall of potential 5 volts. Therefore the machine must develop at least 115 volts at its terminals in order to maintain 110 volts at the nearest lamp.

Now a further fall of potential would take place along the more distant mains, E, F ; suppose this to amount to 10 volts, then the pressure at the most distant lamp would only be 100 volts, while if this were raised to the desired value of 110 volts by

an increase at the dynamo, the nearest lamp would then be working at 120 volts, or 10 volts above its proper pressure. Even ignoring the waste of power, such a difference could not be permitted if similar lamps were used throughout the system, as some would be giving far above and others far below their normal candlepower.

It would, however, be practicable, but very inconvenient, to employ different types of lamp, placing those made to run at 110 volts at the end near the dynamo, and others constructed for 100 volts at the further end of the line, and so on. But even then, if the dynamo were perfectly regulating-that is to say, capable of maintaining a constant potential difference at the brushes under all circumstances the potential at the far end of the mains would rise considerably when any number of the nearer or intermediate lamps were cut out of circuit.