174 Proceedings, &c., for 1885. 



E 



in the lamp, thus the current C = =j— - ■ — > 



r ' R + r -f r 



E 2 



This transforms — .Watts from horse-power into heat 



R + r + r 



(since there is no mechanical movement in the outer circuit), and 



R c 2 t 

 since this is distributed according to Joules' Law H = — - — 



J 



it is evident that we must make the resistance of the lamp high 



compared with that of the rest of the circuit. 



Let us now consider what is done in practice. 



There are two distinct methods of treating the case, known as 

 the series and the parallel arrangements. In the former we pass 

 the same current through lamp after lamp, and in the second, 

 increasing the current in proportion to the number of lamps, we 

 divide it up between the various lamps of the system. 



All other methods are modifications or combinations with slight 

 modifications of these two great systems. 



To bring before our mind more clearly the points of resemblance 

 and of difference of these two systems, we may examine the 

 following case : — 



And first take the series system. 



Suppose we are dealing with 100 50-volt lamps, taking 1 ampere 

 of current and developing 16 candle power, we shall require an 

 E.M.F. at least equal to 5000 volts — a tension that is approaching 

 the dangerous — and even then only 100 lamps are supplied. If 

 we calculate on a 90 per cent, efficiency we have R-f V = 500 

 ohms. This would mean an inexpensive dynamo of small size, 

 and, supposing the lights to be one mile away, a No. 28 main wire 

 (•34 mm.) 



But another difficulty is, that the failure of one lamp of the 

 series interrupts the circuit and puts all the others out. 



This is not likely to occur in arc lamps, and they are almost 

 always worked in this way ; but an incandescent lamp may fail 

 at any minute, and hence we shall have to find some method of 

 providing against this contingency. With such a high E.M.F. , 

 too-switching is almost impossible, and a special arrangement must 

 be introduced in the dynamo itself. 



The pure series system, then, will not be troubled with the 

 problem of keeping down the resistance of the dynamo and line, 

 but will, without any difficulty, have nearly all the energy repro- 

 duced in the lamp. But it will require to be provided with the 

 most perfect insulation and safety arrangements, and some special 

 apparatus for providing against failure of the lamp and with the 

 requisite quantity of current. 



We must next consider the parallel system. Suppose, as before, 

 we have 100 fifty-volt lamps, then we shall require a current of 



