ELECTRIC CANDLES. 223 



candle having carbons of 10 millimetres in diameter, he 

 placed the same carbons in one of Serrin's regulators, ad- 

 justed so as to give the same length of arc. The conditions 

 thus being the same in each case, he found that the luminous 

 intensity was also the same; and he points out that the heat 

 employed in volatilizing the insulator was not entirely lost, as 

 is generally supposed. With certain insulators, particularly 

 with kaolin, there may be, according to Jablochkoff, some 

 diminution of light from the somewhat considerable deriva- 

 tion of the current through the fused part of the insulator ; 

 but with plaster insulators this inconvenience does not exist, 

 and above the candle there is a flame, which he considers 

 useful. Be that as it may, these candles are not exceptional 

 in being attended with a loss of electricity, for in the regu- 

 lators there is also a loss of current, due to the introduction 

 into the circuit of a coil of more or less resistance. But we 

 have seen that the Lontin arrangement, by derivation, in a 

 great measure avoids this inconvenience. 



The Jablochkoff candles require to be used with alternately 

 reversed currents: and we have seen that to obtain such 

 from the Gramme machine it was necessary to invent a 

 special machine, which at the same time would allow the 

 action to be distributed over several circuits. It is easy to 

 see why the currents must be reversed, if we remember that 

 with direct currents one of the carbons (the positive one) 

 would be consumed quicker than the other, and the distance 

 between them would gradually increase, until it would become 

 so great that the light would go out. Jablochkoff asserts, 

 however, that this may be prevented by increasing the 

 diameter of the positive carbon sufficiently to make it burn 

 as slowly as the other. This plan has not hitherto been tried. 



In order to obtain a slower rate of consumption, Jabloch- 

 koff has lately used metallized carbons. We have seen (p. 137) 

 that this plan, which was invented by Reynier, has given ex- 

 cellent results, and in experiments recently made at Geneva 

 the candles were thus prepared. The metal was, however, 



