ELECTRIC CANDLES. 237 



re-lighted ; but in that case it is necessary that the insulator 

 should be made of a moderately conducting substance, the 

 resistance of which is calculated accordingly. This arrange- 

 ment moreover enables the current to make a derivation 

 when the light goes out, and this prevents the generating 

 machine from " running away," as often happens when the 

 resistance opposing its working is suddenly withdrawn. We 

 have seen that Jablochkoff used a similar plan for re-lighting 

 his candles, and I am told his plan was patented a year 

 before that of which we are speaking. 



In order to keep the luminous point stationary in electric 

 candles, the Abbe Lavaud de 1'Estrade has contrived a kind 

 of candlestick in which the candle is supported by a float. 

 For this purpose the candlestick has three vertical rods, two 

 of which terminate in cylindro-conical vessels plunged into 

 two tubes filled with mercury. The third rod serves as a 

 guide, and a rack enables the mercury tubes to be placed at 

 any suitable height. The object of the vessels is to keep the 

 electric candle at any givea height, and their capacity, as well 

 as the volume of the rods which they supported, are calcu- 

 lated accordingly. This height may, it is true, be modified by 

 means of the rack, but the equilibrium between the weight 

 of the candle and the tendency of the vessels to rise up being 

 always the same, so long as the candle keeps its weight the 

 vessels are immersed to the same extent, whatever may be 

 the position of the system supporting the candle. It is only 

 when the candle comes to be consumed that this equilibrium 

 is destroyed and the candle rises, the amount being deter- 

 mined by the relation between the volume of the carbon con- 

 sumed and that of a cylinder of mercury having a diameter 

 equal to that of the rods and the weight of the carbons con- 

 sumed. If the rods attached to the vessels have such a 

 diameter as that the weight of the mercury they displace 

 precisely corresponds with the weight of the carbon consumed, 

 and such that the height they project from the mercury shall 

 be exactly the same as that by which the carbons are short- 



