466 FRAGMENTS OF SCIENCE 



substance between them melts like the wax of a candle. 

 The comparison, however, only holds good for the act of 

 melting; for, as regards the current, the insulating plaster 

 is practically inert. Indeed, as proved by M. Rapieff and 

 Mr. Wilde, the plaster may be dispensed with altogether, 

 the current passing from point to point between the naked 

 carbons. 



M. de M^ritens has recently brought out a new can- 

 dle, in which the plaster is abandoned, while between 

 the two principal carbons is placed a third insulated rod 

 of the same material. With the small De Meritens ma- 

 chine two of these candles can be lighted before you; 

 they produce a very brilliant light.* In the Jablochkoff 

 candle it is necessary that the carbons should be consumed 

 at the same rate. Hence the necessity for alternating cur- 

 rents by which this equal consumption is secured. It will 

 be seen that M. Jablochkoff has abolished regulators alto- 

 gether, introducing the candle principle in their stead. In 

 my judgment, the performance of the Jablochkoff candle 

 on the Thames Embankment and the Holborn Viaduct is 

 highly creditable, notwithstanding a considerable waste of 

 light toward the sky. The Jablochkoff lamps, it may 

 be added, would be more effective in a street, where 

 their light would be scattered abroad by the adjacent 

 houses, than in the positions which they now occupy in 

 London. 



It was my custom some years ago, whenever I needed 

 a new and complicated instrument, to sit down beside its 



^ The machine of M. de Meritens and the Farmer-Wallace machine were 

 •worked by an excellent gas-engine, lent for the occasion by the Messrs. Crossley 

 of Manchester. The Siemens machine was worked by steam. 



