474 



its removal from the tomb and exposure to the air — proving 

 that these lamps were not supplied from any bituminous 

 source, or volcanic fire. He considers the requisites for 

 an everburning lamp to be, — a perpetual wick, which 

 might be made of gold wire, or asbestus; and a perpetual 

 supply of fuel, which he imagines the bituminous springs 

 of Pitchford, in Shropshire, or the inflammable gases is- 

 suing from fissures in coal mines, would afford. That such 

 could supply fuel for a flame, so long as the bituminous 

 spring existed, or the gas continued to exhale from the 

 mines, is evident ; but it no more deserves the appellation of 

 an everburning lamp, than does a fire arising from any 

 volcanic source. The desideratum for such a lamp is, that 

 it should contain, within itself, a renovating principle, such 

 as, probably, does the luminous atmosphere encompassing 

 the body of the sun, supposed by Sir William Herschel 

 to be electrical. 



** That electricity was the principle upon which such a 

 lamp could be constructed having occurred to me some 

 years ago, I reflected, upon the different means by which a 

 constant light could be produced from this source, and 

 concluded that, if by an arrangement of metals a thermo- 

 electric current could be produced of sufficient intensity 

 to decompose water, the heat produced by the burning of 

 the two gases arising from the decomposition, would be suf- 

 ficient, when applied to the alternate metallic junctions, to 

 continue the electrical current of the thermo-electric pile ; 

 while the gases, which in burning become aqueous vapour, 

 might be condensed by passing through a long tube, through 

 which being conveyed to the closed vessel in which the 

 water had been originally placed, they would again under- 

 go decomposition, recombination, and condensation. Such a 

 thermo-electric arrangement has been discovered by Prof. 

 Botto of Turin, who has obtained decomposition of water 

 from a series composed of a great number of wires of 



