WILLIAM SIJ-:MI-:\S, I-\R.S. 251 



cent, of incombustible admixture) and it requires -'- -- = 56 Ibs., 



1 Oj^tU \j 



or just half a hundredweight, of this coke to produce the same 

 heating effect as 1000 cubic feet of gas. But 1000 cubic feet of 

 gas cost on an average 3. 6d., and half a hundredweight of coke 

 not more than Gd. (at 20s. a ton), or only one-seventh part of the 

 price of gas. 



If heating gas could be supplied at a much cheaper rate than 

 at present, it would in many cases be advantageous to substitute 

 incombustible matter, such as balls of asbestos for coke or anthra- 

 cite. The consumption of gas would in that case have to be 

 increased very considerably, but the economical principle involved 

 (that of heating the air of combustion by conduction from the 

 back of the grate) would still apply, and would produce economical 

 results as compared with those obtained by the gas-asbestos 

 arrangements hitherto used. 



To illustrate the efficiency of this mode of heating the incoming 

 air by what would otherwise be waste heat, I will show you 

 another application of the same principle which I have made 

 very recently to the combustion of gas for illuminating purposes. 



Gas engineers were formerly under the impression that a supply 

 of cold air was favourable to the production of a brilliant flame. 

 This is a misconception, which was very general also as regards 

 the combustion of solid fuel in furnaces, until it was disproved by 

 Stirling, by Nielson, and by the introduction of the regenerative 

 gas furnace. The " duplex burner " owes its brilliancy to the 

 heating effect of the one burner upon the other ; and my brother, 

 Mr. Frederick Siemens, has quite recently constructed a burner in 

 which the flame of the gas is reversed in its action in order 

 to heat in its descent the ascending current of flame-supporting 

 air. 



By the application of the principle of conduction before 

 described, I obtain the hot-air current in a most simple manner 

 without interfering with the free action of the flame. The con- 

 struction of my burner will be seen from the accompanying Figure 3, 

 Plate 5 : a is an ordinary Argand burner, taking its supply of 

 gas through the vertical copper tube 5. This copper pipe ter- 

 minates in a rod c of highly conductive copper, which passes 

 upward through the burner, and two or three additional conduc- 



