MECHANICS AND USEFUL ARTS. 27 



lar holes are cast in the transverse surface of these bars, so that 

 when placed together on the bearers they present the appearance 

 of a tunnel about 9 inches high pierced with numerous small 

 holes, an arrangement not differing widely from that of Wye 

 Williams, except that the tunnel, being of loose cast-iron pieces, 

 is no more liable to deterioration by heat than common fire-bars. If 

 the mere fact of admitting air to the hydrocarbons at the moment 

 of their generation, and in minutely divided lines, is sufficient to 

 insure their combustion, surely nothing could do so more effect- 

 ually than this arrangement. But the result. 



A~large quantity of fuel being placed upon the incandescent car- 

 bon in the furnace, we have, after the expiration of a few seconds, 

 a splendid display of white flame, not entirely smokeless, but com- 

 paratively smokeless, unless the quantity of air admitted is very 

 large ; white flame and intense heat, evidence of the precipitation 

 of the carbon particles and of their combustion after precipitation ; 

 smoke-burning, not smoke-prevention; greatly increased tem- 

 perature of the furnace-door, evidence of increased radiation of 

 heat. But, as I said before, in almost all cases a loss of efficiency in 

 the furnace, a reduction in the absolute temperature of the flame. 

 Was Mr. Williams deceived by that radiant heat ? I cannot avoid 

 the conclusion that he was in some cases at least. But the fur- 

 naces adopted with economical results contained elements not yet 

 described. The ash-pit was divided into 3 chambers by 2 vertical 

 sheet-iron partitions, made fast to the longitudinal bearers in 

 such a manner that all air entering it at the central chamber must 

 pass through the arched bars, while that entering by the two side 

 chambers reaches the fuel in the ordinary manner. Now, observe 

 the difference : Here we have a long central fire-chamber open to 

 the air only at one end. The air before entering the fire-cham- 

 ber passes over the surface of highly heated sheets of iron, trav- 

 erses in turn the cross-pieces of the little arched liars and the 

 heated surface of the ribs. Even with this simple change the 

 results are, I believe, in all cases, altered from failure to success. 

 A heating effect has been obtained where a cooling effect only 

 could be produced before. 



To sum up my own observations on this subject, I find : (1.) 

 That the admission of cold air in quantities sufficient for the com- 

 plete combustion of the gases in ordinary furnaces is attended 

 with a loss of efficiency in all cases, even if that admission takes 

 place in finely divided streams immediately over every portion of 

 the fuel from which the gases are rising. Radiant heat, and con- 

 sequent temperature of the furnace door, are enormously in- 

 creased; smoke, however, is considerably reduced. (2.) That 

 by the comparatively slow motion of air over heated surfaces, and 

 its consequent rarefaction and increase of velocity when issuing 

 from the orifices of the arched bars, a much more perfect chemi- 

 cal union is insured. The flame is not so luminous, but a higher 

 rate of efficiency is obtained. Radiant heat is decreased, the fur- 

 nace door is rendered less hot, and smoke is more perfectly 

 prevented. The old Cornish system of dead-plate firing, when 

 conducted very carefully, and in such a manner that the incan- 



