88 ' Bolkrfar Sitam.—Btcil Furnace, 



is higher when deep than when fhallow, it is found to be of fome confequence In the dift!!- 

 lation of fermented liquors, that the charge fliould not be deep. But this appears to be left 

 on account of any fuppofed rapidity of evaporation, than becaufe the flavour of the product 

 may be altered, by fuch increafe of temperature. 



To afford a regular fupply of water to the boiler, without depending upon the engine-keeper, 

 is certainly an objcdt of no difficulty, and is adtually done in many fteam-engines, by means 

 fimilar to that fuggefted by my correfpondent. In Mr. Kier's engine, defcribed in this 

 Journal, I. 422. the water is admitted through a valve, kept fliut by a float, which falU as 

 the water fubfides by evaporation. On which particular, I may here add, that the body im- 

 merfed in the water is not a fimple float, but a piece of ftone that hangs from the valve,. 

 which is affifled in its tendency upwards by a balance-lever and counter weight. When the 

 ftone is immerfed to a certain depth, the Weight attached to the lever predominates, but when 

 the water falls below that depth, the flone exerts more of its gravity, and keeps the valve. down„ 

 till the requifite fupply of hot water has flowed in. 



2i/?4in7, 1799. 



W. N. 



X. 



Defirlptian of the Furnace for converting Ear-iron into Steel, ly Mr. JOSEPH CoLLIER*. 



H E furnaces for making fleel are conical buildings, about the middle of which are two 

 troughs of brick, or fire-flone, which will hold about four tons of iron in the bar. At the 

 bottom is a long grate for fire. The fleel furnace, however, is not well adapted with 

 ilefcription. 1 fhall, therefore, avail myfelf of an accurate account, which was communicated 

 to me by a gentleman converfant with the manufadlure. 



A layer of charcoal-duft is put upon the bottom of the trough, and upon that a layer of 

 bar-iron, and fo on alternately, until the trough is full. It is then covered over with clay, 

 to keep out the air, which, if admitted, would efFedlually prevent the cementation. When 

 the fire is put into the grate, the heat pafTes round by means of flues, made at intervals, 

 by the fides of the trough. The fire is continued until the converfion is complete, 

 which generally happens in about eight or ten days. There is a hole in the fide, by which 

 the workmen draw out a bar, occafionally, to fee how far the tranfmutation has proceeded. 

 This they determine by the bliflers upon the furface of the bars. If they b« not fufficiently 



* Extraftcd from his Oifervaiions on Iron and Slitl, in the Manchefter Memoirs, V. 109. — The paper itfelf 

 contains a concife and clear, though in fomc refpects imperfect, account of the ufual procelTes of the fmelting 

 and purifying this ufeful metal. This author, who, inadvertently, blames M. Fourcroy and myfelf for this ine- 

 vitable confequence of abridgment in our elementary writings, has himfelf fliewn that the talk of defcriblng 

 complex operations in a few words, is fufficicntly difficult to afford a claim upon the public indulgence. — On 

 Suel, fee Philof. Journal, I. z 10. 248. 381. 468. 575; II. 64. loz. 



changed. 



