STEEL.—Bessemer Process. Gases Evolved. 8G!)
the pig is burnt to carbonic oxide, this, as far as Iam aware, has never been proved to be the case.It will be seen from the analyses that this turns outto be only partially true, and that the compositionof the gases evolved affords us an insight into thenature of the process going on that could not besupplied by analyses of the metal and slag atsuccessive stages.
The gas for analysis was collected by means of along iron gas pipe, on the end of which a swan-necktrumpet mouthpiece of fire-clay was fixed, whichwas inserted in the neck of the convertor after ithad been turned up. This generally got cloggedup with slag during the experiment, and thereforeserved only for one blow. It was allowed to dipinto the vessel, so as to be certain that no air couldbe drawn into the tube by induction. To the otherend of the pipe glass tubes were attached at particularperiods of the blow when the gas was required, andsealed up with the blow-pipe before being removed.A constant stream of gas was allowed to rush throughthe tube the whole time; and as this was done fortwo minutes before the first sample of gas wascollected, it may be safely assumed that the whole ofthe air was swept out, and that the gas analyzed wasa true sample of that produced by the direct actionof the blast upon the metal and its constituents.It was found that during the first part of the blowthe gas would not light at the end of the iron tube,but that from about the commencement of the“boil” to the end of the blow it burnt with the paleblue flame characteristic of carbonic oxide.
Samples of gas were taken every two minutes,and subjected to analysis with the subjoined results.The gases were analyzed in a modified form of theapparatus, described by Professor Frakkland , foranalyzing the gases incident to water analyses, whichwas found extremely convenient. Samples 3 and 5were analyzed by different methods; the carbonicoxide in one case being estimated by absorptionwith a hydrochloric solution of sub-chloride ofcopper, and in the other by explosion with deton-ating gas and oxygen, and subsequent absorptionby potash, after the method of Bunsen. It will beseen that the results agree closely:—
Sample 1.—This was taken two minutes from thestart of a blow which lasted eighteen minutes. Itcontained—
Carbonic acid,. 10'71
Oxygen,. '92
Carbonic oxide,. none.
Hydrogen {not estimated), I
Nitrogen,. f ..
100-00
This I consider to be an unlooked-for result,because it is usually assumed that the carbon isburnt directly to carbonic oxide at all stages of theblow. Taking the nitrogen in this sample at 87 percent., this corresponds to 23-05 volumes of oxygen.10-71 volumes carbonic acid contain 5-35 volumescarbon vapour, which, referred to air as unity,represent 4-43 parts by weight of carbon. Thecarbonic acid contains its own volume of oxygen,
which, deducted from 23-05, leaves 12-34 volumes= 13-62 parts by weight of oxygen, to combine withother elements than carbon. Now, ordinary English Bessemer iron contains little else for oxidation thancarbon, silicon, and iron; the manganese beingusually small in quantity, traces only of the sulphurdisappearing, and the phosphorus remaining un-touched. Of the three oxidizable elements, it isknown from analyses of the slag that the iron is notburnt to any extent till towards the end of the blow,so that we have only to consider the carbon andsilicon. The 13-62 parts of oxygen may, therefore,be assumed to be combining with silicon alone, sothat as 32 parts of oxygen unite with 28 silicon, orin the ratio of 8 to 7, we have 11-91 parts siliconbeing oxidised at this stage, along with 4-43 parts ofcarbon; thus showing that, though both elements arebeing oxidised, it is the silicon which is disappearingmost quickly. Now, this is exactly what an analysisof the metal proves, for a charge containing, onrunning into the convertor, 3'57 per cent, carbon and2-26 per cent, silicon, was found by Barker in a fewminutes to have lost -53 per cent, carbon and 1-305per cent, silicon. As an example of the nature ofthe gas evolved at a point between the times whenthis first sample was taken, and that obtained twominutes later, I give here analyses made by myfriend W. Thorp, F.C.S., of the Rivers’ CommissionLaboratory, of two consecutive tubes of gas which Icollected, four minutes from the commencement ofa blow which, from the convertor being new, or theiron cold, lasted twenty-nine minutes—
No. 1 tube
No. 2 tube
contained
contained
Carbonic acid,.
8-94
9-296
Oxygen,.
•916
•113
Carbonic oxide,.
•078
•044
Hydrogen and nitrogen,.
90’066
90-544
100-000
100-000
Samjtle 2, taken four minutes from commencement,contained—
Carbonic acid,.’.. 8"57
Carbonic oxide,. 3'95
Hydrogen,. -88
Nitrogen,. 86-58
100-00
It will be noticed that carbonic oxide is graduallyincreasing, while carbonic acid is decreasing inquantity.
Sample 3, taken six minutes from start, contained—
By Absorption.
Carbonic acid,.
7-89
8-2
Carbonic oxide,.
4-65
4’52
Oxygen,.
absent.
absent.
Hydrogen,.
—
2'00
Nitrogen,.
87-55
85‘28
100-00
100-00
Dissecting these results, after the manner of thosefor No. 1, we find that 85'28 nitrogen correspond to22-61 volumes oxygen. The carbon is combined with