October 16, 1885.] 



SCIENCE. 



343 



can be most naturally and economically supplied 

 by the large Bessemer works in our manufacturing 

 centimes, the magnitude of whose operations enables 

 them to profitably employ the best talent and ma- 

 chinery and to produce at the smallest cost, geo- 

 gTapliical conditions occasionally favor the erec- 

 tion of small steel works ; for example, where a 

 special demand for Bessemer steel, too limited to 

 warrant the erection of full sized works arises in a 

 place remote from aU existing Bessemer works, 

 and where pig-iron is cheap, owing to the imme- 

 diate vicinity of iron blast-furnaces. Here it might 

 be cheaper to convert the local pig-iron into steel 

 at local works, even if they be so small that the 

 cost of treatment is somewhat liigh, than to trans- 

 port the iron to distant works, have it there con- 

 verted into steel, and then bring it back to the 

 starting point. 



To meet such cases several small and compara- 

 tively cheap arrangements of the Bessemer plant 

 have been designed, and one of these, the Clapp- 

 Griffiths, has kindled quite a glow of interest 

 in this country, which judicious and energetic 

 fanning and puflSng bid fair to convert into a 

 veritable craze. Since the arrangement aims 

 at a comparatively small output, some sacrifice 

 of rapidity and cheapness of working are properly 

 made in order to diminish the cost of the plant 

 itself. The costly rotating converters of the ordi- 

 nary plant are replaced by the cheap Swedish 

 stationary converter. The blast is introduced, not 

 as in the ordinary converter at the bottom of the 

 deep bath of metal, but near its upper surface, so 

 that, having little resistance to overcome, blast at 

 low pressure, and hence furnished by cheap blow- 

 ing apparatus, may suffice. Moreover, towards the 

 end of the operation and while the steel is being 

 tapped out of the converter, the blast is admitted 

 very slowly, to avoid ' over-blowing ' ; and a hole 

 is provided in the shell of the converter at such a 

 height that the slag runs out through it during the 

 converting operation. I mention these latter de- 

 tails because they are supposed to play an import- 

 ant and unlooked-for part in the chemistry of the 

 process ; indeed, the plant itself, of good but not 

 remarkable design, is of interest to the readers of 

 Science, chiefly because it is claimed that it re- 

 moves sihcon more uniformly and completely than 

 the ordinary Bessemer plant does. The effect of 

 phosphorus in rendering steel brittle has long been 

 known to increase with the proportion of carbon 

 present. A percentage of phosphorus which would 

 have httle effect on steel containing only 0.15^ 

 carbon would change steel with 0.5^ carbon from 

 a valuable ductile metal to a worthless brittle one. 

 While some have maintained that silicon counter- 

 acts the effects of phosphorus, many have long be- 



lieved that like carbon it greatly exaggerates them. 

 This belief is somewhat strengthened by the fact 

 that phosphoric samples of Clapp-Griffitlis steel, 

 when low in both carbon and silicon, are surpris- 

 ingly ductile. But whether their ductility be due 

 merely to low carbon or to the combination of low 

 carbon with uniformly low silicon, it is interesting 

 to inquire whether it be due to conditions which 

 can be regularly imitated in the large scale Besse- 

 mer works ; if it be, then, since the magnitude of 

 their operations enables them to produce more 

 cheaply, an important if not the chief ulterior re- 

 sult of the development of the Clapp-Griffiths 

 plant and practice will probably be to teach the 

 metallurgists of our large works how to produce 

 more uniformly ductile steel from given pig-iron, 

 and, aiming at a given degree of ductility, to em- 

 ploy more phosphoric, and hence cheaper, pig-iron 

 than heretofore. Let us, therefore, consider the 

 explanations which have been advanced of the re- 

 sults obtained in the Clapp-Griffiths practice. 



1. The uniformly thorough desilicidation has been 

 attributed to the unusually low blast pressure em- 

 ployed. While it is conceivable that, by increas- 

 ing the tendency of carbon and oxygen to disso- 

 ciate this might favor the oxidation of sihcon, 

 this explanation seems far fetched and insufficient. 

 But, if low blast pressure be the cause, the ordi- 

 nary Bessemer works can employ it by making 

 their vessels wider and the bath of metal shallower 

 than at present. 



2. It has been attributed to admitting the blast 

 near the top instead of at the bottom of the bath 

 of metal ; this is supposed to cause a local excess 

 of oxygen in the upper part of the bath with the 

 formation of iron-oxide (the copious evolution of 

 red smoke at the commencement of the operation 

 is adduced as evidence of this) which is supposed 

 to attack silicon rather than carbon. But the early 

 appearance of iron-oxide in the flame of the Clapp- 

 Griffiths converter may indicate, not that it is a 

 more active, but actually a less active agent than 

 in the ordinary converter (I will not pretend to say 

 what its true significance is). If we confine our 

 ideas to a very minute quantity of metal immedi- 

 ately in front of any one tuyere of the ordinaiy 

 converter we realize that, in this restricted space, 

 oxygen is nearly, or perhaps quite, as much in ex- 

 cess as it is in a similar space in front of a Clapp- 

 Griffiths tuyere. If iron-oxide forms in the latter, 

 it will also, and perhaps to an equal extent, in the 

 former. We do not see it escaping from the ordi- 

 nary converter, probably because it is reduced by 

 the carbon and sihcon and slagged by the sihca it 

 encounters in its long upward path through the 

 superincumbent metal, wliile in the Clapp-Grif- 

 fiths converter, dragged along by the blast, its 



