way, and with one exception the method of manufacture was pretty much
the same now as at that early date. The improvement was the employment
of gas furnaces instead of the old coke holes for melting. Important
economies had resulted from this introduction. Where before it required
3 tons of coke to melt 1 ton of steel, the same thing was now done with
35 cwt. of very poor slack. Though it was apparently easy to make
crucible steel castings, it was not in reality easy to make a true
steel, that was to say, to make a metal that contained only the correct
proportions of carbon and silicon and manganese. The only real way to
make crucible castings of true steel was to melt the proper proportions
of cast steel scrap with the proper amounts of silicon and manganese to
produce that chemical composition which was known to be necessary in
best castings. It was in consequence of this difficulty that many makers
resorted to the addition of hematite pigs. The Bessemer process was used
much more extensively upon the Continent than in this country in the
manufacture of castings. It seemed likely that Mr. Allen's agitator for
agitating the steel in the ladle so as to remove the gases would be
taken up largely for open-hearth castings and open-hearth mild steel, as
it had a wonderful effect. The Wilson gas producer, working in
conjunction with the open-hearth furnace, had recently produced some
extremely wonderful results. In some large works, steel was by its aid
being melted from slack which was previously absolutely a waste product.
The method of making open-hearth steel castings might be varied greatly.
The ordinary method generally practiced in this country was a
modification of the Terre Noire process. The moulds employed were only
of secondary importance to the making of the steel itself. Unless the
mould was good, no matter how good the steel was, the casing was
spoiled. The best composition which had been found for moulds was that
of a large firm in Sheffield, but unfortunately it was rather expensive.
A good steel casting ought to contain about 0.3 per cent. carbon and 0.3