whole question of the supply from abroad of pure charcoal pig-iron, and had practically tried the famous Algerian iron from Bône and other mines, and also Indian, Nova-Scotian, Styrian, and Swedish pig-irons. Among the latter, I found on analysis, to my astonishment, that certain brands of charcoal pig, which, when delivered in Sheffield, cost only £6 to £7 per ton, were, when decarburised by my process, superior in purity to some of the highest brands of Swedish bar-iron, costing in Sheffield from £16 to £24 per ton. One Swedish brand of pig-iron in particular,costing £6 10s. delivered in Sheffield, was capable of making malleable iron by my process more pure than the far-famed Dannemora L bar-iron, worth £30 per ton in Sheffield, and with which particular brand the small malleable iron gun which I exhibited in May, 1859, at the Institution of Civil Engineers, was made. The analysis of this by Mr. Edward Riley has already been given (page 182 ante).

It will be conceded that if we obtained malleable iron of this extreme purity by my process, steel of very high quality could also readily be produced from that particular class of pig-iron.

Thus fortified by practical working and by actual analysis, and also by the purchase of a large consignment of this pure charcoal pig, we laid ourselves out at the Sheffield works for the production of high class tool steel, which we put on the market at l5s. or 20s. per hundredweight below the ordinary trade prices for this article. My process, so admirably adapted for the production of large ingots, was not so well fitted to make a great number of the 2 3/4 in. square ingots generally used in the Sheffield steel trade for tilting into small bars, which particular size of ingot had all its long-established trade rules and prices connected with it. So we determined to convert our pig into steel in large quantities, and to pour the converted metal into an iron cistern filled with water, in order to granulate the whole charge, and avoid all costs of moulds, casting, etc. By this means, and by the blending of different charges in definite proportions, we insured the production of steel of any desired temper, or degree of carburation, with an accuracy wholly unattainable by the old crucible system. For it must be borne in mind that in the ordinary crucible process the steel melter has to deal with bar-iron that has been subjected for several days, in a very large closed box or chamber, to the action of charcoal powder at a high temperature, during which treatment the iron bars absorb about one per cent. of carbon, more or less, dependent on time and on temperature. The amount of absorption depends also on the relative positions of the bars in the converting-box; hence, when the bars are thus converted into blister steel, it is almost impossible that the ends and the middle of any particular bar should be equally carburised, or that bars occupying different positions should absorb an equal quantity of carbon. After the withdrawal of the bars from the converting-chest, they are broken into short pieces for the melting crucible. Now the only mode of telling how far each piece of the broken bars has been carburised is to examine the crystalline fracture by the eye, and thus class and assort the various fragments for each quality of steel. It is wonderful how accurately a clever practised steel melter will judge of the state of carburation of the metal; but his judgment, after all, can only be approximate. Such visual determination is not like measuring or weighing the constituents of a mixture. Crucible steel is made in separate pots of from 40 lb. to 50 lb. each, and the steel maker cannot afford to make forty- five separate quantitative analyses of every ton of steel he turns out. Even if he could do so, after he had made the metal into ingots, he would not be more secure, since he could not alter the ingot when once cast. As a matter of fact, the precise degree of carburation of each 50lb. of steel produced in the old crucible process depended on the judgment of a man looking at the crystallised fracture of each piece he put into his crucible; and all must agree that it is highly creditable to those engaged in this mere guesswork that they got as near as they did to the quality required.

In the manufacture of tool steel, on the system which I laid down at my Sheffield works, we entirely eliminated this source of inequality, by dispensing with ocular examination of a crystalline fracture, which is subject to numerous modifications in character, from causes other than its precise degree of carburation. We converted five tons of pig iron at one charge, and having granulated it by pouring the molten steel into a cistern of water, we had this quantity of shotted metal in a condition that was still practically fluid as far as the power of mixing was concerned. If each granule weighed, on an average, seven grains, we had in our 50 lb. crucible 50,000 separate pieces, the precise degree of carburation of which had been ascertained by careful quantitative analysis of the whole five tons, which analysis we could afford to make -- and did make -- very carefully.