We produced, as nearly as practicable, three qualities of converted metal, say A, with half per cent. of carbon, B, with one per cent, and C, with one and a-half per cent.; we also made pure iron upon which we could absolutely rely. These four qualities, accurately analysed, were kept in separate bins; the analyst who gave the order to the steel melter to make two or three tons of steel of any precise and predetermined degree of carburation would, say for example, weigh 41 1/4 lb. out of bin A, and put 8 3/4 lb. from bin B into it, thus making the 50 lb. charge, always using the nearest of the three qualities to the one required, and making it a little milder or a little more highly carburised as desired. Most minute differences could thus, at all times, be made with unerring certainty by the simple fusion in a crucible of two metals, the carburation of which had, in each case, been tested by careful analysis. The mixing of these accurately- ascertained qualities in definite weights while in the granulated state resulted in the production of a quality the exact mean of the known constituents of the two qualities mixed. It gave a more certain and a more accurate result than could possibly be obtained on the old system of crucible steel-making, where judgment by the eye took the place of accurate analysis and the weighing machine, as used in my system. Hence it was an undeniable fact that we could, and did, produce commercially crucible cast steel of great purity, and of any precise and predetermined degree of carburation, with greater accuracy than was obtained by the method employed to produce crucible steel in Sheffield.