I nearly agree with Mr. Hixon, and do not think it practical to smelt with much less than 5 per cent. coke continuously; but there is a great saving between the amount of coke used with a moderately heated blast and cold blast. Regardless of either hot or cold blast, however, the fuel consumption depends very much on the character of the ore to be smelted, the amount of matte-fall and grade of matte made. It is not always advisable or necessary to use hot blast for a matting furnace; that is, where the supply of sulphur is limited. It may then be necessary to use as much fuel in the blast furnace to prevent the sulphur from oxidizing as will be sufficient to furnish the heat for smelting. Such conditions existed at Silver City, N. M. , at times, after our surplus supply of iron and zinc sulphide concentrates was used. I understand that they are now short of sulphur there, on account of getting a surplus amount of oxidized copper ore, and are only utilizing what little heat the slag gives them, without the addition of any fuel on top of the forehearth. Before closing this, which I intended to have been brief, I wish to call your attention to a little experience we had with alumina in the matting furnace at Silverton, Col., where I was acting as consulting metallurgist. The ore we had to smelt contained, on an average, about 20 per cent. Al₂O₃, 30 per cent. SiO₂, with 18 per cent. Fe in the form of an iron pyrite, and no other iron was available except some iron sulphide concentrates containing a small percentage of zinc and lead. The question naturally arose, could we oxidize and force sufficient of the iron into the slag, and where should we class the alumina, as a base or an acid? My experience in lead smelting led me to believe that Al₂O₃ could only be classed as an acid in the ordinary lead furnace, and that it would be useless to class it otherwise in a shallow matting furnace; and E. W. Walter, the superintendent and metallurgist in charge, agreed with me. We then decided to make a bisilicate slag, classing the alumina as silica, and we obtained fairly satisfactory results. The slag made was very clean, but treacherous, which was attributed to two reasons: First, that it required more heat to keep the alumina slag liquid enough to flow than it does a nearly straight silica slag; and, second, that we were running so close to the formula of a bisilicate and aluminate slag (about 31½ per cent. SiO₂, 27 per cent. Fe, 20 per cent. CaO, and 18 per cent. Al₂O₃, or 49½ per cent. acid) that a few charges thrown into the furnace containing more silica or alumina than usual would thicken the slag so that it would then require some extra coke and flux to save the furnace. At times the combined SiO₂ and Al₂O₃ did reach 55 and 56 per cent. in the slag, which did not freeze up the furnace, but caused us trouble. PART IX LEAD REFINING