runs very thin; unlike the more siliceous slags, however, which pass through a plastic stage, it sets suddenly on cooling, and becomes quite hard and strong while still at a bright red heat. For anything like good work the composition of these slags, containing 14 to 15 per cent. of metallic zinc, must be kept within very narrow limits for both silica and lime, the limits for the former lying between 231 and 26 per cent., and for the latter 12 to 16 per cent. If the slag contains less than 10 per cent. lime, it carries away matte in suspension and the losses are increased; if there is more than 16 per cent. lime, the slag becomes mushy and gives rise to bad, irregular smelting. If the silica falls to 21 or 22 per cent., the losses of lead and silver in the slag are at once largely increased; if, on the other hand, it rises above 26 per cent., the lead loss increases rapidly without a corresponding increase in the silver loss, which, in fact, is generally lower on account of the larger amount of matte made. The non-formation of matte when smelting an ore charge containing 6 to 8 per cent. of sulphur and producing a normal slag, and its immediate appearance as soon as the silica contents of the slag rise above the normal limit without any change in other conditions, is one of the most remarkable facts brought out by the novel work of Mr. T. J. Greenway on these zinc ores. There seems to be no doubt whatever about the fact, which has been repeatedly observed whenever the ratio of silica to iron in the charge increased either by accident or design. The explanation is by no means obvious, but a clue is afforded by the fact that the matte so formed is not the ordinary iron matte but one very high in lead and zinc, and containing only a small percentage of iron. Further, it has been noticed that a rise in silica is always accompanied by a rise in lead, even though the silver contents of the slag remain low. But for this fact it might be argued that the formation of matte upon the slag becoming more siliceous was due entirely to more perfect separation of heavy metallic sulphides dissolved in the slag through the decreased specific gravity caused by the added silica; and, no doubt, some action of the kind does take place, as such siliceous slags are found to be somewhat more free from sulphur than those ordinarily made. The increased amount of lead in the slag, however, indicates that some other influence is simultaneously at work, and the explanation would seem to be somewhat as follows: The atmosphere in the upper part of the furnace being oxidising or, at all events, non-reducing, a portion of metallic lead is reduced there by direct reaction between the oxidised and unaltered portions of the charge. In practice, somewhat more than one half of the original sulphur contents of the ore is eliminated during roasting, and it may be reckoned that the bulk of the lead is converted into oxide and sulphate, while most of the zinc remains as sulphide. The large excess of lead oxide in the "carbonate ores combines with the silica in the same ores to form silicates, which (together with the unaltered zinc sulphide of the roasted ore, some lead sulphide, and the fluxes) remain without change, except the elimination of CO, from carbonates, until reaching the zone of fusion where they react upon each other. When the furnace is running as usual and producing a normal slag, the reactions which take place may perhaps be approximately represented by the following hypothetical equation : 2(3PbO. 4SiO2) + 4ZnS + PbS + 4Fе,03 + 4CaÒ + 202 1724 = 4(2FeO. SiO2) + 2(2CaO. SiO1) + 2(2ZnO . SiO2) + 5SO2 + 6Pb Slag (85 per cent. of total) = 25 per cent. SiO. Supposing now the ore and iron flux simultaneously become more siliceous (as frequently happens) without any other change in their composition or in the furnace manipulation. Matte is made, as already seen, and the loss of lead in the slag at once goes up. The altered reaction may, perhaps, be shown by the following hypothetical equation: = Ore. Fluxes. Blast. (5PbO. 9SiO2) + 4ZnS + PbS + 3FeO3 + SiO + 4CaÒ + O2 1660 3(2FeO. SiO2) + 4(CaO. SiO2) + PbO. 3ZnO. 3SiO2 + PbS. ZnS Slag (85 per cent. of total) 29 per cent. SiO2 + 3SO2+3Pb. = Matte. The above equations must not be taken as representing definite equivalents, but they may serve to explain the nature of the change which probably takes place when the charge becomes more siliceous. The peculiar practice at these works, though well suited to the abnormal conditions to meet which it has been adopted (including in particular the exceedingly high percentage of zinc), is by no means to be recommended for such conditions as ordinarily prevail in "custom" lead smelting works in other parts of the world. Various other details are given in Table XXX., No. 3. AMERICA. The author not being at liberty to give detailed descriptions of specified works visited has selected as typical examples of American practice the data given in columns Nos. 1, 4, 5, and 6 of Table XXX. The comparative figures there given will, it is hoped, be found useful, and particular attention should be drawn. to the way in which the figures for "hearth efficiency" and "tuyere efficiency" rise in proportion to increased blast pressure. The following examples of unusual conditions may be quoted:Tombstone, Arizona.*-Here, during the years 1883-1887, concentrates from the tailings of pan amalgamation mills containing 50 per cent. Pb, 50 ozs. Ag, and about oz. gold per ton were smelted together with argentiferous manganese and siliceous ores in a round water jacketed cupola, after "bricking" with the finest slime from the settling pits, which itself contained 8 to 10 per cent. lead and a little silver. The total quantity treated was about 40,000 tons, the average lead contents on the charge being only 6-7 per cent. by dry assay, and the base bullion produced averaging 329 ozs. Ag. per ton; yet the losses in slags averaged only 164 per cent. lead and 1.8 ozs. silver per ton, though running much higher when the lead on the charge ran down to only 4 per cent. It was necessary to charge back the bullion produced in order to keep sufficient flow through the furnace. It appeared that good work could be done when the lead produced was not less than 7 per cent. of the total weight of the fused furnace products. The slags made were very fluid (owing to the manganese present, which at times ran as high as 43 per cent.) their average composition is given in Table XXVI. The fuel consumption was high-viz., 20 per cent.owing, no doubt, in part to the oxidising influence of the free oxygen in the manganese ore. No difficulty was experienced except that the perfect fluidity of the basic manganese slag permitted all impurities to settle out and form crusts on the lead. This tendency was increased by the absence of sulphur to form a matte, and by a very small quantity of arsenic which formed a heavy infusible speiss with reduced iron and small quantities of copper, nickel, and antimony present. The total cost of smelting, including limestone flux, was 39s. 11d. per ton of ore. At Casapalca, Peru, 14,000 feet above sea level and 2000 feet above timber line, rich lead and silver ores are smelted in water jacket furnaces with imported English coke, costing $40 per ton. The ores consist of—(1) Lump galena with 45 per cent. Pb and 50 ozs. Ag per ton; (2) jig concentrates with 158 ozs. silver per ton; and (3) Vanner concentrates with 209 ozs. silver per ton. All the concentrates contain 6 to 12 per cent. of lead and v. Church, Trans. A.I.M.E., vol. xv., p. 601; vol. xxiv., p. 559. |