Various gold bullion making methods have been suggested for effecting the elimination of the zinc and other base metals. The chief ones are:

1. Direct fusion with fluxes.
2. Roasting, followed by fusion.
3. Treatment with sulphuric acid, followed by fusion.
4. Lead fusion and cupellation (Tavener process).
5. Volatilisation of the zinc, followed by fusion.

Direct Fusion

The method of Direct Fusion was used in the early days of the cyanide process, but is now superseded, except perhaps on certain isolated mines. If the zinc has been imperfectly separated from the slimes, considerable quantities of nitre must be added. Clay pots are used, and in addition to nitre, other fluxes are added, such as borax, carbonate of soda, sand, and fluorspar.

The slag, which consists of silicates of zinc, soda, corrodes the pots rapidly. Large quantities of zinc oxide are given off as fumes, forming thick crusts in the flues, and evil-smelling products of decomposition of the cyanides are also evolved. The bullion produced by this method varies in colour from a pale yellow to a brownish linnet-green, and is about 650 fine, but cannot be obtained uniform in composition, so that accurate assays are difficult to obtain.

The results of analyses made on three ingots of bullion produced in this way:

The slags obtained in this way are always rich in gold, part of which is sometimes in the form of shots, and may be recovered by crushing and panning. The crushed slag is then fused again with the addition of granulated lead, or of litharge, when all the gold is concentrated in the lead. If the lead thus obtained is granulated, it can be used again until rich enough to be worth cupelling.

When by the careful use of a 40-mesh screen, followed by the squeezing of the slimes in a filter bag, the zinc has been almost all eliminated by mechanical means, very little nitre is required. According to James the zinc can be reduced to about 1 per cent., and bullion obtained 960 fine in gold and silver without the use of nitre, roasting, or acid. The dry slimes are fused in graphite pots with about half their weight of borax, a little soda carbonate, and sand or fluorspar, if necessary. The proportion of the fluxes must, however, be determined in each case by experiment. It is necessary to form a fluid slag.

Roasting followed by Fusion

The dried slimes are spread in a thin layer on iron trays, and heated to a barely-perceptible red heat in the flue of a furnace, or above a special grate under a hood. The carbon, zinc, arsenic, ignite readily, being in a very fine state of division, and roasting proceeds regularly without much stirring, which tends to cause loss through dusting.

Dense fumes of zinc oxide are given off, and the residue consists chiefly of the oxides of lead and zinc with gold and silver and a variable quantity of sand. About 30 per cent, of nitre is usually mixed with the wet slimes to aid in the oxidation of the base metals. Fusion with fluxes is carried out as in the previous case. The bullion produced is said to be about 800 fine. The method, like the ore previously described, appears to have been discarded, owing to the loss by dusting, which was always an unknown quantity.

Acid Treatment

The washed and filtered slimes are passed without being dried to the acid vat. Sometimes, as at Brodie, the slimes are transferred direct into the acid vat without filtering and washing them, and are there allowed to settle partially. The supernatant liquid is then siphoned off into a settling tank. Acid is then added, a little at a time, either in concentrated or dilute form. Violent action ensues, and the fumes are highly poisonous, sometimes containing arseniuretted hydrogen, and must be carried off by a good draught.

The effervescence is at first great and the vat must be deep to prevent frothing over. The charge is sometimes heated by steam, and must be occasionally stirred. The vat may be made of wood, lined with lead, or of enamelled iron.

J. E. Thomas and G. W. Williams have shown that bisulphate of sodium is a cheaper and more convenient solvent for zinc than sulphuric acid is. In a large scale experiment at the Simmer & Jack Mill, a solution of about 34 per cent, of commercial bisulphate, equivalent to a liquid containing 9 per cent, of sulphuric acid, was used in cleaning up with satisfactory results. No arseniuretted hydrogen was produced, and there was no tendency to boil over. The residual solutions contained no free acid.

After treatment with acid the slimes are settled, washed with hot water, and separated by a vacuum filter or a filter press. They are then dried, calcined and fused in the usual way. The cost of recovery per ounce of gold at the Standard Co.’s works at Brodie, in California, is as follows :—

The following analyses of slimes, after treatment with acid:

In Nos. II, and III. the treatment with sulphuric acid was very incomplete. Ingalls states that in fusing these slimes No. I, would require about 50 per cent, of soda carbonate and 25 per cent, of borax, but silica in addition would be required by the others.

The following analyses show the composition of acid-treated and calcined precipitate obtained:

In No. IV., made by A. Whitby, the organic matter, including insoluble cyanide compounds, was not determined. In No. V., given by W. A. Caldecott, some lead existed as sulphate. The slimes were from the Robinson Deep Mill. Nos. VI. and VII., made by G. W. Williams, are of slimes from the East Rand Proprietary. No. VI. is before and No. VII. is after calcination. No. VI. also contained from 7 to 10 per cent, of ferrocyanides (K4FeCy6). About half the sulphur was present as PbS, which was oxidised in the ensuing calcination.

The charge is melted in graphite crucibles, and poured into conical moulds. The slag is fluid and glassy and contains from 20 to 200 ozs. of gold per ton, which is recovered by fusion with litharge. The bullion prepared in this way may be as high as 985 fine in gold and silver, and should not be lower than 850 fine. It is often brittle from the presence of zinc. The loss in roasting and subsequent handling is estimated by Bettel to be from 0.01 to 0.025 per cent., or less than ¼d. per oz.

W. Bettel describes a small reverberatory furnace for melting gold slimes, which he considers cheaper and better than crucibles. It resembles an English cupellation furnace, but has a rectangular instead of an oval opening through which can be removed a brick-lined cast-iron pan with tap-hole and spout forming the bed of the furnace. The pan is supported on two iron bearers built into the brickwork. The calcined slimes and fluxes are mixed, moistened, and compressed before being charged in. The loss by dusting is stated at only 0.04 per cent. The slag melts and is rabbled with an iron tool and tapped off, leaving the gold unmelted, and a fresh charge put in. After several charges have been treated, the heat is raised, and the gold melted and tapped out.

The method of cleaning-up as follows:

The slimes are transferred direct in buckets from the zinc boxes to a filter vat, 6 feet in diameter, and filtered by a vacuum-pump through closely-woven canvas, the clear liquid being returned to the boxes. Water is then added and pumped through until all soluble cyanides have been removed, and the slimes are then baled out (being weighed in the buckets during the process) into a sheet-iron tray. A closed vat, fitted with a stirring apparatus, is then charged with dilute (1 in 10) sulphuric acid, one pound of concentrated acid being added for each pound of moist slimes, the agitator is started, and the slimes added little by little through a hopper, the fumes being carried away by a 3-inch pipe. After all the slimes have been charged in, the agitation is continued for half an hour, and the hopper and sides are then washed down, the agitator washed and removed, and the vat filled with water and allowed to settle. No heat is used, except that caused by the mixing of strong sulphuric acid with water in the vat and by the action on the slimes. The clear liquid is siphoned off, and washing by decantation repeated four or five times, the washings containing about 13 grains of gold per ton of solution. The gold residues are then dried in enamelled iron dishes, roasted gently in sheet-iron trays, and then ground, mixed with fluxes, and transferred to the crucible. The charge fuses quietly with little fume, yielding bullion amounting to 50 to 60 per cent, of the weight of the slime. The average fineness in 1896 was 821.9, and the slags assay 23 ozs. per ton, but only 0.2 per cent, of the gold is locked up in this way.

When a filter-press is used, less gold is lost in the washing, and by the use of hot water, the zinc sulphate is more perfectly removed, so that the subsequent smelting is more easily and cheaply executed.

The composition of acid-treated precipitate in the Transvaal is given above, p. 301. It was formerly customary to fuse such material with nitre, sand, and borax, but in 1902 Johnson and Caldecott described the use of manganese dioxide instead of nitre. Fluxes used recently are as follows:

This gave bullion from 875 to 900 fine.