Some months since, the attention of the author was directed to certain statements to the effect that the gold-contents of foul or extremely dilute cyanide-solutions could not be effectively precipitated in practice by the usual zinc cementation of gold called the “zinc-method”.

Such statements caused all the greater surprise, because it is well known that this difficulty is not encountered in other well- known districts, as, for example, in New Zealand, where solutions containing less than 0.05 per cent, of KCy are regularly reduced in practice to below 1 grain of gold per ton, while in Mexico and the United States almost equally satisfactory cementations are not uncommon.

As a result of these representations, however, the author visited several plants at work in the Transvaal, and noted that while at some few mines the cementation was really excellent, at others the effluent, after treatment, was much too high in bullion-contents.

An investigation into the causes of these unsatisfactory results disclosed the fact that more attention to the close packing of the zinc-shavings in the extractor-compartments tended to improve the results, even with the most dilute solutions. The function of the zinc being one of surface, it stands to reason that a compartment well filled with zinc would be more effective than a half-filled one. Moreover, the quality of the zinc was inferior. This should be in long, tough, thin, light threads; light enough to ignite readily, and tough enough not to disintegrate easily into a slime. The coarse shavings too frequently employed appear to give the reverse of the best practice ; they are brittle, contain less surface, and take up more room. It is almost impossible to pack a compartment properly with such zinc, which permits the existence of channels through which the solution may ascend without coming into effective contact with the zinc at all.

The boxes which do the best work have compartments the depth of which is greater than any other dimension.

But even under the best conditions, solutions are encountered which give trouble when run through in the ordinary way. Such solutions are those charged with alumina salts, and those resulting from the treatment of weathered concentrates containing acid iron salts.

Very thorough investigation has been carried out on these solutions; and it is found that by simply giving them longer contact with the zinc they may be treated with complete success. With solutions the alkalinity of which has been neutralized by acid salts, it is advisable to add a slight amount of soda; but even in the most difficult case an alkalinity of 0.025 per cent., or ½ pound per ton, has been found perfectly effective.

Moreover, with extremely dilute cyanide-solutions, that is, solutions containing under 0.05 per cent, of free cyanide, perfectly satisfactory cementation may be carried out by passing the solution through the extractors at the rate of 1 ton of solution per twenty-four hours for each cubic foot of zinc-thread employed. But even when the solutions are very foul, with iron and alumina salts in solution, and containing no free cyanide, one is able in practice to reduce the bullion-contents to 1 or 2 grains per ton by passing the solution through the extractors at the rate of about ½ ton of solution per cubic foot per diem, the solution being in contact with the zinc for one and one-half to two hours.

This small amount of extractor-space, even with the most difficult solutions, as compared with that required for electric cementation, deserves to be noted.

Liquors derived from the treatment of cupriferous ores and concentrates, and containing 0.2 per cent, of copper in solution, were found to be quite amenable to a treatment similar to that which sufficed for the foul liquors from weathered concentrates.

The following are notes of experiments carried out in connection with the investigation above referred to.

The solution examined was one that was returned in a cyanide-works as containing 44 grains of gold per ton, after it had been twice passed through the extractors in the usual manner. An investigation showed that it contained a considerable quantity of arsenic and nearly 1 per cent, of soluble sulphates in solution, resulting from the treatment of some exceedingly acid weathered concentrates.

The solution was run through the extractors with the following modifications:

(а) The solution was run through a third time at the rate of 2 tons per cub. ft. of zinc per diem.

Assay before treatment, 2 dwts. 12 grs. per ton.
Assay after treatment, 1 dwt. 18 grs. per ton.

(b) There were added 4 lbs. of ammonium chloride per ton of solution, and it was run through at 2 tons per cub. ft per diem.

Assay before treatment, 2 dwts. 12 grs. per ton.
Assay after treatment, 1 dwt 17 grs. per ton.

The addition of the above amount of ammonium chloride had thus no beneficial effect at the above rate of extraction.

(c) There were added per ton of solution 4 lbs- of burnt lime, slaked and made into a cream, and the cementation precipitate was allowed to settle The rate of flow was not accurately determined, owing to the precipitate having reduced the rate of flow in pipe.

Assay before treatment, 2 dwts. 12 grs.
Assay after treatment, 14 grs.

(d) Solution (a) was run through again, for a, fourth time, at 2 tons per cub. ft. per diem.

Assay before treatment, 1 dwt. 18 grs.
Assay after treatment, 5 grs.

(e) The solution was run through again at especially slow rate of ½ ton per cub. ft. per diem.

Assay before treatment, 1 dwt. 11 grs.
Assay after treatment, 2 grs.

This shows a great improvement, and indicates that slow running is, of itself, more practically efficacious and cheaper than the addition of chemicals to assist cementation.

It was found, however, that after the run had been continued for some hours, the assay of the effluent rose to 12 and then to 21 grains per ton. An examination of the zinc showed that it had been thickly coated with a white incrustation which consisted mainly of ferrocyanide of zinc.

In all the above experiments there was only a trace of free cyanide or of alkali present in solution; and the formation of the incrustation pointed to the necessity of the presence of alkali or free cyanide in solution to prevent any such reaction.

(f) Some of the same solution as that used for the previous experiment was now run through the extractor at the rate of about 1/3 ton per cub. ft. of zinc per diem, after sufficient cyanide had been added to the solution to make it show 0.08 per cent, of free KCy.

Assay before treatment, 0 dwt. 19 grs. gold per ton.
Assay after treatment, 0 dwt. 2 grs. gold per ton.
KCy before treatment, 0.08 per cent.
KCy after treatment, 0.023 per cent.

This experiment shows that the addition of free cyanide causes a satisfactory precipitation. There was no incrustation formed on the zinc. The consumption of cyanide, however, amounted to 1 pound per ton; and to avoid this, if possible, the investigation was continued, and

(g) A similar experiment was carried out at the same rate of flow, but adding 4 lbs. of slaked lime per ton of solution instead of cyanide, the cemented sludge being allowed to settle.

Assay before treatment, 0 dwt. 19 grs.
Assay after treatment, 5 grs.
KCy before treatment, nil.
KCy after treatment, nil.

The zinc was washed free from incrustation in dilute solution of caustic soda prior to the commencement of this experiment, and sufficient new zinc was added to replace the consumption of previous runs. At the conclusion of the experiment there was a slight incrustation on the zinc in the first compartment only.

(h) Experiment as above, but with 4 lbs. of caustic soda per ton added in place of lime. No fresh zinc was added, nor was the zinc washed.

Assay of solution before treatment, 0 dwt. 19 grs.
Assay of solution after treatment, 0 dwt. 1 gr.

After the run the zinc was very black, and there was no incrustation.

(i) Experiment as above, but with less soda, and the rate of flow increased to ½ ton per cub. ft. per diem.

Assay of solution before treatment, 0 dwt. 19 grs.
Assay of solution after treatment, 0 dwt. 3 grs.

(k) In this experiment the run was continued for some days at ½ ton per cub. ft. per diem, and samples were taken every 12 hours. There was no caustic soda added, there being present only 1/5 lb. per ton of solution; gold was added to make the solution up to 2 dwts. per ton.

The solution assayed before treatment, 2 dwts. 0 gr.
The solution assayed after treatment, 0 dwt. 2 grs. to 0 dwt. 5 grs.

The above experiments show the perfect practicability of cementation from troublesome solutions by keeping the rate of flow at 1/3 to ½ ton per cubic foot per diem, with the presence of from 0.01 to 0.15 per cent, of caustic soda.

A further test was made on the precipitation of auriferous solution carrying quantities of copper, with the following results:

(l) Used solution from the sumps, containing 0.12 per cent, of copper in solution ; was run through extractors at the rate of 1/3 ton per cub. ft. zinc-space per day. Half a pound of caustic soda was added per ton of solution.

Gold in solution before treatment, 2 dwts. 0 gr.
Gold in solution after treatment, 0 dwt. 0.8 gr.
Copper in solution before treatment, 0.12 per cent.
Copper in solution after treatment, 0.097 per cent.

The zinc was coated with metallic copper, but the precipitation at above rate of flow was practically perfect.

Two more prolonged trials were then undertaken, in which a large quantity of solution was run over the zinc used in (l), without any addition of fresh zinc, for over a week, to note the effect of the metallic copper-coating on the zinc. It was found that the coating became disintegrated in the upper compartment, and that the cementation continued unimpaired, though, as the quantity of zinc in the extractors was gradually lessened, the effluent solutions rose in value proportionately. On replacing, however, with fresh zinc the zinc consumed, the solutions again became practically gold-free, as in (l).

In connection with the above precipitation-experiments, two other points call for especial remark:

1. The amount of alkali added is kept as low as possible, to avoid any great excess beyond the amount required to neutralize the acidity of the ore and for zinc-box reactions.
2. It is well known that the presence of alkali or alkaline carbonates, always present in used solutions, interferes with the silver nitrate test, making the solution appear to be stronger in available cyanide than is really the case. Chemists in charge of works should, therefore, be careful not to be misled by this test, or lowered extractions may result. The effect of caustic potash or soda on double cyanide of zinc and potassium is not to regenerate all the cyanide of potassium as shown by silver-test, the test, as suggested by C. J. Ellis, being rendered unreliable by the silver throwing out the zinc from the double cyanide, and the zinc so displaced being dissolved by the KHO present, thus prolonging the titration.

K2ZnCy4 + 2AgNO3 + 4KHO = 2KAgCy2 + K2ZnO2 + 2KHO3 + 2H2O.