of the vessel used for drying the sample, the fact of the discrepancy has been well attested and it was probably in consequence of this investigation that the Rand practice was developed of "fixing" or precipitating the gold in the slime residues by the copper sulphate method used for solution assays before drying the pulp (see page 52)..

A. Whitby1 states that his practice is to add to the slime residue first a little cyanide solution and then small quantities of solutions of copper sulphate, sodium sulphite and sulphuric acid. After mixing thoroughly the pulp is turned out into an enamelled dish and dried, care being taken to avoid spitting. The sample is then passed through a 100-mesh screen and assayed. It is the custom on the Rand to take exceptionally large charges for the assay and for low-grade pulps it is a necessary precaution especially where a large mill tonnage will multiply an error of a few cents in the assay to a serious sum total of gold in the month's tonnage. From five to ten assay tons of pulp are usually taken for an assay of residue, distributed over two or more crucibles. The flux used by Mr. Whitby for the Rand residues has the following proportions:

Reducer, sufficient to give a button of about 45 grams.

That the error due to neglect of these precautions may be serious is shown by an experience of the writer with a high-grade gold-silver ore in Mexico. It had been the custom to estimate the assay of ore passing to the cyanide plant by assaying the pulp as thickened in the cyanide solution, prior to agitation, the argument being that as no part of the mill solution was withdrawn for precipitation the amount of gold and silver circulating in that medium might be taken as fairly constant and that therefore an assay of the thickened pulp with its normal burden of dissolved 1 Journal Chem., Metall. and Min. Soc. of S. A., March, 1906.

gold and silver would truly represent the value of the pulp leaving the mill, and this value plus that contained in the concentrate would represent the assay of the mill head ore. In practice, however, there was a persistent plus recovery of gold over that shown by assays.

On precipitating the pulp before drying for assay it was found that the gold value as shown by the usual method was much too low though the silver did not appear to be affected. A comparison run daily for 26 days with and without preliminary “fixation" of the gold showed an average gain in the assay of $1.40 per ton due to "fixation," the figures being $13.90 by the usual method and $15.31 by the preliminary "fixation." Of course this is an extreme instance, but in the case of residues a persistent error in the assay of only 10 c. a ton on the low side would make a serious difference between actual and theoretical recovery.

CHAPTER IX

ORES PRESENTING SPECIAL DIFFICULTIES

Telluride and Mispickel Gold Ores.-Gold associated in this way is difficultly soluble in ordinary cyanide solutions and special treatment is necessary for its extraction. There are two methods in use in such cases.

1. Roasting. The ore is ground dry to about 30 mesh and roasted; it is then ground to a slime in water or cyanide solution, usually in grinding pans, utilizing the opportunity simultaneously to amalgamate any gold that may be amenable, and finally treated by agitation in cyanide solution in the ordinary way.

2. Treating Raw with Bromocyanide. This process was used at the Deloro mine, Ontario, Canada, on a mispickel ore, and was developed in Australia by Dr. Diehl for the treatment of the sulpho-telluride ores of the Kalgurli district. The process essentially consists of grinding exceedingly fine, and then agitating with cyanide solution to which bromide of cyanogen is added at intervals. In some mills preliminary amalgamation and concentration are in use, the concentrate being roasted and then cyanided. During bromocyanide treatment the protective alkalinity is kept at the lowest possible point owing to the instability of the reagent, lime sufficient for settlement being added after the treatment is finished.

Julian and Smart1 state that at Kalgoorli when dealing with slimes assaying from 1 to 3 oz. of gold a wash of cyanide solution containing about 4.5 lb. of KCN to the ton of ore is first given, and after an hour or two's agitation cyanogen bromide is added at the rate of about 1 lb. BrCN per ton of ore, and agitation continued for twenty-four hours. The quantities of free cyanide and bromocyanide are varied according to the assay value of the 1 Cyaniding Gold and Silver Ores, page 260 (Second Edition).

ore, any additional bromocyanide needed being added at intervals of several hours. Shortly before agitation is finished sufficient lime is added for the settlement of the slime.

The usual method of making the reagent for laboratory use is to add a strong solution of cyanide to bromine (and not vice versa), until the brown color is just discharged.

The quantity of BrCN may be determined in a working cyanide solution by acidifying with hydrochloric acid, adding excess of potassium iodide, and titrating the liberated iodine with decinormal sodium thiosulphate. Clennell gives the following reactions:1

The bromocyanogen is added to the cyanide solution to be used for extraction purposes in the proportion of about 1 of BrCN by weight to 4 of KCN.

Since BrCN is rapidly decomposed by alkali it is important that free alkali should be almost entirely absent during the treatment, the lime necessary for settlement being added at its conclusion. The reaction is usually illustrated thus:

BrCN+2KOH= KBr+ KCNO + H2O

The following is an extract from a paper by E. W. Nardin, of the Hannan's Star Mill, Kalgoorli, published in the Mining and Scientific Press of October 24th, 1908.

"The bromo-cyanide solution is made according to the following equation:

(1) 2KBr+ KBrO3 + 3KCN + 3H2SO4 = 3BrCN +3K2SO4 + 3H2O. Its action in the treatment vat is supposed to be as follows:

(2)

BrCN+3KCN + 2Au

=

2KAu (CN)2 + KBr.

1 Chemistry of Cyanide Solutions, page 100 (Second Edition).

"The first two quantities in equation 1 are contained in the mixed salts supplied by the London-Hamburg Co., having about 40 to 44% Br as KBr, and 20 to 22% Br as KBrO3; the proportion of Br as bromide being about twice that of Br as bromate. A 30-lb. charge is usually made up, and for this the following weights are taken:

The KCN is 93%, and the H2SO4 63% (chamber acid) strength. The solution is made in a closed wooden vessel, stirred by rotating arms, holding about 200 gal. In making up a charge, a portion of the water and all the H2SO4 are first mixed, and allowed to cool to normal temperature. The KCN, which is dissolved in a separate vessel in sufficient water to fill the mixing vessel, is then run in, and at the same time the proper weight of 'mixed salts' is gradually added. The whole is then agitated for 6 hours before being used, and in a closed vessel it will retain its strength for some days. The cost of a 30-lb. charge of BrCN is about $21.88, made up as follows: 50 lb. H2SO4 at 4 c., 20 lb. KCN at 17 c., 36.8 lb. salts at 44 c. From each charge mixed, a dip-sample is taken and tested with a standard Na2S2O3 solution, using potassium iodide as an indicator.

Thus BrCN corresponds to 2Na2S2O3.

"The standard solution is made so that 1 cc corresponds to 0.02 gm. BrCN, and for this about 93.6 gm. of ordinary photographic crystals, Na2S2O3 5H2O, are dissolved in one litre of water."

(5)

(248)

"A solution of copper sulphate is used for standardizing the above. 2 CuSO4 + 4KI = Cu2l2 + 2K2SO4 + I2

126.8 Cu liberates 254 I, so that it corresponds to 106 BrCN and 496 Na2S2O3 5H2O. The solution is made by dissolving one gram of pure copper foil in acid, converting to sulphate, and dissolving in 100