through a 30-mesh sieve. It is crushed with a Wedgewood mortar and pestle, and the whole passed through a 60-mesh sieve, this operation being performed as quickly as possible. It is then placed in a clean, dry, wide-mouthed bottle fitted with a tight-fitting dry glass-stopper so as to prevent any access of air. Two grams of the sample is carefully weighed out and agitated with 1 litre of a 2 per cent. cane-sugar solution (or 1 gm. with 1⁄2 litre of a 2 per cent. sugar solution). If a shaking machine be available, 2 hours continuous agitation should be given, if not, 6 hours intermittent agitation, every care being taken to prevent coagulation of the lime, in order that the lime and solution may be brought into the most intimate contact during this period. When the agitation is finished the solution is filtered as quickly as possible, and aliquot portions titrated with N/10 or N/5 acid, using rosolic acid as indicator, avoiding delay so as to obviate undue exposure to the atmosphere. The distilled water used in the above determination must be made neutral to rosolic acid to counteract the presence of dissolved CO2.

Zinc Determination.-To ascertain the presence of zinc add a few drops of sodium sulphide solution to 50 cc of the solution to be tested (which should be strongly alkaline), and heat to boiling point. If zinc is present it will come down as a flocculent white precipitate. It sometimes happens that a cyanide solution will contain an appreciable quantity of alumina, in which case a precipitate of aluminium hydroxide may be formed on adding sodium sulphide. This precipitate, however, is gelatinous and can be easily distinguished from the flocculent precipitate of zinc sulphide.

If silver is present also it should first be removed by shaking up some of the solution with a little aluminium dust and filtering, as the black precipitate of silver will mask the white sulphide of zinc.

To determine the amount of zinc, the method recommended by Gerard Williams1 is probably the most generally useful. Silver 1Journal Chem., Metal. and Min. Soc. of S. A., May, 1904.

and lead, if present, interfere and should be removed by precipitation with aluminium before proceeding with the test. Lime, also, has in some cases appeared to impair the accuracy of results, and is better removed.

The reagents needed are aluminium dust, commercially known as aluminium bronze powder, crystals of neutral potassium or sodium oxalate, sodium sulphide solution, a decinormal solution of iodine and a decinormal solution of sodium thiosulphate (Na2S2O3.5H2O).

The thiosulphate solution is made by dissolving 24.827 grams of the salt in distilled water and making up to 1 litre. If the crystals are chemically pure and have been kept well stoppered so that they are neither damp nor coated with white powder (indicating dehydration), this solution may be taken as standard for the purposes of this test, though for accurate chemical research work it would be necessary to standardize it against a strictly decinormal iodine solution, or by other methods given in works on volumetric analysis.

A decinormal solution of iodine contains 12.7 grams of the element per litre. To expedite the dissolution of the iodine Clennell1 recommends the following procedure: weigh out 18 or 20 grams of potassium iodide and dissolve about 15 grams in 20 or 25 cc of distilled water, reserving the remainder of the iodide crystals until the finish. Then weigh up 12.7 grams of c.p. iodine and transfer quickly to the beaker containing the iodide solution. Stir with a glass rod until no more will dissolve, dilute with distilled water, and decant carefully into a litre flask. Add the remainder of the iodide to the beaker with a little water, and the rest of the iodine should soon dissolve; if not, a gram or two more of the iodide may be added. When all the iodine solution has been transferred to the litre flask, fill up to the mark with distilled water, and transfer the whole to a stoppered bottle, which should be kept in a dark cupboard.

To check this solution against the decinormal thiosulphate take 25 cc in a small beaker and run in the thiosulphate solution 1 Private communication.

from a burette until the brown color is almost dispelled. Then add a few drops of a solution of boiled starch as indicator, and continue the titration until the blue color disappears. Should it need more or less than 25 cc of thiosulphate to produce this result, the difference should be noted on the label of the iodine bottle and the proper correction made for it when using the iodine solution for subsequent titrations If it appears from the look of the thiosulphate crystals that they are of doubtful hydration, it will be better to take the iodine solution as the standard and make the necessary correction on the thiosulphate solution. For instance, if the 25 cc of thiosulphate solution was neutralized by 24 cc of iodine solution, 1 cc thiosulphate = 0.96 cc iodine. Then if the thiosulphate be considered standard and if in an analytical titration 13 cc of iodine is needed, the correction for the iodine would be 0.96:1:13:13.54, indicating that 13.54

N 10

cc of strictly iodine had been used. If on the other hand the

iodine be considered standard, then

and

24 cc of iodine = 25 cc thiosulphate

=

1 cc of iodine 1.04 cc thiosulphate

so if in an analytical titration 13 cc of thiosulphate is needed, the correction for the thiosulphate would be 1.04:1::13:12.5,

indicating that 12.5 cc of strictly

N
10

thiosulphate had been used.

To determine zinc take about 150 cc of the cyanide solution to be tested and dissolve in it a few crystals of potassium oxalate. This will precipitate the lime as an oxalate and substitute caustic potash in the solution.

Should silver and lead be absent the zinc determination may be proceeded with after filtering out the lime, but if either of them is present, it should be first removed. To effect this, take the filtrate after lime precipitation and place in a small stoppered bottle, add 40 or 50 mgm. of aluminium powder and shake briskly for 4 or 5 minutes. A drop of the liquor may then be withdrawn on a glass rod and brought in contact with a piece of filter paper moistened with sodium sulphide solution. Should any stain

result the precipitation is not complete and agitation must be continued, with addition of more aluminium powder if necessary.

Since caustic soda is needed in order to precipitate metals in a cyanide solution by means of aluminium the solution must contain from 0.05% to 0.1% of free alkali; if it carries much less than this, a tiny splinter of caustic soda may be added.

When precipitation of silver and lead is complete the solution is filtered, and the filtrate is then ready for the zinc determination. Measure out 100 cc of this filtrate, make strongly alkaline with caustic soda, add excess of sodium sulphide solution, and heat almost to boiling point. By this time all the zinc should have come down as zinc sulphide, which is filtered out, and well washed with hot water, the washing of precipitate and filter paper being continued until the water coming away shows no stain when tested with lead acetate paper or solution, indicating that all soluble sulphide has been eliminated. The filter paper with precipitate is then transferred to a small stoppered flask and agitated violently with a little distilled water to disintegrate the paper. An amount of decinormal iodine solution is then run in from a burette, agitating the flask meanwhile, until an excess is present, as indicated by a permanent dark brown color. This excess is then titrated back with decinormal thiosulphate.

The original number of cc of iodine solution run in less the excess indicated by thiosulphate represents the quantity used up in decomposing the zinc sulphide.

The filter paper pulp usually acts as an indicator in this case, rendering the use of starch unnecessary, but the latter indicator may be used in the ordinary way if desired.

Copper Determination. The first essential is completely to decompose all the cyanogen compounds. For this purpose Clennell1 recommends evaporating 100 cc of the solution to be tested with 5 cc of nitric and 10 cc of sulphuric acid until dense white

1 Chemistry of Cyanide Solutions, page 134 (Second Edition).

fumes are freely given off. Sharwood' prefers to evaporate 100 cc of the solution to dryness twice with nitric acid and then dissolve the residue in dilute sulphuric. If in either case a yellowish flocculent precipitate remains undissolved an incomplete decomposition of cyanogen compounds should be suspected. In some instances the writer has found it necessary, after evaporating to dryness twice with nitric acid, to add strong sulphuric and evaporate to white fumes in order to complete the decomposition of the cyanides.

When a large quantity of solution is taken for evaporation the lime separates out as calcium sulphate at the finish so it is better in such cases before starting the analysis to precipitate the lime with sodium carbonate or potassium oxalate and filter it out. When decomposition is complete, whatever method be used, the assay is diluted with water and boiled until all anhydrous salts are dissolved. Sometimes a slight insoluble residue is left which may be filtered out, washing the filter paper with hot water and adding the washings to the filtrate. Next place in the solution a cylinder of thin sheet aluminium and boil until all the copper is precipitated. If hydrogen is not freely given off add more sulphuric acid. When the action is sufficiently brisk the copper deposited on the aluminium will be detached and form spongy masses, leaving the aluminium surface clean. When the copper is precipitated remove the aluminium cylinder and wash it with distilled water. Should any copper adhere to the surface it may be detached with a rubber-tipped glass rod and washed into the beaker. Decant the solution carefully and wash the spongy copper several times by decantation with hot water, avoiding exposure to the air so as to minimize the risk of oxidation.

The copper is now dissolved in 5 to 10 cc of nitric acid, and boiled until all brown fumes are expelled, and the determination made by any of the standard methods, such as electrolysis, titration with cyanide, or the potassium iodide method.

The Cyanide Method for Copper.-Unless a high degree of accuracy is required the cyanide method may be made to give quite satisfactory results.

1 Engineering and Mining Journal, 1898, page 216,