(a) All the liquids which contain gold, except those in which there is a cyauide, are strongly acidulated by sulphuric or hydrochloric acid, unless they are already acid, and then largely diluted with water. Precipitate the precious metal by a solution of sulphate of protoxide of iron (copperas), and, after a few hours standing, it is ascertained that the liquor does not contain any more gold when a new addition of iron sulphate does not produce any turbidity. The precipitated gold is in the form of a red or black powder; collect upon a filter, wash, and dry in an iron pan, with weights equal to its own, of borax, saltpetre, and potash carbonate. Gradually introduce the powder into a refractory crucible heated to a white heat in a good air-furnace. When all is introduced, increase the heat and close the furnace, so that all the metal may fall to the bottom of the crucible. After cooling, extract the button of pure gold which remains. If it is desired to dissolve the powdered gold left on the filter in aqua regia, it will be necessary to wash it several times with dilute sulphuric acid, to remove the iron sulphate with which it is impregnated.

This mode of reduction is adapted to an impure gold chloride, to the baths by dipping with soda bicarbonate or pyrophosphate, and also to the ungild-ing acids; but it is imperfect with baths holding a cyanide, which never completely part with all the gold they contain by this process. The best manner of treating the latter liquors is to evaporate them to dryness in a cast-iron kettle, and calcine the residue to a white heat in a good crucible. A small proportion of borax or saltpetre may be added to aid the fusion, but generally it is unnecessary. The resulting button of gold at the bottom of the crucible is red when saltpetre has been employed, and green with borax; but these differences of colour have nothing to do with the purity of the metal.

(6) Gold may be separated from liquors which contain no cyanide, by an excess of tin protochloride, which produces a precipitate easily reduced by heat. Sulphurous acid will also reduce the gold; but in this case, the liquor should be heated. Granulated gold is obtained by running the molten metal, in a small stream, and from a certain height, into a large quantity of cold water.

(c) Sweepings, saw-dust, residues from the bottoms of scratch-brushing tubs, filters, papers, and rags, must be collected, mixed, and burned in a furnace constructed for the purpose. The ashes are finely pulverised, sifted, and thoroughly mixed with a quantity of mercury, which combines with the gold and silver. The amalgams, separated by washing, are then distilled in cast-iron retorts of a peculiar shape. The mercury volatilises, and the gold and silver remain in the retort. For separating these metals, granulate the solid mass and treat with pure nitric acid, which dissolves the silver, and is without action upon the gold. The latter metal collects at the bottom of the vessel in a black or violet powder, and is pure, after having been washed in distilled water. If an ingot contains only a little silver and much gold, melt previously with a certain proportion of the former metal, in order to more easily dissolve in nitric acid. The ingots of silver and copper should be boiled in cast-iron kettles with concentrated sulphuric acid, which transforms the copper into soluble sulphate, and silver into sulphate, only slightly soluble. The separation of the two may be partly effected by washing, but, generally, the silver is precipitated by plates of copper.

The alloy, previous to its solution, should be granulated.

(d) An easy method to recover gold from solutions, particularly from old toning-baths of photographers, has been made known by Haugk. It consists in filtering the solution into a white glass flask or bottle, making it alkaline with sodium bicarbonate, and then adding, drop by drop, a concentrated alcoholic solution of aniline red (fuchsine), until the liquid retains a deep strawberry colour. The flask is then exposed to the sunlight for 6-8 hours, at the end of which all the gold still present will have been precipitated as a dark violet powder, and the liquid will have become colourless. After pouring off the liquid, the flask, with its precipitate, is kept until a fresh quantity of solution has to be precipitated, and this is continued until the deposit in the flask is sufficiently large to make it worth while to remove it. It is then transferred to a filter, washed, dried, and burned with the filter. The residue containing the filter-ash is dissolved at a gentle heat in aqua regia, filtered, and the solution is evaporated to dryness.

The quantity of impurity caused by the simultaneous solution of the filter-ash is too insignificant to be objected to.

(e) Although sheet zinc, or zinc and iron sheets, serve well for the precipitation of silver, they cannot be employed for the recovery of gold. The latter separates out in such a case very incompletely and as a firmly adhering lustrous film in the zinc. On the other hand, finely divided zinc, the so-called zinc dust, is an excellent substance to employ for precipitating gold quantitatively and in the form of powder from spent cyanide liquors. When zinc dust is added to a spent gold bath and the liquid is periodically stirred or shaken, all the gold is precipitated in 2-3 days. The amount of zinc to be added naturally depends on the quantity of gold present. Freshly prepared gold baths for gilding in the cold contain on the average 3.5 grm. gold per litre; whilst those used for the hot process contain 0 * 75 grm To precipitate all the gold in the original bath, 1.74 grm. or 0.37--0.5 grm. zinc dust would be necessary, and, of course, a much smaller quantity would be sufficient for the spent liquors.

Since the precipitation takes place more rapidly when an excess of zinc dust is present, it is generally advisable to add J or at the most 1/2 kilo, of zinc dust to every 100 litres of solution.

The precipitated gold, which contains zinc dust and usually silver and copper, is washed, freed from zinc by hydrochloric acid, and then from silver and copper by nitric acid and thus obtained pure.

A spent bath treated in this way gave the following amounts of gold per litre: -

1st experiment .. 0.2626 grm. 2nd „ .. 0.2634 „ Mean...... 0.2630 .

The presence of gold in the residual cyanide solution could not be qualitatively detected. The potassium cyanide of the solutions obtained by this process should be converted into ferrocyanide by heating with ferrous sulphate and milk of lime, since this substance is not poisonous and can therefore be got rid of without danger. It would, however, be more economical and, considering the large amount of cyanide present, more profitable to work it up into Prussian blue.