b. 5. Sulphate of iron is detected by tincture of galls striking a black colour with the water after it has been boiled, and has cooled. Its quantity may be estimated by precipitating the iron by ferro-cyanide of potassium.1 c. Hydrochloric Acid, either uncombined or combined, is detected by nitrate of silver, which forms with it a white precipitate insoluble in nitric acid; but the alkaline carbonates, if any, must be first saturated by nitric acid, and any sulphuric acid removed by nitrate of baryta. The proportion of uncombined acid is ascertained by saturating it with barytic water, and then precipitating the baryta by sulphuric acid. For every 100 grains of the ignited precipitate, set down 21 grains of real hydrochloric acid. Very minute portions of chlorides may be detected by putting into some pure nitric acid. contained in a porcelain capsule a minute quantity of finely divided gold, precipitated from its solution by sulphate of iron, and then adding the supposed chloride. If it be a chloride, a light tint will be gradually formed round the gold.2 d.

The Chlorides and Hydrochlorates contained in mineral waters are incompatible with the following articles in the second column : -

1 To make the calculation, the weight of a precipitate produced by the prussiate in a solution of a given weight of sulphate of iron in water must be previously determined.

2 Ann. de Chim. xxviii. 26.

d. 1. Hydrochlorates of soda and of potassa are detected in water by acetate of silver: but any earthy nitrates and chlorides must first be decomposed by sulphuric acid, and the sulphates separated by alcohol and nitrate of baryta. To ascertain whether the precipitate be hydrochlorate of soda or of potassa, evaporate to dryness, then dissolve the acetate in alcohol, and again evaporate to dryness. If it be acetate of potassa the salt will deliquesce; but if acetate of soda, it will effloresce. To estimate the quantity of these salts, if they be unaccompanied by other salts, it is only necessary to dry and weigh the precipitate by nitrate of silver, setting down for every 100 grains of chloride of silver thus thrown down, 52 of chloride of potassium, and 41 of chloride of scdium. If alkaline carbonates be present, they must be first saturated with sulphuric acid, and sulphate of silver used to precipitate the hydrochloric acid.

d. 2. Chloride of barium is detected by sulphuric acid. It is rarely found.

d. 3. Chloride of calcium. To detect this salt the water must be first freed from the sulphates, then filtered, evaporated to dryness, the dry mass treated with alcohol, and the residue, after evaporating the alcohol, dissolved in water. If this solution yield a precipitate with acetate of silver, the water contained chloride of calcium.

d. 4. Chloride of magnesium is detected by separating the sulphates, and proceeding as in the former case. If the aqueous solution of the dry mass treated with alcohol afford no precipitate with carbonate of lime; and if sulphuric acid and evaporation, with the addition of a little alcohol, occasion no precipitate, the solution contains only magnesian salt.

d. 5. Chloride of aluminum is detected by first saturating any alkali the water may contain with nitric acid, and separating any sulphuric acid by nitrate of baryta; and then adding carbonate of lime, which produces a precipitate if this salt be present. This process also precipitates chloride of iron and of manganese, if any be present.

To estimate the quantities of these salts, which may all be contained in the same water, the earths, after separating any sulphates that may be present, are to be precipitated by baryta water, and redissolved in hydrochloric acid. They are then to be separated by the rules already mentioned, and separately weighed. For every 50 grains of lime, set down 100 of dried chloride of calcium; for 30 grains of magnesia, 100 of chloride of magnesium; and for 21.8 grains of alumina, 100 of chloride of aluminum. The barium of the chloride of barium, which the addition of the baryta water had formed in the mineral water by precipitating the earths, is now to be separated by sulphuric acid, and its hydrochloric acid expelled by heat; after which the chloride of sodium, which the water originally contained, is to be obtained by evaporation.

e. Nitric Acid never exists in an uncombined state in mineral waters, and even the nitrates are comparatively of rare occurrence. The nitrates may be detected by an experiment, the inverse of that employed for detecting the chlorides; that is, by putting the fragment of gold into pure, colourless hydrochloric acid, and adding to it the suspected nitrate.

f. The nitrates are incompatible with the salts in the second column of the following table: -

f. 1. Nitrate of potassa may occur in mineral waters in conjunction with sulphates and chlorides; the former of which must be decomposed by acetate of baryta, and the latter by acetate of silver, before the nitrates can be estimated. After these previous steps, filter the water, evaporate it to dryness, and treat with alcohol; which dissolves the acetates, and leaves the nitre.

f. 2. Nitrate of lime is detected by first concentrating the water, and separating the sulphates by alcohol; then filtering and distilling off the alcohol, and separating any hydrochloric acid by acetate of silver; afterwards filtering again, evaporating to dryness, and dissolving the residue in alcohol, which must be also distilled off, and the dried residue dissolved in water. If oxalic acid detect lime in this solution, the mineral water contains nitrate of lime; the quantity of which may be estimated by precipitating with sulphuric acid, and calculating the quantity of lime contained in the sulphate; and for every 35 grains of lime, setting down 100 grains of dry nitrate of lime. J. 3. Nitrate of magnesia is detected by nearly the same means: but to the last watery solution, instead of oxalic acid, add potassa, as long as any precipitate appears. Filter this solution; evaporate and treat the dry mass with alcohol. If a residue of nitre remains, the mineral water contained nitrate of magnesia. Such is the general method of ascertaining the components of mineral waters, and the proportion of the ingredients contained in any particular water.

To render the analysis complete, many minutiae must necessarily be attended to; but the detail of these would far exceed the limits which a work of this kind can admit of.