a. Alkalies and their carbonates are detected in mineral waters by the tests already mentioned; and by the water, after being boiled, throwing down a precipitate on the addition of chloride of magnesium. The volatile nature of ammonia easily distinguishes it if present; and the best test for determining whether the fixed alkali be potassa or soda is chloride of platinum 3,which forms an immediate precipitate with potassa or any salt containing it, but is not at all affected by soda. The quantity of an alkali is determined by saturating it with sulphuric acid, and noting the quantity of real acid 4 necessary; setting down, for every 100 grains of real acid used, 121.48 of potassa, or 78.32 of soda. The loss of weight produced by the effervescence on dropping in the acid, being added to the above, shows the quantity of an alkaline carbonate.

b. Iodine is detected by starch. Iodides by starch and chlorine.

c. Earthy carbonates. If the water contains sulphuretted hydrogen gas, this must be separated by exposing the water for a considerable time to the air, before the quantities of the earthy carbonates can be estimated. After this, boil the water for fifteen minutes, filter it when cold; and treat what remains on the filter with hydrochloric acid, which will dissolve the carbonates of lime, of magnesia, and of iron. The residuum, which may contain carbonate of alumina, and, perhaps, sulphate of lime, is to be dried in a red heat, and its weight noted; and then boiled in a solution of carbonate of soda. The soda is next to be saturated with hydrochloric acid, and the mixture boiled for half an hour, which precipitates carbonate of lime and alumina. This precipitate being dried, the lime is to be separated by acetic acid, and the alumina that remains dried and weighed; so that, by subtracting its weight from the original weight, the proportion of sulphate of lime is ascertained.

1 Dr. Henry.

2 This test is sufficiently delicate to detect soda when it amounts to 1/2217 th part only of the water.

3 Sulphate of lime likewise produces the same effect.

4 For a rule to determine the quantity of real acid in a diluted acid, see Introduction.

To estimate the contents of the hydrochloric solution, add to it ammonia as long as it throws down a reddish precipitate, which is the iron united with a portion of magnesia. Separate the magnesia by acetic acid, the precipitate being previously dried by exposure to the air, in a heat of 200°, and the solution added to the hydrochloric solution. To determine the weight of the iron, redissolve it in hydrochloric acid, then precipitate it by an alkaline carbonate, and dry, and weigh.

Sulphuric acid is now to be added to the hydrochloric solution and the sulphate of lime, thus obtained, is to be heated to redness, and weighed; setting down for every 100 grains of it 74 of carbonate of lime. From the solution the magnesia is lastly to be separated by carbonate of soda, dried and weighed : then evaporate the remaining solution to dryness, and wash the residue with distilled water, so as to dissolve the chloride of sodium. This residue is carbonate of magnesia, the weight of which, when dried, must be added to the former; which gives the entire weight of the carbonate of magnesia. 3. Mineral Acids exist in mineral water, sometimes uncombined, but generally combined with alkalies and earths, forming sulphates.

a. Sulphuric Acid is readily detected by chloride of barium, when it does not exceed the millionth part of the water. To render this test certain, however, the chloride must be diluted; the alkaline carbonates, if the water contain any, must be previously saturated with hydrochloric acid; and the precipitate must be insoluble in hydrochloric acid. The hydro-sulphurets are precipitated by chloride of barium, but their presence is easily detected by their odour.

The proportion of sulphuric acid is easily estimated by saturating it with barytic water, and heating the precipitate to ignition: every 100 grains of this sulphate of baryta indicate 34 of real sulphuric acid.

b. The Sulphates contained in mineral waters are six in number, and are incompatible with the following salts placed in the opposite column : -

b. 1. Sulphate of Soda. To detect this salt, first evaporate the water to one half, and add lime-water as long as any precipitate falls. This precipitates all the earths except sulphate of lime, which may be separated by evaporating the fluid till it becomes concentrated, then adding a little alcohol, and after filtration a little oxalic acid. If lime-water produces a precipitate in the water thus treated, immediately, or after a little alcohol be added, either sulphate of potassa or of soda is present. To determine which, add acetate of baryta, which will precipitate sulphate of baryta; then filter and evaporate the filtered fluid to dryness, and dissolve the residue by digesting it in alcohol, and evaporate to dryness. If the sulphate be sulphate of potassa, the dry salt thus obtained, being acetate of potassa, will deliquesce; but if it be sulphate of soda, the acetate will effloresce.

The proportion of the alkaline sulphates is found by precipitating their acid, by nitrate of baryta, from the purified water. If soda be the base of the salt contained in the water, for every 100 grains of this precipitate ignited, set down 61.2 grains of dried sulphate of soda; if potassa be the base, for 100 grains of ignited precipitate set down 74.8 of dry sulphate of potassa.

b. 2. Sulphate of lime is detected by an immediate precipitate being formed by oxalate of potassa. To determine its quantity, first saturate any earthy carbonates that may be present with nitric acid; then evaporate the fluid to a few ounces; and having precipitated the sulphate of lime by means of proof spirit, dry and weigh it.

b. 3. Alum is detected by carbonate of magnesia, chloride of calcium, chloride of magnesium, or succinate of ammonia. Twelve grains of alumina precipitated by carbonate of magnesia, heated to incandescence, indicate 100 grains of crystallised alum, or 49 of the dried salt.

b. 4. Sulphate of magnesia may be detected in any water (previously freed from any alum or uncombined acids) by hydro-sulphuret of strontian, which produces an immediate precipitate with this salt, and with no other. If no other earthy sulphate be present, the sulphuric acid may be separated by a barytic salt; every 100 grains of the ignited precipitate indicating 51 grains of dried sulphate of magnesia. If sulphate of iron be present, mix the water with a portion of argil, and expose it for some days to the air, during which time oxide of iron and sulphate of alumina are precipitated, leaving the sulphate of magnesia alone in solution; which may be then estimated by the above method.