If a few drops of a solution of nitrate of silver occasions a milkiness with the water, which vanishes again by the copious addition of liquid ammonia, we have reason to believe that the water contains a salt, one of the constituent parts of which is muriatic acid. EXPERIMENT. If lime water or barytic water occasions a precipitate which again vanishes by the admixture of muriatic acid, then carbonic acid is present in the water. EXPERIMENT. If a solution of phosphate of soda produces a milkiness with the water, after a previous addition to it of a similar quantity of neutral carbonate of ammonia, we may then expect magnesia. The application of this test is best made in the following manner: Concentrate a quantity of the water to be examined to about 1/20 part of its bulk, and drop into about half a wine-glassful, about five grains of neutral carbonate of ammonia. No magnesia becomes yet precipitated if this earth be present; but on adding a like quantity of phosphate of soda, the magnesia falls down, as an insoluble salt. It is essential that the carbonate of ammonia be neutral. This test was first pointed out by Dr. Wollaston. The presence of oxygen gas loosely combined in water may readily be discovered in the following manner. EXPERIMENT. Fill a vial with water, and add to it a small quantity of green sulphate of iron. If the water be entirely free of oxygen, and if the vessel be well stopped and completely filled, the solution is transparent; but if otherwise, it soon becomes slightly turbid, from the oxide of iron attracting the oxygen, and a small portion of it, in this more highly oxidated state, leaving the acid and being precipitated. Or, according to a method pointed out by Driessen, the water is to be boiled for two hours in a flask filled with it, and immersed in a vessel of water kept boiling, with the mouth of the flask under the surface of the water: it is to be inverted in quicksilver, taking care that no air-bubble adheres to the side of the flask, and being tinged with infusion of litmus, a little nitrous gas is to be introduced: if the oxygen gas has been sufficiently expelled from the water, the purple colour of the litmus does not change; while, if oxygen be present, it immediately becomes red.[16] If we examine the different waters which are used for the ordinary purposes of life, and judge of them by the above tests, we shall find them to differ considerably from each other. Some contain a large quantity of saline and earthy matters, whilst others are nearly pure. The differences are produced by the great solvent power which water exercises upon most substances. Wells should never be lined with bricks, which render soft water hard; or, if bricks be employed, they should be bedded in and covered with cement. METHOD OF ASCERTAINING THE RELATIVE QUANTITY OF EACH OF THE DIFFERENT SUBSTANCES USUALLY CONTAINED IN COMMON WATER.