Pure water is simply oxygen and hydrogen, eight-ninths being oxygen and but one-ninth hydrogen; the latter gas, if pure, having, like oxygen, neither taste nor smell. Rain-water is the purest type; and, if collected in open vessels as it falls, is necessarily free from any possible taint (except at the very first of a rain, when it washes down considerable floating impurity from the atmosphere, especially in cities). This mode being for obvious reasons impracticable, cisterns are made, and rain conducted to them through pipes leading from the roof. The water has thus taken up all the dust, soot, and other impurities found upon the roof, and, unless filtered, can not be considered desirable drink. The best cistern will include a filter of some sort, and this is accomplished in two ways. Either the cistern is divided into two parts, the water being received on one side, and allowed to slowly filter through a wall of porous brick, regarded by many as an amply sufficient means of purification; or a more elaborate form is used, the division in such case being into upper and under compartments, the upper one containing the usual filter of iron, charcoal, sponge, and gravel or sand. If this water has a free current of air passing over it, it will acquire more sparkle and character; but as a rule it is flat and unpleasant in flavor, being entirely destitute of the earthy salts and the carbonic-acid gas to be found in the best river or spring water.
Distilled water comes next in purity, and is, in fact, identical in character with rain-water; the latter being merely steam, condensed into rain in the great alembic of the sky. But both have the curious property of taking up and dissolving _lead_ wherever they find it; and it is for this reason that lead pipes as leaders from or to cisterns should _never_ be allowed, unless lined with some other metal.
The most refreshing as well as most wholesome water is river or spring water, perfectly filtered so that no possible impurity can remain. It is then soft and clear; has sufficient air and carbonic acid to make it refreshing, and enough earthy salts to prevent its taking up lead, and so becoming poisonous. River-water for daily use of course requires a system of pipes, and in small places is practically unavailable; so that wells are likely, in such case, to be the chief source of supply. Such water will of course be spring-water, with the characteristics of the soil through which it rises. If the well be shallow, and fed by surface springs, all impurities of the soil will be found in it; and thus to _dig deep_ becomes essential, for many reasons. Dr. Parker of England, in some papers on practical hygiene, gives a clear and easily understood statement of some causes affecting the purity of well-water.