The power of retaining certain matters dissolved or suspended in water is possessed naturally by all porous and granular materials that are not themselves soluble in water - as sand, gravel, loam, clay-soil and charcoal, etc. Sand is the least powerful in proportion to its bulk. It does not act except in large quantities, as applied in water works in layers of from two to four feet deep. It is, therefore, practically useless for domestic filters. Sand, and particularly loam, take away an appreciable quantity of salts in solution to the extent of from 5 to 15 per cent, of the amount originally in the water. Of all the materials capable of filtering, animal charcoal is incomparably the best. The active portion (the carbon) forms about 10 per cent, of its weight, the remainder being made up principally of calcic and magnesia phosphate. The extraordinary power of animal charcoal in removing actually dissolved matters from water will be seen if a decoction of logwood be submitted to filtration, the highly-colored liquid passing through the filtering medium perfectly colorless. If instead of the decoction of logwood we take dirty water, porter or port wine, smell, taste and color will in like manner nearly or entirely disappear. This wonderful power of animal charcoal in effecting the entire removal of the coloring matter is only an illustration of what it effects in the case of organic matters which may impart no color. Animal charcoal is largely used in sugar-refining, the dark-colored syrup made by dissolving the raw sugar in water passing through the charcoal perfectly clear and bright, and capable on crystallization of yielding a perfectly white crystalline lump-sugar. When a sufficient thickness of the animal charcoal, in bulk or in layers, is used, it is capable of removing upwards of 85 per cent, of the organic, and 25 per cent, of the mineral matter from the water filtered through it. In this property vegetable is inferior to animal charcoal. Gaultier de Claubry states that one part of animal charcoal purifies 136 times its weight of very impure water.

Leo. Liebermann (Zeitsch. f. Analyt. Chem.) has shown that animal charcoal not only retains many salts when their solutions are filtered through it, but that not a few of the salts are actually decomposed, the charcoal retaining generally the base, with a portion only of the acid.

But, apart from this, charcoal has another property of great interest and importance from the special point of view from which we are regarding it. In the case of shallow wells, receiving, as they are almost certain to do, sewage-contaminated water, the water has often proved harmless, simply because the soil through which it has filtered has effected the oxidation of the organic matter and converted it into harmless products. We noted, moreover, that, when danger resulted from drinking the water of shallow wells, it was when the power of the soil in effecting oxidation failed, and the organic matter found its entrance into the well in an unchanged condition. Now the power of charcoal in absorbing, or rather condensing, oxygen into its pores is most remarkable, and it is to this wonderful property that the value of animal charcoal as a water-filter is particularly due. It acts like the earth, but is infinitely more powerful and effective. Thus, even supposing water contains animal organic matter by ineffectual natural filtration, its passage through charcoal will probably effect the oxidation of the last remnants of these bad ingredients, and so prevent the danger that might otherwise accrue. Charcoal, indeed, when properly used, renders a pure water purer, and also in an impure water renders the deleterious portions of organic matter innocuous by oxidizing them. Many materials, it is true, possess some of the qualities necessary for a purifying medium, but they are all open to objections to which animal charcoal is free. Fibrous organic substances, as wool, cotton, hair and sponge, are objectionable as filters, since they decay and dissolve in the water, and so serve to render it worse after filtration than before. The unpleasant taste of filtered water, which is often complained of in filters fitted with sponge, is simply a minor degree of foulness derived from the sponge, which, in decaying, makes itself distinctly perceptible to taste and smell.

Pulverized coke has been used, and is considered a filtrant, but less effective than charcoal. Wood or vegetable charcoal, however, when powdered, acts merely in a mechanical manner as a strainer. Spongy iron, or pulverized hematite, mixed with sawdust and roasted, pulverized magnetic iron ore and clean scales from a blacksmith's anvil, pulverized and mixed with clean, sharp sand, have been much used and successfully experimented with, and will not only make fetid water sweet, but it is also claimed that the iron mixtures destroy bacteria and their germs.

Magnesia and magnesia compounds with charcoal are also used. As a mechanical filtering medium they may be recommended. Their chemical purifying action depends on the quantity of charcoal introduced and its frequent renewal.

There is one filtering material which is little known in this country, which has all the properties of animal charcoal, and is said to give higher results. This is magnetic carbide, and consists of protoxide of iron in chemical combination with carbon. It is considered that the purifying effect is produced by its power of attracting oxygen to its surface without the latter being acted on, the oxygen thus attracted being changed to ozone, by which the organic matter in the water is consumed.

Whatever filtering material may be adopted, it must be properly applied. If it be so used as to be soon choked up, and at the same time to be inaccessible for cleaning or renewal, it will be useless.

Whenever sand is to be used, alone or in connection with charcoal, it must be first prepared for filtering purposes, be purified from organic matter and lime. To destroy the organic matter it invariably contains, sand must be heated to a high degree. To separate the lime, pour diluted muriatic acid in a suitable vessel over the sand, stir, drain off the acid, and wash carefully with fresh water until the water running off does not turn blue litmus-paper red or pale.