As the action of spongy iron is dependent upon its remaining covered with water, whilst the materials which are employed in perhaps all other filters lose their purifying action very soon, unless they are run dry from time to time, so as to expose them to the air, the former is peculiarly suited for cistern filters.

Cistern filters are frequently constructed with a top screwed on to the filter case, by means of a flange and bolts, a U-shaped pipe passing down from this top to near the bottom of the cistern. This tube sometimes supplies the unfiltered water, or in some filters carries off the filtered water, when upward filtration is employed. This plan is defective, because it practically gives no access to the materials; and unless the top is jointed perfectly tight, the unfiltered water, with upward filtration, may be sucked in through the joint, without passing at all through the materials. This is remedied by loosely surrounding the filter case with a cylindrical mantle of zinc, which is closed at its top and open at the bottom. Supposing the filter case to be covered with water, and the mantle placed over the case, an air valve is then opened in the top of the mantle, when the air escapes, being replaced by water. After screwing the valve on again, the filter is supplied with water by the siphon action taking place between the mantle and filter case and the column of filtered water, which passes down from the bottom of the filter to the lower parts of the building. These filters are supplied with a regulator arrangement on the same principle as ordinary domestic filters. The washing of materials, on starting a filter, is easily accomplished by reversing 2 stop-cocks, one leading to the regulator, the other to a waste pipe.

The use of spongy iron has now been applied on a large scale to the water obtained from the river Nette, for the supply of the city of Antwerp. Dr. Frankland has visited the Antwerp Waterworks at Waelheim, about 15 miles above that city, and reported on the result of his inquiry. He attaches especial value to the fact that spongy iron filtration “is absolutely fatal to Bacteria and their germs,” and he considers it would be “an invaluable boon to the Metropolis if all water supplied from the Thames and Lea were submitted to this treatment in default of a new supply from unimpeachable sources.”

Many preparations of iron have long been known to possess a purifying influence on water containing organic impurities. Thus Scherer, years ago, recommended a solution of iron sulphate where the impurities were present in large quantity. Later still, iron chloride was proposed as suitable, the salt being precipitated in the presence of organic matter as ferric oxide, the oxide thus formed acting also mechanically on the suspended impurities in course of precipitation, very much as white of egg acts in clarifying liquids, when it coagulates and carries impurities with it to the bottom. Other iron preparations have a similar action, notably dialysed iron, while several oxidising agents, such as potash permanganate, are also well known to possess a powerful effect on organic impurities. It will at once be seen, however, that all such substances are inadmissible as filtering media, or purifying agents for potable waters, for the reason, that in the case of some at least of the agents mentioned, decompositions take place, which in themselves might prove dangerous, while in the case of all an excess (and it would be almost impossible to avoid an excess) of the purifying agent would be equally bad, and would render the water quite unfit for domestic purposes. It has been found, however, that various kinds of native rock containing iron protoxide effect the filtration of water very completely, and Spencer, acting on this idea, after experimenting, found that when the iron protoxide was isolated as magnetic oxide, it both freed the water from turbidity and effected decoloration very quickly. Thus bog-water, as dark as porter, when filtered through it speedily lost its colour and became clear and sweet, the carbonic acid given off during the process of decomposition rather tending to improve the water. The purifying power of the magnetic oxide does not deteriorate with use. The oxide gets coated with a slimy deposit, owing to the deposition of decomposed organic matter, but this being removed, it is as powerful as ever in its purifying action. Unfortunately this iron rock is not found native to any extent, but the fact of its action being determined, Spencer continued his experiments with the result that it can now be produced artificially, and forms one of the most efficient and useful filters for domestic purposes.

Metallic iron is employed by Jennings & Hinde. The filtering material consists of fine iron or steel shavings, filings, turnings, or borings obtained from the swarf or skin of cast iron, wrought iron, or steel; this material may either be used by itself, or it may be used with other materials, either mixed with them or in separate layers. The iron or steel shavings, &c., are obtained from iron or steel that has been brought to a state of fusion either by melting or the processes necessary for making cast iron, wrought iron, or steel, and being separated from many of the impurities contained in the ore from which it was obtained, will have but a comparatively small portion of earthy impurities mixed with it, and will be for this reason superior to iron which is obtained from native ores or oxides without fusion.

By filtering water through small divided swarf or skin of cast iron, wrought iron, or steel, free oxygen will be withdrawn from the water, and consequently any insects or animalculæ contained in the water will be deprived of life, and any germs contained in the water will be deprived of the oxygen necessary for their development and life, and the water will be consequently purified and rendered wholesome. A convenient way of forming a filter is to use a layer of the turnings, shavings, &c., together with layers of other filtering material resting upon a perforated partition placed across a closed vessel. The materials are cleaned by boiling them in hot water with a small quantity of ordinary washing soda, to remove any oil or grease that might accidentally be associated with the materials above mentioned. Afterwards the iron borings should be well washed before being put into the filter. The filter vessel may be of any ordinary construction and shape. If sand is used in conjunction with the above-mentioned materials, it is preferable to place some of the sand at the bottom of the filtering vessel, and the iron or steel materials, or both, over the sand, and then more sand over them. These materials are disposed so that they may be partially separated from each other by perforated plates of earthenware, glass, or other suitable material. But this partial separation, though convenient, is not essential, as the perforated plates may be dispensed with and the material placed over and under each other in layers without plates to separate them.

Porous Pottery.—Chamberland has found that the liquid in which microbes have been cultivated becomes absolutely pure if passed through unglazed porcelain. Its purity can be demonstrated by mixing it with liquids sensitive to the action of microbes, such as veal broth, milk, and blood, in which it produces no alteration.

16. Chamberland Filter.

A tube a (Fig. 16) of unglazed porcelain is enclosed in another b of metal, and the water to be filtered is admitted to the space between the two by turning a stop-cock. Thence it slowly filters through to the inside of the porcelain tube, and flows out at the bottom. Under a pressure of 2 atmospheres, or 30 lb. to the sq. in., a tube 8 in. in length, with a diameter of 1 in., will yield about 5 gal. of water daily. For a larger supply, it is only necessary to increase the size or the number of the tubes.