an instrument or apparatus contrived for clearing liquids from any impurities mechanically combined with them. Filters are of various kinds, according to the nature of the liquor requiring to be clarified; but they all act on the same general principle, namely, straining the liquor by passing it through earth or other porous material; the process being thus exactly analogous to that of sifting, as used for powder or other dry substances. The operation of filtering is extensively used in chemistry and the arts, and also in purifying water for domestic purposes, and for the use of ships in long voyages; and hence the construction of artificial filters on a large scale for the supply of towns and cities with pure water becomes a very nice and difficult operation, and now forms an important branch of civil engineering. The purity of spring water arises entirely from a process of filtration. Such water generally issues from a considerable depth beneath the surface of the ground, as is evident from the force with which it is discharged, and the equable temperature which it preserves throughout the year, being cool in summer, and in winter never freezing. Hence, in its course from the surface downwards, it must pass through a variety of soils, subsoils, and porous rocky strata; and it is by the perfect straining which it receives in percolating through such vast masses of filtering materials as must occur in the bowels of the earth, that it comes out at last with that admirable purity and transparency which we observe. But notwithstanding of this, spring water contracts, often among the rocks and minerals with which it comes in contact, various impurities, which become chemically dissolved in it, and from which it cannot be freed by filtration or any other mechanical process. It becomes often, therefore, unfit on this account for domestic use or the purposes of the arts, and is rarely equal, and never superior, to rain or river water properly filtered. The great importance of this operation, therefore, is evident, more especially from the frequent and serious failures which have taken place. But it may be better to defer the consideration of this branch of the subject till it can be resumed along with other hydraulic operations. See WATER-WORKS.

On a smaller scale, for domestic use, water filters have at different times been brought forward, and for many of them patents taken out. But as they are all very similar in principle, and none appears greatly distinguished above the rest, it is unnecessary to enter into a minute description of them. They are all liable to the objection of clogging up, and getting foul in the course of time. But if regularly cleaned, any of them will answer the purpose sufficiently well, the process of filtration being so simple and so easily effected. Several ingenious plans have been proposed for self-cleaning filters; the water, for instance, being made to ascend, or to pass alternately in opposite directions, and thus carry off the impurities by the process itself. But as none of these have succeeded completely, it will be found most convenient in small filters to clean them regularly; and those are therefore to be preferred which afford the means of doing this most conveniently.

In these filters the filtering material generally used is gravel, sand, porous stone, or unglazed earthenware. Charcoal is sometimes added, but this is more for its antiseptic and purifying quality, and is only necessary in the case of fetid or very impure water. Gravel, though the best of all materials on a large scale, is rather coarse in itself for small filters. But where it occurs in nature in very thick beds, the large mass of material makes up for the coarseness of the grain; and many instances occur where springs from this source are scarcely inferior in purity to distilled water. But in small filters it requires a bed of sand above it to detain the impurities. Sand is perhaps, after all, the best filtering material; it should, however, be clean sea or river sand, with no tincture of iron; this last is indicated by its reddish colour, which, if deep, will communicate an iron taint to the water, a circumstance harmless enough in itself for most domestic uses, but very prejudicial for the purposes of bleaching, paper-making, &c., in the arts.

In filters constructed with gravel or sand, all the apparatus that is necessary is a vessel to contain the filtering material, with a space above for holding a quantity of the water to be filtered, and a vessel below for receiving the filtered water. A simple and convenient apparatus of this kind was contrived by Messrs Hearman and Dean, potters, Redriff, described in Nicholson's Journal, vol.viii., and represented in Plate CCXLV. fig. 1, which exhibits the whole apparatus, and shows, in fig. 2, a shaded section. A is a vessel of stone-ware perforated with holes, m, at bottom, upon which coarse gravel, h, is laid, and on that again a stratum of fine gravel, and, lastly, fine sand, g. Or, otherwise, the bottom may be covered with a coarse cloth, which will render the graduated fineness of the gravel and sand less necessary. Upon the top of the sand is laid a perforated and loaded board or plate of earthenware, to prevent the sand from being disturbed when the water is poured in. B is a lower vessel, into which the filtered water from A drops, together with any sand that may escape from above. The clear water flows out through the neck ac into the vessel d for use.

On this plan an excellent filtering apparatus may be constructed at little expense, and in various ways, according to the means or conveniences most readily at hand; and the sand and gravel possess a decided advantage over filtering stones or other materials, from the great facility of cleaning, the latter being liable, from long use and the impurities lodging in their pores, to give the water an unpleasant taint.

A very simple arrangement in the above plan is to have a cylindrical jar of earthenware, as at fig. 5, with two partitions or false bottoms, dividing the height into three equal portions, the partitions resting each on a shoulder formed all round on the jar, or otherwise supported; the middle division for the filtering material, the upper one for the water to be filtered, and the lower one for containing the filtered water, with a stop-cock for drawing it off at pleasure, and a vent or opening at the top for the admission and discharge of air. The stop-cock must be placed three or four inches from the bottom, to allow full space for any sand or other material to deposit in the bottom; and the air vent must be carried up by a small pipe to the top of the jar, to prevent the water ever rising and overflowing by it. The partitions are made of earthenware, with a number of small holes to allow the water to descend. If sand only is used, it will be necessary to spread a layer or two of cloth under it, to prevent the sand getting through the openings. But if gravel can be had, it will be better to lay a bed of this, of two inches or more in thickness, on the bottom, and the sand above it. After the sand is filled in to the height of the upper partition, this is placed on as a cover, which having also a number of small openings in it, the water flows from the upper division downwards. The cover should press on the top of the sand, in order to prevent its being disturbed by the entrance of the water, or a layer of gravel should be placed on the top of the sand. If a layer or two of cloth be laid over the partition, it will serve to detain some of the impurities; and being frequently washed, it will make the filter continue longer clean. In order to clean this filter it is only necessary to remove the upper partition, take out the upper stratum of sand to the depth of two or three inches, wash it thoroughly, and replace it; or, if sand can be had readily, let the whole be replaced with a fresh quantity, clean and pure. This is on the whole, a very simple, perfect, economical, and convenient filter. It may be constructed of wood or other convenient material, and is both very easily formed, and easily kept clean and in order; and by varying the thickness of the bed of sand, it can be adapted to the exact degree of filtration required. The vessel also admits of any ornamental shape which taste may direct; and several elegant figures of this kind, contrived by the manufacturers, are shown at the bottom of the plate. If the lower division also were left unglazed and porous, an evaporation would go on from the outside, so that in summer it would act as a cooler as well as a filterer. Should much sediment be found collecting in the lower vessel, perhaps it would be better to place the whole apparatus upon a second jar, so wide as to allow the first to stand over it, and having a pipe in the bottom rising nearly to the top of the vessel, and open at the top. The filtered water would flow over this, and descend into the lower vessel, which being detached, could always be kept perfectly clean in the bottom. The upper end of the pipe should be covered with a fold or two of cloth. An ingenious plan, introduced by Mr Parkes, chemist, for drawing off the water in the reservoir of a filtering or purifying apparatus, though applied rather on a larger scale, might often be used with advantage in smaller filters, such as those for bleachers, paper-makers, and others. It is represented at fig. 6, and consists in connecting the stop-cock for the discharge of the water with a pipe and flexible joint terminating in a rose-head and floating ball, which keeps the head always immersed a little under water, and thus discharges it from near the surface, although the level of the water may vary considerably. The water here is also let in, not directly into the reservoir, but by an upright pipe having a funnel at the top, and descending to near the bottom, which causes the water to enter below, where it must deposit any impurity before this can rise to the surface.

In the sand and gravel filters here described, the water descends during the whole of the process, and this causes a slight tendency to carry minute particles of sand or other impurities along with it. To remedy this defect, various plans have been proposed for making the water after descending, again ascend; and this plan certainly possesses advantages, and is a decided improvement on the other. A very neat contrivance of this kind is that proposed by Professor Parrot of Paris, in Nicholson's Journal, and is represented at fig. 3.

It consists simply of an inverted syphon, which could be easily made of lead, having the bend filled with sand, and being covered on the top with gravel in each branch. The water entering the longer branch from a reservoir, passes through the sand, and rising up, is discharged at the other end, which should be three inches below the level of the first. When the section of the tube is four inches by two, and eighteen inches the length of sand, it was calculated that six Paris pints of pure water would be passed in an hour.

Another plan, proposed by Sir Henry Englefield, in the same work, is shown at fig. 4. It consists of a conical vessel placed within the one containing the sand, and allowing the water to descend on the inner vessel, and ascend again in the outer one.

But the simplest mode of effecting the ascent of the water is merely to divide the jar or filtering vessel by a vertical partition, open at the bottom. The water being then poured into the one side, descends to the bottom, passes under the partition into the other side, and ascends to the same level at which it entered, and can then be drawn off by a stop-cock or other means at pleasure. In that case a single vessel only is wanted, and contains itself the means of holding separately the filtering material, the water to be filtered, and the water after filtration. Upon the whole, this appears by far the simplest, most perfect, most economical, and most convenient filter of this, or perhaps of any kind, that can be contrived. All that is necessary is to have a pretty deep vessel of earthenware, wood, or other material; a tub or barrel will answer perfectly. Put a vertical partition in the middle, reaching nearly to the bottom, fill it half or three fourths full of sand, with a layer of gravel on the top, and it is ready for use. It may be found convenient, to prevent the gravel being disturbed, to place a cover over it, with holes. The water may either be drawn directly from the pure side with jugs, or let off by a stop-cock. Instead of sand and gravel, some have proposed to use charcoal, either by itself or in combination with the others; and an apparatus of this kind was contrived by Mr John Isaak Hawkins. Charcoal is only, however, when the water has an impure taint, filter by which cannot be removed by sand, that the charcoal should be resorted to, as it is more expensive, and more difficult to be cleaned. In long voyages, however, charcoal, as proposed by Hawkins, may, in conjunction with sand, be found very useful.

Instead of the water descending and then ascending, several ingenious methods have been contrived where the water only ascends through the filter, being let in below, and forced upwards by the pressure of an external head of water. The advantage of this is, that the filter can be cleaned by letting the water return and run out at the place where it usually enters. These plans, however, on a small scale, are hardly ever found to supersede cleaning. On this principle was the patent granted in 1791 to Mr Peacock, which he intended for small filters as well as those on a large scale. His apparatus consisted of a vessel of wood, lead, or other materials, of any shape or dimensions required, divided near the bottom by a horizontal grate or partition, upon which the filtering medium was spread in layers, beginning with the coarsest and ending with the fine. The lower division had a communication with a more elevated reservoir, from which the water flowed, and rose through the filtering bed. This consisted of gravel, sand, broken and pulverized glass or pottery, all well washed and sifted with sieves. On a similar principle Mr Brahmah had a patent for various modes of conveying and drawing off beer and other liquors, in which he describes a filtering apparatus which he proposed to place in the pipes to conduct the liquor. This was a cylindrical vessel, made of copper tinned within, or of any other materials; the upper end was closed by a lid screwed on by a flaunch, and the lower end terminated in a cone; a pipe was inserted into the vessel near the lower part, to introduce the liquor; and from near the upper end a similar pipe proceeded to convey it away. Between these two was fixed the filtering floor, consisting of flannel, sand in bags, sponges or other similar substances, which were spread upon a horizontal plate pierced full of small holes, and fixed in the cylinder above the filtering substance; and another similar plate was fixed and screwed down to confine it. The liquor was introduced by the lower pipe by means of a pump or otherwise, and ascending through the filtering floor, passed off in a clarified state; the impurities which are separated collect in the conical bottom of the cylindrical vessel, whence they can be drawn by a cock. This contrivance is adapted to give a slight filtration to a great quantity of liquor, such as beer; and Mr Collier, in his patent for the purification of oils, claimed the principle of filtering per ascensum combined with hydrostatic pressure.

Such are the different contrivances for filtering water, where sand and gravel are the materials; but notwithstanding their simplicity, filters of porous stone are those which have hitherto been most generally in use. Filter stones are procured from beds of sandstone, and sometimes a kind of limestone. They are found in Northamptonshire, Derbyshire, and other parts of England, and are often imported from the island of Teneriffe. The construction of the filter is extremely simple. A thick bowl or basin is formed of the stone, and mounted in a frame. The foul water, being poured into the basin, passes slowly through the substance of the stone, and drops into a receptacle below, in a perfectly transparent state. When the water is foul, a small quantity of mud or slime is found to collect at the bottom of the basin, and must occasionally be cleared out; but in the course of time the more minute impurities will insinuate themselves into the pores of the stone, and at last clog up the passage of the water. This is remedied by chipping away the interior surface of the stone one half or three fourths of an inch, because the impurities do not penetrate deeper into the stone. This will restore its action for a long time.

The basin shape used in these filters is obviously not the most proper, as it leaves an unequal thickness in the bottom for the water to penetrate, and also the pressure of the fluid is unequal. The bottom should be flat and of equal thickness, and in fact it would be much better not to make the stone a reservoir at all; but to use a filtering vessel such as above described, substituting for the sand and gravel a plate of the filtering-stone, of such thickness as may be found best adapted for the nature of the water. If a thin plate also were placed on the top of a thick one, it might then be removed without any trouble when it got foul, and replaced by another, without disturbing the lower mass; or several plates could be laid one above another till the requisite thickness for filtering were obtained.

Instead of pouring the water into the basin and allowing it to filter downwards, the water may be contained in a vessel in which the basin is nearly immersed; and in this way the water will filter partly laterally, but chiefly per ascensum, and leave much of its deposit behind, so as to enable the stone to continue longer clean. A filter on this plan was contrived by Mr Moult, and is described in the Transactions of the Society of Arts. The filtered water runs in and fills the basin, from which it can be taken out or let off by a pipe. This principle was also proposed by Mr Collier, and it possesses all the advantages we have already described in filtering upwards with sand; and if it were used with plates of the stone instead of a basin, it would be still more effectual.

The unglazed earthenware for filtering has been introduced as a substitute for the filtering stones. It is made of different compositions of pottery. The original patent was granted to Mrs Johanna Hempel, potter, Chelsea, in the year 1790. The proportions were four out of nine of tobacco-pipe clay, and five out of nine of coarse sea, river, or pit sand. This was intended for basins not exceeding one gallon. For larger capacities it is liable to fly or crack in the fire. A second composition, therefore, consisted of equal parts of tobacco-pipe clay and sea or river sand; or if this was found insufficient to prevent the cracking, three parts tobacco-pipe clay, one of Stourbridge clay, or clay from the surface of the coal mines, one of Windsor loam, or other loam of equal quality, and four of coarse river or sea sand; or four parts of tobacco-pipe clay, three of sand, and one of the burnt ground clay of which crucibles are made. The above ingredients are brought to the proper basin or bowl shape on the wheel, and are then hardened in the furnace like other pottery ware. The same objection, however, applies here to the basin shape as to the stones; in fact, it would be better to have this composition formed into plates of different thickness, which could then be used, one or more in number, in the filtering vessels, instead of sand or stone; one could be taken out when foul, and another put in, with the utmost facility; and this would certainly be attaining the perfection of a filtering apparatus.

Another still simpler plan would be to have one filtering vessel, with a vertical partition of the porous ware reaching to the bottom. The one half of the vessel would then be filled with the impure and the other with the pure water; and if several of these plates were at hand, one partition could be taken out when foul, and another inserted in its place, or there might be two porous partitions, and a solid one between them. The one would then ascend and run over the top of the solid partition, and again descend and pass through the second porous partition, which would therefore hardly ever require renewal.

A filter by Mr Robins of London, and for which he has taken out a patent, has been lately much recommended to the public. The external appearance is varied according to taste, and figs. 13, 14, 15, already referred to, represent some of the neatest and most elegant patterns. We are not acquainted with the interior construction, but understand it consists of various strata of filtering material, such as we have above described, arranged so as to act with peculiar efficacy. At the top there is a sponge neatly inserted into a recess above the upper partition, which is a very judicious and useful mode of detaining any impurities, and preventing them from getting into the filterer, the sponge being easily taken out and washed from time to time. The patentee, however, claims far higher powers for his invention than that of mere filtration. He states that the interior process involves a voltaic action, which decomposes soluble substances, and renders the water equal in purity to distilled water. We confess we have great doubts in regard to this, and think it irreconcilable with any idea of voltaic or chemical action with which we are acquainted. Still we have no doubt, from the various testimonials brought forward, that the filter is a good one. Fig. 14 is about three feet high, and one and a half broad, and is calculated to filter fifty gallons a day; the others in proportion.

From the great success with which water has been purified by filtration, attempts have been made to purify oils in the same manner; and an ingenious apparatus for this purpose was contrived, and a patent taken out, by Mr Joshua Collier, applicable also to water and other liquids, and of which an account will be found in the Repertory of Arts, vol. x. But in oils there is generally so many impurities, chemically or otherwise united, that filtration can do little for their purification. Charcoal is the substance chiefly employed. A filtering machine contrived by Mr Innes of Edinburgh is said to answer well for oils and other fluids. By it the fluid is forced through the pores of wood by the action of a forcing-pump. Such a degree of pressure is unnecessary for water; but for oil, quicksilver, and other valuable fluids, it appears certainly well adapted.