s much used in chemical processes. The materials employed are unsized paper, cloth, flannel, tow, sponge, sand, pulverized glass, flints, porous stones, earth-pan, earthenware, &c. In fig. 12 is represented a neat little filtering apparatus constructed by Mr Parkes; and the following remarks and directions are taken from Mr Faraday's excellent work on Chemical Manipulation. Of the above substances, the first is almost exclusively in use at the laboratory, a few of the others now and then being resorted to only on particular occasions.

"Funnels are continually necessary to support the paper through which filtration is to take place. The ordinary funnels required in the laboratory for the passage of fluids answer the purpose very well, but they are of use for supporting filters, when, from their necks being broken off, they are otherwise unserviceable. They may be either of glass or good Wedgwood or other earthenware; those of glass are to be preferred, because the progress of the filtration and the state of the filter can be better ascertained in them. Metal funnels should not be allowed in the laboratory.

"The funnels containing filters may be frequently supported by the glass or jar intended to receive the filtered fluid, and at other times by the rings of retort stands. But notwithstanding these facilities, so frequently is the operation required, a laboratory should be supplied with at least one filtering stand of considerable size. The end of such a stand is represented in fig. 7. It should be fifteen inches wide, three or four feet long, and have an interval between the bottom and top of twelve inches. The top should have a number of round holes made in it at intervals of six inches, differing in diameter from one and a half to three inches; two or three small ones of half an inch or an inch in diameter being intermixed. The whole should be well made and firm, so as to support considerable weights, the top having frequently to answer the purpose of a table, and to sustain jars full of solutions, as well as to bear funnels and filters.

"It is no easy matter for the chemist to obtain unobjectionable filtering paper, and yet it is of such importance in reference to the duties it has to perform, that he should not spare pains to procure the best possible. It should be so porous as to admit the free and ready passage of fluids; so close as to retain the finest solid particles; so strong as to bear the weight of a considerable quantity of fluid; and so pure as to give nothing to the solution, or if heated with the substance retained upon it, to occasion no mixture of ashes. Some chemists use plate paper, that is, the paper of copperplate printers. It is very porous, and yet there are few precipitates that will pass through it; at the same time it is often tender, generally yields a considerable quantity of ashes when burnt, and is inapplicable for minute filters when very small quantities of fluid only are to be worked upon, because of its thickness, and consequent waste of the portion of solution imbibed by it.

"It is amongst the thinner varieties of unsized paper, or white blotting paper, kept by some of the stationers, that the chemist will probably find the kind best suited to his purpose. It should be so strong, that a single filter of it, capacious enough to hold a pint of water, should not break with that quantity, even though some degree of agitation be given to the funnel containing it. Its porosity, that is to say, its comparative freedom from size, for it is mostly sized in a slight degree, may be judged of by holding it to the tongue, and observing how it absorbs moisture; and by a cautious pull its strength may be ascertained whilst in such moistened state. The student who is unused to the examination of papers will, however, better judge of its capability of allowing fluid to pass, by actual trial with water; a pint filter filled with clean water should allow the fluid to run in a considerable stream.

"The best method of judging of the purity of paper is to burn it and examine its ashes: the fewer it yields the better is it adapted for filters. A demy sheet should not yield more than one and a half or two grains of ashes altogether. If it contains more, their solubility or insolubility should be observed, that the student may be aware of the impurities that may probably be imparted to solutions in very delicate experiments. In minute cases of investigation, sulphuric acid may frequently be traced to the sulphate of lime existing in the filtering paper.

"Filtering paper should be cut ready for use into different sizes. A demy paper furnishes useful sizes for filters, where the sheet is separated into four, or six, or nine parts; and parcels of each of these sizes being prepared, they should have a string passed through the corners, and be preserved for use in a clean place. The previous cutting of the paper in this way is very convenient, as readily supplying the sizes that will be wanted, and in preventing the waste that would occur by carelessly tearing up a sheet each time a filter is required.

"On preparing a filter, the piece of paper should be first examined, by looking through it against the light, to ascertain that it is free from holes. The simplest filter is made by folding the paper twice in opposite directions, so as to bring the four corners together, and by opening one corner from the other three, so as to produce an irregular conical cavity, as in fig. 8. Such a filter being put into a funnel, and then filled with liquid, will immediately permit its passage; but from the similarity of form between the filter and the funnel, and the close adhesion of the former to the latter over by far the greater part of its surface, considerable obstruction is opposed to the passage of the fluid, and the operation is retarded. For this reason different contrivances have been recommended, to separate the filter here and there from the funnel, and allow passages for the fluid. Lavoisier recommended small glass rods inclined along the funnel before the filter is put in. Straws are used in a similar way, and they certainly open channels in their immediate neighbourhood, by which the fluid may flow down. By other chemists ribbed funnels are recommended; but it is difficult to find a funnel so deeply ribbed as to support the paper in such a manner that it shall not touch the glass in every part, and if they do not perform this they are of no use.

"The best expedient by far is so to fold the filter that ribs may exist in the paper itself; and this may be done so as not only to allow numerous free passages for the fluid between the filter and the glass, but also to allow of ready transmission through its whole surface, and not of one half only, and even that imperfectly, as in the former case. For this purpose, the paper is first to be doubled, and in this state is again to be folded in half; each half folded into quarters, and each quarter into eighths, the folds being all on the same side, and radiating at equal distances from the middle of the folded edge to the other edges. Fig. 9 represents the doubled paper thus divided into eight parts. Each eighth is now to be divided into half by folds in the opposite direction, but in lines still originating at the same centre, which makes the doubled piece resemble a child's paper fan, both when closed and when a little open: it is represented by fig. 10. Whilst in this state, the projecting corners should be taken off by a knife; folding the whole up tight like a closed fan, and making the section at about a. Being now allowed to expand a little, the originally doubled sides are to be separated from each other for the first time, but without disturbing the angles or bending the ridges or ribs which they form. Having opened it sufficiently to separate the cut edges from each other, it will be found that the paper is equally divided into parts forming alternate external and re-entering angles, except at the two edges bb, where two external angles come together. Here the intervening portion of paper between the two contiguous external angles should be folded, by bringing the latter together, and creasing the paper down, so as to form a re-entering angle between them: this should be done at both places. Then opening the paper sufficiently to bring the bottom into proper shape, by thrusting out the part which is convex within, so as to make it project externally, the filter is completed, and being put into a funnel, is ready for use. Its appearance, when perfectly formed, is represented at fig. 11.

"It is necessary, in making these filters, that the folds be not continued to the point, but that they should stop about half an inch short of it; for if completed to the bottom, the frequent action of the fingers in folding will..." so far break down and destroy the texture of that part, that a hole probably will appear before the filter is finished; or if not, it will be so weakened as to be unable to bear a quantity of fluid without breaking. Hence that part of the filter will, during the folding, assume a concave form; and the regularity of the folds, which by practice may be easily attained, must be strictly attended to in the upper part, but may be dispensed with at the lower. When opened out, therefore, for the completion of that part, before the filter is put into the funnel, all that will be required is to push that side of the bottom outwards which is convex inwards, doing it so carefully as to cause no injury. The folds of the filter should be distinct and sharp, an effect which should be obtained by a single decisive pressure, and not by much fin- gering. No wrinkle or mark should appear in a well-made filter, except the folds described, and the portion of paper between the folds should be as stiff as when first taken up. The filter should be handled lightly during the whole of the operation, and never be opened out more than is represented in the last figure. It is best to keep the folds as close, and the whole filter as compact, as possible, dropping it loosely into the funnel, and permitting the fluid poured in to do the office of opening it out. It should be so regularly made, that when thus expanded by the fluid, the external angles of the folds should touch the glass at equal distances from each other, except at two opposite places where smaller divisions exist; and unless a large quantity of fluid be present, the angles at the upper part should remain nearly as sharp when the liquid is introduced as they were before. Below they gradually pass into the rounded surface, forming the centre or bottom of the filter, which will be about the size of a sixpence or shilling, according to the dimension of the funnel in that part. These filters leave such free space between themselves and the glass, as to admit of the passage of a far greater quantity of fluid than is necessary; and no obstruction being placed against the external surface of the paper, the whole acts in the filtration, and that in the most favourable manner.

Of these two kinds of filters, which may be distinguished as the plain and the folded, sometimes the one and sometimes the other is preferable. When the object is to cleanse and purify the fluid, that being the valuable part, it is most rapidly and effectually attained by a folded filter; but when the precipitate is the part required to be taken care of, it is generally desirable not to spread it over the irregular and extensive surface of such a filter, but, by using a plain one, to retain it on a surface of only half the extent. It is then likewise of such uniform thickness in every part, as best to admit of the operation of washing by having successive portions of water passed through it, and when the filter is opened out, the precipitate is delivered in one continuous portion, and not divided into many parts, as happens with the folded filter.

When a single filter is judged too weak to hold the mass of fluid it is required to sustain, a double one should be used. If the filter is to be a folded one, then a double thickness of paper is to be taken; but if plain filters are used, the two should be made separately, and put one into the other, in such a manner that the three thicknesses of the one may come against the single thickness of the other. Occasionally it is proper to strengthen the bottom of the filter, which, not being supported by the glass, has to support the greatest column of fluid; but not so as to obstruct the upper part, where a filtration as free as possible is required. In such a case a smaller filter is to be added to the exterior of the large one, so that it may not interfere with the precipitate within, when it is necessary to remove it from the paper.

The filter and funnel being ready, and placed on the stand in one of the holes before mentioned, over a glass ready to receive the liquid, the mixture to be filtered is to be poured in. It should not be poured from a great height, nor upon the middle of the filter, but down the side, the force of its descent being diminished as much as possible by good management, lest it break a hole through the paper; for the same reason it is better to pour it down the rod. If the first portions of fluid which pass be not clear, they should be returned into the filter, and a second glass should be placed beneath. The solid matter in the liquid to be filtered will soon, by adhering to the paper, cause clear filtration, except perhaps in one or two particular cases, as with precipitated oxide of tin, &c., on which occasions a double thickness of paper must be used. On changing the vessels beneath the filter for the removal of the clear solution, it should be done so that no drop be lost. By inclining the empty vessel, its edge may be brought under the funnel before the full one is removed; and in no case should a vessel be left under the filter, in which there is not sufficient room for the contents of the latter if it should break. Such an accident ought not to occur, but ought always to be provided against.

Sometimes hot filtrations are to be performed. Oils filter better hot than cold; tallow and cocoa-nut oil may be passed through paper when hot; and many solutions must be filtered at high temperatures, because of the greater solubility of the ingredients under those circumstances. In these cases the fluids should be heated in a flask or basin before they are poured into the filters, and the filter funnels being placed in the glasses or jars, and after the fluid is poured in, covered over with a basin or glass plate, the whole should be enveloped in a piece of flannel or a dry cloth. If, from the length of the operation, or other causes, the flannel or cloth be not sufficient, it may be dispensed with; and the glass and filter being placed on a warm part of the sand-bath, should be covered with a box or vessel large enough to rest on the sand, and thus form a hot air-chamber for the process. When the operations are upon a small scale, a paper cone is sufficient to cover the vessels on the sand-bath, and keep them hot.

When the object is hastily to filter a fluid for the purpose of removing pieces of dirt, then a little loose tow at the bottom of a funnel, or a piece of sponge slightly thrust in, is often sufficient for the purpose. Upon some few occasions filters for acids are required. These are generally recommended of powdered glass, being arranged somewhat in the manner described for lixiviation. Pieces of glass are put into the neck of the funnel; upon these smaller fragments, then again other layers of particles diminishing in size, until a moderately fine powder has been used, the top being finished with a layer of small fragments, to prevent disturbance by pouring. The pieces of glass should be well cleaned before pulverization, and, when arranged, some water should be passed through the filter, to remove alkali or other matters that may be separable from the glass, which should be that of wine-bottles, and not flint glass.

Another mode of separating a fluid from the finely-divided solid matter it may contain, is to allow the latter to deposit, and then to remove the former. This is called decantation, and is a process much superior to filtration in many analytical experiments, and recommended in preference both by Lavoisier and Berzelius.

For further information on this subject see Repository of Arts; Hempel, vol. ii. p. 230; Collier, vol. x. p. 389; Peacock, vol. xi. p. 221; vol. xiii. p. 140; Nicholson's Journal; Hearman and Dearn, vol. viii. p. 126; Parrot, vol. ix. p. 40; Sir H. Englefield, vol. ix. p. 95; Lillach, Phil. Mag. vol. vi. p. 240; Mém. Acad. Par. 1745, Hist.