Textbooks generally remark at the outset that it is very necessary to use a funnel, the sides of which form an angle of 60°, this being the angle formed by the folded paper. Symes takes exception to this very exacting requirement. We do not get our straining bags or percolators made of such a shape, and that because our experience teaches us how much more suitable is a form in which the angle is decidedly more acute, the same volume of liquid in this latter form producing a longer column, and consequently a greater downward pressure. Then, as to the paper fitting the funnel, we know quite well that all else being equal, the less perfectly it fits, the more rapidly filtration proceeds, so that, for any useful purpose, it is quite unnecessary to insist on this very orthodox shape. One has, say, a pint of fluid to filter, and for this purpose a funnel of about 8 oz. or 10 oz. capacity is taken. Symes would use one of the long French pattern, fold the filter in plaits, and before opening it out, place it fairly well down in its position in the funnel, or, if there were reasons for not plaiting the filter, then it should be folded first in half, and then the two outer portions, representing rather more than 1/8 each of the entire paper, should be turned back so as to overlap each other slightly at the top, and not to form a very acute point.

In either case, the paper, whilst being fairly well supported, would have comparatively little surface adhesion, and but small resistance would be offered to the passage of the fluid in any part. Funnels of this shape, in much larger sizes, can be used with advantage, but it is then desirable to have them ribbed. The ribs of funnels (especially of large ones) to be of any real value, should be much deeper than they usually are, and should not run vertically, but spirally. A piece of muslin placed between the paper and funnel not only strengthens and supports the paper, but assists filtration by preventing adhesion; a cone formed of coarse hair cloth is still better. For larger sizes, say of 4 to 8 pints, it is advantageous to dispense with the funnel altogether, and to use an inverted cone formed of linen or stout calico, the edges being fastened to a wooden hoop, which, resting on a deep earthenware pan, forms an efficient support for the paper, the liquid passing through with equal facility over the entire surface, a suitable cover placed over it excluding the air, and the process goes on under comparatively satisfactory conditions.

A self-feeding arrangement can be fitted to this if it be so desired, in a very simple manner.

When, by exhausting the receiver, atmospheric pressure is brought to bear on the liquid in a funnel, then the latter should be of the orthodox shape, as with it air is less likely to pass; but this requirement militates against the advantage that such a method would otherwise possess. The point of the filter should be supported by a cone of platinum, or zinc, or by a packing of tow or prepared wool.

English paper-makers do not appear to have devoted much attention to the production of filters in any variety, and for this reason we derive our supplies chiefly from the Continent. It is a well-known fact that holding almost any of the common filters up before a strong light they are seen to be perforated more or less with minute pinholes, so that, when in use, it is only after these have become filled up that the whole of the solid matter is separated, and the liquid passes through bright. Each time a fresh portion of liquid is added the disturbance caused thereby is liable to remove some of the particles which are acting as a filling, and if this occurs, filtration again becomes imperfect. These filters, although very cheap, do not pay to use if time and convenience are taken into consideration. There is, however, considerable difference in the efficiency of the various kinds of filtering papers, even when free from this defect. The presence of animal matter, as in the grey filter, increases the strength, but diminishes its working capabilities, and the existence of mineral matters therein does the latter, but not the former.

The papers specially prepared by Schleicher and Schull are practically free from all extraneous matters, the pulp having been treated with hydrochloric and hydrofluoric acids, etc. They are an example of what can be accomplished in this respect, but at the same time they are too expensive for general pharmaceutical purposes, and, indeed, are only made in comparatively small sizes, suitable for analytical work. For operations requiring filters of 7 in. diameter (before folding), the Rhenish papers, No. 595, are, in Symes' opinion, the most suitable; for larger sizes, the French stout plaited or plain papers, taken in all their qualities, give the best results. The French also make a paper specially suitable for syrups, thick to support the weight, and yet sufficiently pervious to allow of fairly rapid filtration. Symes finds, however, in very large sizes, a double sheet of Rhenish paper in an inverted case of linen, as already described, answers even better.

Filter-paper which has been immersed in nitric acid (sp. gr. 1.42) and washed with water is remarkably toughened, the product being pervious to liquids, and quite different from parchment-paper made with sulphuric acid. Such paper can be washed and rubbed without damage, like a piece of linen. The paper contracts in size under the treatment, and the ash is diminished; it undergoes a slight decrease in weight, and contains no nitrogen. Whereas a loop formed from a strip 25 mm. wide of ordinary Swedish paper gave way when weighted with 100-150 grm., a similar loop of toughened paper bore a weight of about 1.5 kilo. The toughened paper can be used with the vacuum pump in ordinary funnels without extra support, and fits sufficiently closely to prevent undue access of air, which is not the case with parchment-paper. An admirable way of preparing filters for the pump is to dip only the apex of the folded paper into nitric acid and then wash with water; the weak part is thus effectually toughened.