As before stated, the action of a filter is either mechanical or chemical. Solid particles which are too large to pass the pores of the filter are arrested; other particles adhere to the surface of the filtering material, even after they have been wholly dissolved. Furthermore, the air contained in the pores of the filtering substance oxidizes the dissolved organic matter and thus destroys it. It follows that the more extensive the area of filtering material the greater the power of holding impurities by adhesion; while the more frequently and thoroughly it can be cleansed and aerated the more efficient its action.

More or less elaborate arrangements are provided for cleansing public filter beds. But as a rule this is irregularly and carelessly done, and hence just in proportion to the efficiency of the filtering material does it become clogged. "Inadequate area and infrequent cleansing/' says Prof. Nichols, "are the common faults of many so-called filters. The most that can be said of the majority of such filters is that they act with greater or less efficiency as strainers, but they do not remove the finer and more dangerous impurities".

Furthermore, as the filter beds are not covered, the exposure of the shallow water to the hot sun in summer assists the development of vegetable life, which causes a disagreeable odor and taste in the water. Doubtless organic putrefaction may be assisted in like manner. In cold weather the filter beds are frozen and cannot be used.

Prof. Ripley Nichols states that sand is the best material yet used practically on a large scale for artificial filtration. Visible suspended particles and an appreciable proportion of organic matter actually in solution may be thus removed. He lays special stress upon the need of abundant area, frequent cleansing and renewal of the filtering material, constant supervision, protection from the sun, and prompt distribution of the filtered water to consumers. Where a water supply is taken from deep wells, basins or collecting galleries which are fed by "ground water," the supply will go through a process of natural filtration. But the water from such sources is not always potable.

Dr. Smart, in his paper recently read, testified to the satisfactory results achieved by natural filtration, in the percolation of the rain-fall through sand, gravel and other porous soils. If a public water supply could be subjected to the same process of filtration by passing it through a sufficient mass of material, equally good results would follow. In the case of the soil, there are usually intervals between rain-fall during which matters caught in its pores are oxidized, otherwise the pores would become clogged and a source of evil. This further illustrates the need of frequent and thorough cleansing of all filters.

With regard to filters, it is essential that the material employed should not act injuriously upon the water. The mechanism should be simple and the appliance inexpensive; the filter should be easily cleansed or the material renewed; and lastly, not only all suspended particles, but also, so far as possible, all dissolved organic matter should be removed.

The Japanese use a porous sandstone filter, hollowed in the shape of an egg, through which the water percolates into a receptacle underneath; the Eygptians resort to a similar device; the Spaniards use a porous earthen pot. But these devices cannot be thoroughly cleansed; some impurities will remain in the pores of the stone. Spongy iron and car-feral are open to the same objection. The various forms of filters that are screwed to the faucet have not enough filtering material in them to be of much utility, and they very soon become foul and offensive. Buck says: "There is no material known which can be introduced into the small space of a tap-filter and accomplish any real purification of the water which passes through at the ordinary rate of flow". Complicated closed filters which cannot be cleansed condemn themselves. Parkes, in his "Manual of Practical Hygiene," says: "Filters where the material is cemented up and cannot be removed ought to be abandoned altogether/' Filters in which the water comes in contact with metallic surfaces, either iron, lead, tinned iron or zinc are objectionable from their appreciable influence upon the water retained in them for any considerable time. Pure block tin is the least objectionable of any of the metals. The aim of most filters is to remove impurities from the water as rapidly as it escapes from the faucet. Effective filtration cannot be accomplished when the water does not remain long enough in contact with the filtering material to become purified. Slow filtration or purification is therefore best. Of all the filtering materials mentioned, sand and charcoal are the two that accomplish the best results.

The radical objection to most filters is that, to use Prof. Franklin's words, "the polluting matter removed from the water is stored up in the pores of the filter, and in time develops vast numbers of animalculae, which pass out of the filter with the water, rendering the water more impure than it was before filtration. It is, therefore, necessary to remove and purify the material".

These statements demonstrate the vital necessity of filtration. The question next arises, How far does or can filtration purify? To what extent can it be depended upon to guard the public against the dangers from the pollution of a water supply, and is it applicable for use upon a large scale? After studying the results obtained both abroad and in this country, this inquiry can be answered emphatically in the affirmative. But to be practicable the undertaking must be carried on upon a large scale. By this I mean that the water to be purified must be passed through a body of filtration material of sufficient volume to insure the complete removal of all matters held in suspension, however minute. On this account the numerous patented appliances for domestic filtration cannot be recommended. They are too small to perform their duty. It is like setting a child to do a man's work.

It is capable of demonstration that the water supply of the largest cities, no matter how great its volume, can be effectually and economically filtered. It is simply a question of ways and means. There are today in use in many industrial establishments in this country and elsewhere, including paper-mills, breweries and bottling factories, which consume enormous quantities of water, filtering appliances which have borne the test of years of trial, and which are delivering large volumes of filtered water of a purity, transparency and general quality which would astonish the average water-drinker in our principal cities and towns.

Without going into the question of what mineral matters it is better without, or what salts it is well to have dissolved in water, the best means of freeing it from those organic impurities that lead to all sorts of trouble in practical work should principally be considered.

If we dissolve some pure sugar or salt in a glass of clear water we will be unable to tell, from the appearance of the water, that it has taken into solution any foreign body, showing that the clearness of water is no guarantee of its purity. This can also easily be seen by filtering some dirty water, say, from a drain, through blotting paper until it is quite brilliant, and then boiling it in a glass beaker, when flocks of animal or vegetable matter will be visible floating in it, which have been coagulated by the high temperature, and which were before invisible, showing that though the nitration brightened the water it did not purify it. The energies of filter makers, therefore, have for many years been devoted to perfecting some method whereby both the floating bodies that impair the clearness of water may be removed, and also any hurtful organic matters that it may hold in solution.

The result has been an immense number of patent filters of every imaginable pattern and design, but many having some defect that renders them unfit for use in manufactories where a large quantity of pure water is required. They do very well, however, for domestic use where the quantity of water treated is small. The best substance that has been found, among others, suitable for filtering purposes is charcoal, and those favor-ing it are loud in its praises. Indeed the properties of animal charcoal render it a very desirable filter-medium. It clarifies the water, separates from it floating and visible matters, and at the same time oxidizes and removes the organic impurities held in solution. Charcoal has always proven eminently satisfactory, if proper attention has been paid to it, that is, if it has been frequently renewed or regenerated. To sum up, a water-purifying apparatus should combine the following:

1. Capable of acting both upon the impurities held in solution and upon those in mechanical suspension.

2. It should be composed of materials incapable of communicating the slightest taint to the water passing through it. Here animal charcoal is pre-eminently excellent, for it is absolutely insoluble in water. Thus it is impossible that the charcoal can itself be a means of imparting impurities

3. Its action should remain unaltered, and should be attended with the least possible trouble or necessity for attention on the part of servants.

4. It should be capable of retaining its purifying properties for a long period.

5. It should be so arranged that the filtering medium may be easily got at for cleansing or for renewal.

6. It should have sufficient filtering area to supply the necessary quantity of water without delay or inconvenience.

7. And finally, the price of the filter should be such as to place it within the reach of all. As there are filters and filters, so there are waters and waters; and we would remind the carbouator that too much should not be expected of them. Despite the most careful filtration, some water will remain impure. In such cases the fault should not be laid to the filter. Nothing short of chemical treatment will solve the problem. Where a first-class filter does not accomplish the desired result, rest assured some more searching agent is needed. Put the blame where it belongs, on the water.

As a matter of course, each man who displays a filtering apparatus, has the best. However, it is not for us to settle this mooted point. The trade is quick to recognize a good thing, and a first-class filter will not long remain unappreciated.