IIANICAL FILTRATION is, of all, a straining process, in which the natural capabili ties of the filtering medium are aided by coagu lation. The gelatinous coagulating material, combined with the more or less sticky organic matters in the water, and with the finely divided clay and silt, form a layer on the surface of the filter and for a greater or less extent on the sides of each sand grain. Thus the effectiveness of the strainer is greatly increased by the reduction in the interstitial passages and by its adhesive qualities. After a comparatively short period, ranging from say two days to twelve hours, the filter becomes clogged. The impurities penetrate the whole mass, so all the filtering material must be washed. This is done by the simple mechani cal process of reversing the flow of through the filter, so it passes upward from the bottom. The dirty water is wasted at the top. Prior to or in connection with the reversed Ilow the sand is loosened by means of power-driven revolving rakes, gradually lowered into the filtering ma terial. In place of the rakes. compressed air, admitted from the bottom, is used in me chanical filters. The filter sand is supported on a false hot tom, in which are placed the pipes for collecting the filtered and admitting the wash water. Strainers of perforated metal plates or wire netting give the water access to and permit it to flow from these pipes. The coagulant is admitted to the water before it comes to th filter and is given from a few minutes to severa hours' time to act before filtration takes plaec If the period is brief a coagulating chamber is placed directly beneath the false bottom of each filter tank; otherwise one or at most two larger coagulating basins, or settling reservoirs in which a coagulant is used, are employed. The coagu lant is made of the desired strength by proper dilution. The quantity applied to the water be ing treated is proportioned by various ingenious devices to the changes in rate of pumpage or water consumption. Among these may be men tioned some form of pump, driven or controlled by the flow of water through the supply pipe to the basin or filter. The amount. of coagulant used varies from about one-half to three grains per gallon of water, but rarely exceeds one grain except in waters with high turbidity or that have some other unusually troublesome feature. The filtering material used in mechanical filtration is generally confined in wooden or steel tanks which give a filtering surface with a diameter of only some 10 to 16 feet. The depth of sand is about the same or a little deeper than in slow sand filters, but the rate of filtration runs from 90,000,000 to 125,000,000 gallons an acre. The tanks and operating mechanism are almost in variably placed in a building, and are therefore quite tree from the extremes of heat, and from rain or snow. Under proper conditions of de
sign and operation, mechanical filtration will remove almost if not quite as many of the bac teria as the slow sand process; more turbidity and color; but less dissolved organic matter. It is more commonly applied to turbid, highly col ored waters than to those which are objection able on account of sewage pollution. In first cost and in labor charges, mechanical filtration has the advantage, but this is largely and some times wholly offset by the cost of the coagulant, of operating the machinery and providing wash water. If the slow sand filters have to be cov ered the capital charges may be so increased as to place them at a financial disadvantage, as corm pared with mechanical filtration, although cov ering filter beds often materially reduces their cost of operation. The choice between the two systems depends upon such a variety of local conditions that each ease should he most care fully considered on its merits before adopting either system.
llnicrontuAL SKETCH. 'Flu- history of filtration, as applied to public water supplies, date* from 1829, when James Simpson built some filter beds for the Chelsea Water Company, of London, England. Following that action, slow sand filtra tion was first gradually, then rapidly adopted in England and on the Continent of Europe. It was not until about 1887 that the real action of slow sand filters was Prior to that time they were supposed to be strainers only, effeeting little or no other chemical change upon the water and leaving the bacteria unharmed. In fact, it was not so many years before this that the relation of bacteria to disease (see InsEAsE, (Imo Titway oF). and partieularly of water borne germs, was established. The perfection by loch of methods applicable to the enumeration of water bacteria made possible the announcements by Percy Frankland and others about 1887 that filter beds removed nearly all the bacteria in water. England, Dermany, and America have each played important. parts in the scientific and practical development of the various aspects of the bacterial purification of water. The work in America has been done very largely by the 3,1assachusetts State Board of Health, at its Lawrence Experiment Station (see annual re ports, 1887 to date), as has been mentioned under SEWAGE DISPOSAL. At Lawrence slow sand filtra tion has been studied in great detail and a mass of scientific data has been accumulated and valuable conclusions drawn therefrom. Other American investigations, begun a number of years later, are mentioned farther on.