FILTRATION. The purification of water is effected by mechanical means on a larger scale at the present time than has ever before been known. To filter a small quantity of water is not a difficult !natter, but to filter millions of gallons a day involves engineering problems of magnitude. In most of the systems employed abroad sand-filters are used. The water is usually allowed to remain at rest in settling basins until the heavier matters have deposited, and then is passed to the filter-bed, through which it oozes slowly. This type of filtration has several serious objections. It is slow, and hence unable to meet heavy drafts on it. as in the case of fire. ']'he filter-beds acting tardily may become foul, which leads to the rapid and enormous development of bacterial life in them, and this may cause the water to become biologically less pure after passing through them than in its original state. There is no quick way of cleaning the filter-beds. In fact, there is no method of simple filtration known that is competent to handle on a commercial basis the water-supply of a large city.
The next step in the evolution of successful mechanical filtration was the addition to water of substances which react chemically with the bicarbonate of lime present in all natural waters, and form a precipitate which assists in removing the suspended matters by filtration. The addition of chemical substances to aid in clarifying water is very old. The most efficient of these substances are those which produce in the water precipitates of a gelatinous nature. The gelatinous precipitate thus formed in the water entangles and agglomerates the minute particles of suspended matter, he they mineral particles or microbes, and forms masses of sufficient size to be easily removed by the filter. Of the substances which produce in natural waters gelatinous precipitates, alum is the most readily obtained and is not surpassed in effi ciency by any. The alum and the bicarbonate of lime \Odell is in the water react on each other chemically. The alum is decomposed, and a gelatinous precipitate of aluminic hydrox ide, mixed with a basic aluminic salt, is thus formed. The most searching chemical examina tion fails to show the slightest trace of alum in water that has been treated with the proper amount of it and then filtered.
Alum has been used for many years as a "coagulant" for water with excellent results. The treatment usually consisted in adding a certain amount of alum to the writer. mixing it well and allowing the water to stand until the precipitate settled, after which the clear, super natant water was run off to the filters. While in this way a bright water was obtained, there were still difficulties which prevented commercial success on a large scale. The subsidence plant was unwieldy, and the same difficulties existed with the filters that have been mentioned. Three obstacles remained to prevent the commercial success of filtration of water on the im mense scale that large cities require. The first was the difficulties attending the cleaning of the filter-beds; the second was the time required for filtration: and the third, the great size of the filtration plant. It was reserved for us in America to solve the problem in a most in genious way, and to devise a process that has made the cleaning of the filter-beds simple and effective; that has diminished the time of filtration to a practical minimum, and has greatly reduced the size of the apparatus.
The principles of the process now generally in vogue here are briefly as follows: On its way to the filter the water receives the addition of a minute amount of a saturated solution of the coagulant, usually alum. The amount of coagulant added varies with different writers, and even with the same water at different times of the year. Usually it amounts to about one fifth to one third of a grain to the gallon. The water having received this small dose of coagulant, so small that it seems incredible that it should produce such remarkable results. passes, without stopping to settle, directly to the filters. The most generally adopted form consists of large closed cylinders of boiler-iron filled with sand, or a mixture of sand and coke. The coagulated water passes down through these filter-beds and collies out clear and spark ling, as delicious and as tempting as a mountain spring.