In 1893 Mr. Edmund B. Weston, C. E., of Providence, R. I., conducted for the water department of that city a series of experiments upon the purification of the water of the Pawtuxet River by means of mechanical filters. Though less extensive than the experiments above mentioned, they are of historic interest as giving the first adequate demonstration of the possibilities of that method of purification.
The system of mechanical filtration, or the 'American System,' as it is sometimes called, differs from natural sand filtration by the use of alum or some similar coagulating substance before sedimentation and filtration, by the higher rate of filtration employed and by the use of certain mechanical devices for cleaning the sand beds. The application of this process to the treatment of turbid water was next investigated. In 1895 the Louisville Water Company undertook a most extensive series of experiments to determine the relative efficiency of various types of mechanical filters in the purification of the water of the Ohio River. The work was placed in charge of Mr. Geo. W. Fuller, C. E., who was assisted by a large corps of trained assistants. For nearly a year the experiments were carried on without interruption: the filters were operated by the companies interested in them, and their efficiency was determined by Mr. Fuller on behalf of the water company, who had at hand a complete laboratory equipment and who used every means known to science in the analysis of the water before and after treatment. The most important result of these experiments was to prove beyond doubt the applicability of mechanical filtration to the purification of water rendered turbid by the presence of fine particles of clay.
The experiments in Louisville were followed in 1898-9 by a somewhat similar investigation at Cincinnati, 0., also conducted by Mr. Puller. As in Louisville, the water supply is taken from the Ohio River, but the character of the water at this point is not in all respects the same as that farther down stream. The problem in Cincinnati was to determine whether the English system of sand filtration or the American system, involving the use of a coagulant, was best suited to the purification of the water, and whether any preliminary treatment of the water before filtration was advisable. To solve this problem the Board of Trustees, Commissioners of Water Works, decided to appropriate for needed experiments a sum equivalent to about one year's interest on the probable cost of a plant for filtering the supply of the city. The equipment consisted of four steel tanks, each with a capacity of 100,000 gallons, fifteen experimental filters, arranged for operation under different conditions, and a large laboratory fully equipped for chemical and bacteriological work. After a period of continuous operation, covering about ten months, the evidence showed that either the American system or the English system operated with preliminary coagulation and sedimentation would satisfactorily purify the water, but that the American
