The water companies have large storage and sedimentation basins with an aggregate capacity equal to nine days’ supply, but the proportion varies widely with the different companies. It is desired that the water held in reserve shall be alone used while the river is in flood, as, owing to its increased pollution, it is regarded as far more dangerous than the water at other times; but as no record is kept of the times when raw sewage is discharged, and no exact information is available in regard to the times when the companies do not take in raw water, it can safely be assumed that a considerable amount of raw sewage does become mixed with the water which is drawn by the companies. The water drawn from the river is filtered through 113 filters having an area of 116 acres. None of the filters are covered, and with an average January temperature of 39° but little trouble with ice is experienced. A few new filters are provided with appliances for regulating the rate on each filter separately and securing regular and determined rates of filtration, but nearly all of the filters are of the simple type described on page 48, and the rates of filtration are subject to more or less violent fluctuation, the extent of which cannot be determined. The area of filters is being continually increased to meet increasing consumption; the approximate areas of filters in use having been as follows: 1839 First filters built 1855 37 acres 1866 47 acres 1876 77 acres 1886 104 acres 1894 116 acres There has been a tendency to reduce somewhat the rate of filtration. In 1868, with 51 acres of filters, the average daily quantity of water filtered was 111,000,000 gallons, or 2,180,000 gallons per acre. In 1884, with 97 acres of filter surface, the daily quantity filtered was 157,000,000 gallons, or 1,620,000 gallons per acre; and in 1893, with 116 acres of filter surface and 195,000,000 gallons daily, the yield per acre was 1,680,000 gallons. Owing to the area of filter surface out of use while being cleaned, the variations in consumption of water, and the imperfections of the regulating apparatus, the actual rates of filtration are often very much higher and at times may easily be double the figures given. Evidence regarding the healthfulness of the filtered river-water was collected and examined in a most exhaustive manner in 1893 by a Royal Commission appointed to consider the water-supply of the metropolis in all its aspects with reference to future needs. This commission was unable to obtain any evidence whatever that the water as then supplied was unhealthy or likely to become so, and they report that the rivers can safely be depended upon for many years to come. The numbers of deaths from all causes and from typhoid fever annually per million of inhabitants for the years 1885-1891 in the populations receiving their waters from different sources in London were as follows: