FILTRATION 



373 



first the upper and purer portion of the oil is forced through the tube b into tho filter, 

 and thonco out through the pipe I. When the fouler oil follows, it deposits its im- 

 purities in the space under the partition c, which may from time to time be drawn off 

 through the stopcock Jc, while the purer oil is pressed upwards through the filter. In 

 this way the different strata of oil in the cask may be filtered off in succession, and 

 kept separate, if found necessary for sale or use, without running any risk of mixing 

 up the muddy matter with what is clear. According to the height of the water-cistern 

 n will be the pressure, and of course the filtering force. When the filter gets choked 

 with dirt, it may be easily re-charged with fresh materials. 



It has been for many years the custom of the Water Companies to send the water 

 taken from the river through filter beds, prepared usually of sand and gravel. It Was 

 long thought that the effect of these filter beds was merely to separate the solid in- 

 soluble matters suspended in the water. It has, however, been shown by the investi- 

 gations of the late Mr. Henry M. Witt (a chemist of peculiar promise, lost too soon to 

 science, and ere yet the world could recognise his powers), that these filter beds had 

 the power of separating many of the dissolved substances from the water ; that, in 

 fact, the soluble salts of lime, and the like, were removed by some peculiar physico- 

 mechanical force, resident, as it appears, as a surface-force, in all porous masses. 

 There are many very remarkable examples in nature of the operation of this power in 

 producing beds charged with metalliferous matter, some of which will be described 

 under the head of MINING. 



Mr. H. M.Witt communicated to the 'Philosophical Magazine' for December, 1856, 

 an account of some experiments on filtration, which are of much value. Many of his 

 experiments were made at the* Chelsea Water-works, and they appear of such interest 

 that we quote the author's remarks to some extent: 



' The system Of purification adopted by the Chelsea Water- works, at their works at 

 Chelsea, consisted hitherto (for the supply has by this time commenced from Kingston) 

 in pumping the water up out of the river into subsiding reservoirs, where it remained 

 for six hours ; it was then allowed to run on to the filter-beds. These are large square 

 beds of sand and gravel, each exposing a filtering surface of about 270 square feet, 

 and the water passes through them at the rate of about 6|- gallons per square foot of 

 filtering surface per hour, making a total quantity of 1687*5 gallons per hour through 

 each filter. 



' The filters are composed of the following strata, in a descending order : 



ft. in. 

 2 6 

 



1. fine sand . . . 



2. Coarser sand . . . . . . . .1 



3. Shells 06 



4. Fine gravel . . & -. .-. -.; . . .03 



5. Coarse gravel . .*._..-,. . . . .33 



These several layers of filtering materials are not placed perfectly flat, but are dis- 

 posed in waves, and below the convex curve of each undulation is placed a porous 

 earthenware pipe, which conducts the filtered water into the mains for distribution. 

 The 'depth of water over the sand was 4 feet 6 inches. The upper layer of sand is 

 renewed about every six months, but the body of the filter has been in use for about 

 twenty years. 



' Samples of water were taken and submitted to examination : 



1 1st, from the reservoir into which the water was at the time being pumped from 

 the middle of the river. 



' 2nd, from the cistern, after subsidence and filtration.* 



Experiments were made at different seasons of the year ; but one of Mr. Witt's 

 tables will sufficiently show the results. 



