2G-> 



THE CIVIL ENGINEER AND ARCHITECTS JOURNAL. 



[ACGIST, 



and t}ie action of the winds, tlje waters are still further purified, 

 as they have a tendency to j,^ve i>ff, wlien thus exposed, many of 

 the gases they may have combined with during the decomposition 

 of animal and vegetable matter, which gases cannot otherwise be 

 removed by nitration: it has likewise the effect of removing hard- 

 ness from the water, and rendering it more fit for all domestic and 

 other purposes. The Thames water, for instance, owing to its 

 being much loaded with organic matter received from the towns 

 and villages on its banks, after being kept for two or three months 

 in a closed cask, when opened, the water is found to be black, 

 nauseous, and unfit for use; but on being exposed and agitated it 

 deposits a ([uantity of slimy mud, and becomes clear and sweet. 



The water, after having been allowed to subside, the next point 

 to be attended to is the process of filtration, in order to remove all 

 the remaining impurities. In all filters, the great principle to 

 which attention should be most particularly directed, is that of 

 having a large extent of filtering surface, it being greatly owing to 

 want of attention to such an important point that is to be attri- 

 buted the inefficiency of the filters of many water-works. The 

 great aim in the construction of many filters, is that of having the 

 bed of sufficient fineness, so as to prevent the impurities from 

 passing through, causing the filter bed thus to act like a sieve. 

 Besides this, however, in order to filler water thoroughly, another 

 important principle should be brought into operation — namely, 

 that of attraction; and the only way by which advantage of this 

 principle can be obtained, is by having a large filtering surface, so 

 that the water may percolate very slowly through the filtering 

 material. 



If a stone, for instance, be suspended in muddy water, it will be 

 found very soon coated all round with the impurities in the water, 

 caused by the attraction wliich exists between the impurities and 

 the stone, the latter, as it were, forming a nucleous to which these 

 impurities adhere. On the same principle, if the water is allowed 

 to percolate very slowly through the material composing the bed 

 of filter, while the whole surface will still act like a sieve to prevent 

 the passage of many of the impurities, each particle of the material 

 is brought more fully into operation in removing the finer particles 

 of the impurities carried along with the water. The filtered water 

 is thus rendered more pure and pellucid; very fine material 

 becomes unnecessary, and the filter bed will continue for a much 

 longer time in good and efficient working order. 



AVhen the filter bed is too small, as the water must of course 

 pass quickly through, fine material becomes almost indispensable, to 

 prevent the quick percolation of the water from carrying many of 

 the impurities with it ; the consequence is, that it soon becomes 

 silted up, and requires continual cleansing; and previous to being 

 cleansed, recourse is not unfrequently had to that of forming holes 

 in the material, in order to make it more open; or, as I have 

 already observed, filter as much as they can, and make up the 

 deficiency by unfiltered water. It may, indeed, be taken as a 

 general principle, that the smaller the bed of filter, in proportion 

 to the amount of filtered water required, the finer must be the 

 material it contains, in order to remove tlie impurities; and, on 

 the contrai-y, the larger the bed the coarser the material. 



^V'ith regard to the speed with which the water should be allowed 

 to percolate through tlie filtering material, much will depend on 

 the state of the water previous to being let on to the bed of the 

 filter. In the most of filters at water-works the speed will be 

 found to range from 25 feet and upwards per day, and even that is 

 not regular in many cases; but in general, the' water should not 

 be allowed to percolate more than from 10 to 20 feet per day, 

 passed regularly through during the whole 21. hours. To many, 

 this speed being only from 5 to 10 inches per hour, may seem much 

 less than there is any need for; but, taking everything into con- 

 sideration, water companies, by making their filters so as to come 

 within this limit, they would be enabled to filter all the water sent 

 in for supply, and very materially diminish the yearly expense for 

 cleansing. 



Tabfe showing the number nf Cubic Feet and Im/ierial Galhns of Filtered 

 Water One Acre of Filterint/ Surface is capadlc of furnishing^ the Water 

 bcin'i allowed to percolate llirough the maleriat from 10 feet to 20 feet 

 per day. 



Cubic feet. 



Feet per 

 Doy. 



10 «.5,fiOO 



11 479,160 



12 522.720 



13 hCS.iSa 



14 60S.S40 



15 6.13,400 



Imp. Gul. 



2.722.r.00 

 2.Ua4,750 

 3 2l!7.0il0 

 3.53a,2.'>0 

 3,81 1. .WO 

 4.IW3.750 



Water can be filtered by passing it through the material in an 

 upward, downward, or horizontal direction; in springs it passes 



Feet per 

 D;.y. 



16 



17 



18 



la 



20.... 



Cubic feet. 



loip. gal. 



. ft«i,!«;0 

 . 7411..') 20 

 . 7s4.f!i0 

 , S27.C40 

 . S71,200 



4,3.'i6,000 

 462S.2.'iO 

 4,'JUl>,.'i00 

 5.l72,7,-i0 

 5,445,0110 



along in various directions, according as it finds a passage through 

 the different strata in its course. ^Vith artificial filters, however, 

 the same will not apply, as in the construction of these it is neces- 

 sary to take into account the process of cleansing. Upward filtra- 

 tion is, no doubt, the best, as the sediment, on account of its 

 weight, tends to fall downward while the water is flowing up- 

 ward; but in the cleansing of such filters there are many difficulties 

 to contend with, as since the sediment or silt lies mostly at the 

 bottom, the whole material requires to be taken out before the silt 

 can be either partially or wholly removed. Horizontal or oblique 

 filtration has similar objections. In practice the downward system 

 has been found to suit best, as the great body of the impurities lie 

 near the surface, and the bed of filter can l)e partially cleansed by 

 scraping, as is sometimes done, or more effectually by reversing 

 the direction of the water, as will be afterwards explained. 



The material through wliich to filter water should be of such a 

 nature as will remain unchanged, be imputrescent, capable of 

 allowing the water to pass through, and which does not change in 

 its mechanical structure— such as broken granite, trap rock, and 

 hard gritty freestone, silicious sand, pebbly and hard gravel, clean 

 ashes, &e.; that which is most generally employed is silicious sand 

 and gravel; broken trap rock and freestone being only used when 

 good coarse gravel cannot be got in sufficient quantities; they, 

 however, suit exceedingly well for a coarse filter bed, and in many 

 cases are even preferable to the gravel. Indeed, it has been 

 observed, that when water passes along a bed composed of rocks of 

 the trap, or amygdaloid species, a kind of natural filtration is 

 effected — so much so, that even moss water is rendered in some 

 cases comparatively pure. As to the fineness of the silicious sand, 

 none finer than that obtained on exposed parts of the coast should 

 be used, and it is even advisable to free that of much of its finer 

 particles. Sometimes only the finest of the sea sand is employed; 

 but when the filter is of large extent compared to the quantity of 

 %vater required to be furnished, a good bed of coarse silicious sand, 

 plentifully intermixed with pebbly gravel, makes a much better 

 filter. The use of fine sand is one of the great mistakes committed 

 in many filters at present in operation, as they always become 

 ineffective at times when most required, besides entailing a large 

 annual expense for cleansing. 



The next important matter connected with filtration is the best 

 construction of filters, so as to act effectually, and admit of being 

 easily cleansed. I have already observed, that the most practical 

 system of filtration is xvhen the water is allowed to flow downwards 

 through the material, as by this means the filter admits of being 

 much easier cleansed; and further, it may now be observed, that in 

 order to get advantage of the whole surface of filter bed, the filter 

 should be so arranged that the water, when let on, will spread 

 equally amongst the whole material of said bed. 



In many existing filters the water is let on by various openings, 

 and the filtered water is taken away at a level with the bottom, 

 either by drains or by a false bottom below filtering material. In 

 such filters the water which is let on, when the material is quite 

 clean, will be observed to spread only a small distance from the 

 inlet, and then disappear; but as the material becomes silted up, it 

 gradually spreads farther on to the bed of filter, and it is only 

 when the whole is silted up that the water spreads over the entire 

 surface. In such a filter the actual filtering surface becomes only 

 a part of the whole, and the coarser the material the less will that 

 part be; and, even although the whole surface be of large extent, 

 very slow filtration is not obtained; fine material, consequently, 

 becomes necessary, in order to cause the water to spread over a 

 larger part of the entire surface, as well as to prevent the impu- 

 rities from passing through. Were the water, by some means or 

 other, caused to fall on the entire surface in drops like rain, the 

 whole would be brought at once into operation, and by having a 

 large surface slow filtration would be obtained, and fine material 

 become unnecessarv. 



That of getting the water to fall equally over the whole surface 

 not being easily attained, especially in a large filter, the next best 

 system is that of raising the level at which the filtered water 

 escapes, thus causing the water, as it is let on, to be dammed back 

 amongst the filtering material, which consequently becomes fully 

 s.iturated, and is all brought more or less into operation : besides, 

 tlie pressure of water, by raising the outlet, being taken off, the 

 sand bed especially does not become so soon consolidated, as will 

 be observed to be the case when the outlet is at the lowest part of 

 the filter. \Vhere such a system has been adopted the water is 

 observed to rise in the bed of filter as the material becomes silted 

 up, forming a thin sheet over the whole surface. Dui-ing hot 

 summer weather this thin sheet of water has a tendency to become 



