Apbil 1, 1895.] 



KNOWLEDGE 



81 



State Board of Health, in his last report, stating that one 

 of the chief objects of water filtration, in most cases, is the 

 removal of the disease-producing germs. Formerly the 

 efficiency of a filter might be accurately measured by 

 testing it with some dark-coloured, finely divided material, 

 such as gunpowder, charcoal, or ultramarine, and ascer- 

 taining whether these particles were kept back by the pores 

 of the filter. Many of the domestic filters still commonly 

 employed, when tested in this way, yield a filtrate which is 

 clear and bright, proving that the pores in the filtering 

 medium are sufficiently tine to strain back such particles. 

 It is obvious that these filters will ensare the removal of ; 

 all suspended matter which is of the same size, or larger, ' 

 than such particles, and that if a transparent bright liquid 

 only is aimed at, any filters passing such a test might be 

 recommended. Unfortunately, the pathogenic or disease- 

 producing bacteria are much smaller than such particles, 

 and find their way through the pores of most filters. 



If, therefore, a domestic filter is to fulfil the condition 

 laid down by Dr. Drown, it must be so constructed as to 

 ensure that these minute organisms cannot possibly pass 

 through, so that on testing the water, before and after 

 filtration, the number of micro-organisms must be sensibly 

 diminished. An ideal filter is one which will allow no 

 organisms to pass through, but, as shown by Drs. Sims, 

 Woodhead and Wood, there are practically none in exist- 

 ence which fulfil this condition except the candle filters, 

 and these can only be used when a pressure of water is 

 available. The partial removal of the micro-organisms 

 is, however, a matter of less difficulty, and filtration 

 through sand is found to be very efficient in improving the 

 bacterial condition of water. In America, during the last 

 few years, experiments on a large scale have been 

 conducted, with the result that it has been shown to be 

 possible to construct filters which will yield two million 

 gallons of water per acre daily, and remove 99'5 per cent, 

 of the bacteria in the applied water. 



In these researches a particular bacterium (B. prodigiosus) 

 was used as a test organism for ascertainmg how filters 

 made in different ways, witb sand of varying dimensions, 

 behaved. This organism was cbosen because it is very 

 easUy distinguished from ordinary water bacteria, and in 

 its mode of life in river water resembles very closely the 

 behaviour of the organism which is associated witb typhoid 

 fever. It was supposed that if the sand removed this 

 organism from the water, a filter constructed in the same 

 way might be recommended for the filtration of the water 

 supplied to a town, and thus ensure that such filtered 

 water would be free from the germs of typhoid fever. At 

 first it was thought that the sand used for such filters 

 must be very fine, and that the rate of filtration should be 

 slow ; but last year a high degree of bacterial efficiency 

 was attained by new filters of coarse sand, and operated at 

 high rates of filtration. Thus, in two experimental filters, 

 coarse sand having an effective size of 0-26 to 0-29 milli- 

 metres when used to a depth of sixty inches, removed over 

 98 per cent, of the water bacteria, and over 99 5 per cent. 

 of the applied B. prodirjiosus, even when the rate of filtra- 

 tion was upwards of nine million gallons per acre. The 

 knowledge gained by experiments ^f this character has 

 been made use of on a large scale for filtering the water 

 supply of the city of Lawrence, in Massachusetts, with 

 most satisfactory results. This city is supplied with river 

 water which is contaminated with drainage from other 

 riparian towns, and for many years the death-rate from 

 typhoid fever was three times higher than that of other 

 towns of the same size. It was also noticed tbat the 

 annual epidemic was later at Lawrence than at Lowell, a 

 town higher up the same river, so that it was clearly 



evident that the disease was water borne, and caused from 

 the inhabitants drinking the unpurified river water. In 

 1892 the cholera scare iadacei the autborides to be»ia 

 the construction of this filter, which was finally completed 

 towards the latter end of 1893, and now tbat it has been 

 at work for over a year, the effect of the filtration of tbe 

 water can be estimated. Chemically, the cban;i;e has con- 

 sisted in the conversion of the free ammonia and organic 

 ammonia into nitrates, whilst the hardness of the water 

 has been slightly increased. Tae colour, as might be 

 expected, has been much improved, but the most important 

 change has been in the removal of the micro-organisms 

 present in the water. As a result of working this filter, it 

 has been shown that, when properly attended to, tbe filter 

 increases in efficiency after some days. The removal of 

 the bacteria is, therefore, not only due to a straining 

 process, but also to the conditions for the life of the 

 different organisms being made unsuitable. It is well 

 known that the bacteria living in water require organic 

 matter as their food, and as the filter allows the oxygen of 

 the air to oxidize the organic matter in the water, it 

 follows that those organisms which are successful in 

 passing through the filter find themselves in a medium 

 which contains no food supply, and they, therefore, quickly 

 perish. The total cost of the filter was about £13 000, and 

 although the results above recorded are of considerable 

 interest, this expenditure of money has already been justified 

 by the marked improvement in the health of tbe town. 

 Thus, since the filter has been in use, the death rate of the 

 town from typhoid fever has been reduced to forty per 

 cent, of the former mortality, and it has also been shown 

 that of this remaining forty per cent, nearly one half of 

 the cases are attributable to the use of unfiltered water 

 drawn from the canals. At present the returns from the 

 medical officers for the autumn months of last year are not 

 available, so that the effect of the filter in preventing the 

 communication of diarrhoeal and other diseases carried by 

 water cannot be given, but the physicians report that 

 there has been a very marked improvement in the health 

 of the population since the filter came into use. These 

 experiments clearly prove tbe fact that no reliance must 

 be placed on the self-purifying power of streams, as the 

 previous his:ory of tbe city of Lawrence test:ties to the 

 ready conveyance of typhoid fever down a stream, by 

 sewage-polluted drinking water. Tue results are also 

 so satisfactory that there can be no doubt that the 

 construction of this filter will be speedily copied by other 

 towns similarly situated with regard to tbeir water supplies, 

 as the practicibility of protecting a community against an 

 infected drinking water supply by natural sand filcraiion 

 has been established. 



Although the Lawrence experiments are by themsi I tes 

 sufficiently conclusive, there is an abundance of evidence to 

 show that the filtration of water, if properly carried o'lt, 

 has a very material influence upon its quality. During the 

 cholera outbreak in Hamburg in 1892, it was manifest 

 that the filtration of the river water, as supplied to Altona, 

 practically rendered that part of the district immune from 

 attacks of cholera, and Prof. Koch obtained conclusive 

 evidence to show that drinking unfiltered Elbe water was 

 the cause of the tremendous mortality from this outbreak. 



Since then, Hamburg has had new filtration works 

 ei'ected, and these are now at work. The essential 

 features of the Hamburg filters are somewhat different 

 from tbose of the one at Lawrence, as the water is 

 first allowed to settle in four sedimentary basins, each 

 holding enough for one day's supply, and having an area 

 of about twenty acres, before being pumped on to the 

 filter-beds. These filter-beds are so arranged that they 



