82 



KNOWLEDGE. 



[April 1, 1896. 



can be disconnected without interfering with one another, 

 and consist of basins about two acres in extent. The 

 material used for filtering consists of about three feet of 

 fine sand, resting on eight inches of coarse sand, below 

 which is an eight-inch layer of gravel and an eight-inch 

 layer of small stones. The water is carefully pumped on 

 to each of these filters in turn, special precautions being 

 taken not to disturb the surface of the beds, and the water is 

 farther not allowed to rise to a higher level than three 

 feet above the surface of the bed. In this way the rate of 

 filtration is maintained at a uniform rate of about one 

 hundred cubic metres per hour. The system is under careful 

 bacteriological control, samples of water being frequently 

 taken for examination, and as a rule the number of 

 bacteria found in the filtered water do not exceed from 

 ten to forty per cubic centimetre. When the number rises 

 to one hundred the filter-bed is closed and remade, so that it 

 will be seen that the only test as to the efliciency of the 

 filter-beds is bacteriological. In Warsaw the town is 

 supplied with water from the river Vistula, and elaborate 

 waterworks have recently been completed for the filtering 

 of this supply. The system adopted is similar to that 

 which obtains in Hamburg, but the settling tanks are 

 smaller, while both these and the filtering-beds are 

 enclosed, so that the whole of the works are under cover. 

 In this way the influence of temperature, light and dust 

 are minimized, and the water supplied into the city 

 mains is of very uniform quality. At present it is difficult 

 to say how a few feet of sand can possibly have such 

 a beneficial eSect upon the bacterial contents of a water. 

 The surface of a sand filter, after it has been in use 

 for some time, becomes covered with a film of glutinous 

 silt or slime, and the action of a filter in removing organisms 

 seems to be due in a large measure to the surface action of 

 this film. Under the microscope it is found to be composed 

 of a mass of bacteria in the zooglcea condition, and amongst 

 the organisms present are those which have nitrifying 

 properties. It seems, therefore, probable that the surface 

 and first few inches of a sand filter contain innumerable 

 bacteria which, in presence of air, oxidize the organic 

 matter contained in the water. In intermittent filters the 

 necessary oxygen is derived from the air, but in many 

 waters there is sufticient dissolved oxygen to render any 

 special aeration of the filter-bed unnecessary. This 

 biological action destroys the food supply of the water 

 bacteria, and thus ensures their speedy death, or possibly 

 the organisms growing on the filter excrete ptomaines 

 which have a specific poisonous effect upon them. Whatever 

 the explanation may be, the action is most effective when 

 the surface is well covered with this colloidal growth, since 

 when it is scraped away the filter does not remove the 

 bacteria nearly so well. Prof. Eay Lankester compared 

 the action to dialysis through a fine jelly, and hence it is 

 important that there should be no cracks or breaks in the 

 continuity of the straining surface. The filter has to be 

 cleaned from time to time owing to the growth of the 

 surface layer downwards, which so clogs the filter as to 

 render the filtration too slow for practical purposes. The 

 renewal of the filter from time to time seems also desirable 

 on other grounds, since it is quite possible that with an 

 undisturbed filter the organisms might grow right through 

 the bed, and thus infect the filtered water. In London, the 

 water companies have used sand as a filtering medium for 

 many years, and the recent inquiry into the water supply of 

 London caused considerable attention to be devoted to the 

 subject. The filter-beds in London are not constiucted with 

 the same care as those at Hamburg and Lawrence. For 

 many years the number of bacteria present in the water 

 supplied by the several companies have been counted. Thus 



Dr. Percy Frankland showed that, taking a period of three 

 years, 97' 5 of the micro-organisms present in the Thames 

 water were removed by the companies, and that the East 

 London Water Company, which uses Lea water, succeeded 

 in removing 95'7 per cent. Although the chances of an 

 epidemic being spread by such a water are far less than if 

 no filtration were adopted, it is still possible that among 

 the few organisms which succeed in running the gauntlet 

 of the filters there will be some that are pathogenic in 

 character, and therefore a second filtration at the house of 

 the consumer through a Pasteur-Chamberland filter is 

 desirable. In Germany, the city authorities of Worms 

 have constructed some large cylinders of artificial stone 

 through which the water is forced under pressure, and in 

 this way hope to produce a sterile water for domestic use, 

 and more recently it has been seriously proposed to subject 

 the water supply of towns to a high temperature for a 

 short time under pressure, with a view to the absolute 

 destruction of all the micro-organisms present. 



The use of river water for a town supply has been 

 obviated in several towns, such as Fraukfort-on-Main, 

 Buda-Pest, and Dresden, by making use of natural filtration. 

 When a shallow well is used for a water supply, there is 

 always a danger of surface water which has not undergone 

 any process of filtration finding its way into the well 

 through openings in the sides. When an iron pipe is em- 

 ployed, this danger is to a great extent removed, and the 

 water which collects at the bottom of the well will have 

 passed through several feet of subsoil and be practically 

 filtered from any bacteria existing in the water. Such subsoil 

 water can be collected on a large scale and utilized for the 

 supply of towns, although there are several factors to be 

 taken into consideration when it is proposed to take a large 

 quantity of water from beneath the surface in this way. 



At Dresden there is a collecting gallery about five 

 hundred and seventy feet long, beneath the surface of the 

 ground and some little distance from the river. It ia 

 doubtful, however, in this case whether the subsoil is 

 sufficiently impervious to prevent leakage of the river 

 water directly into the collecting gallery. At Frankfort 

 a long tunnel of white bricks extends for nearly a mile 

 and a half below the surface of the ground at a depth 

 of about fifty feet, at which depth the subsoil water is 

 met with. Copper pipes are sunk from this subway at 

 intervals into the subsoil water, and from the channel it is 

 pumped direct into the mains of the town. This water is 

 of uniform temperature all the year round, and is free 

 from organic matter and bacteria. The surface from 

 which the water drains above the works is a large forest 

 of about three square miles in area, and this is kept free 

 from undergrowth and from habitation. 



In London one of the companies derives its supply from 

 deep wells sunk in the chalk, and this water, as might be 

 expected, requires no filtration and is remarkably free 

 from bacteria. 



Such underground supplies are to be desired in localitiea 

 in which there are severe frosts in winter, as sand filters 

 are useless in times of frost. In Berlin, to prevent the 

 freezing of the filters, they are covered, but at present it is 

 doubtful whether the action of light upon the surface 

 growth of filter-beds is not advantageous. 



Other methods of purification are in use in diSerent 

 towns, involving a combined chemical treatment with 

 filtration. Thus at Antwerp Anderson's system is adopted, 

 in which the water is treated in revolving cyUnders 

 containing iron, and in the United States the Hyatt 

 process is in use in many towns. It will be beyond the 

 scope of the present article to discuss these chemical 

 methods for improving the bacterial quality of a water at 



