THE DISTRIBUTION OF MICRO-ORGANISMS 187 



upon the exhaustion of the food supply by the growth of bacteria 

 themselves, upon the destructive influence of direct sunlight, 

 and the dilution of the contamination by a large volume of 

 water. ^ It is not usually to be relied upon as a means of freeing 

 the water-supply from pathogenic bacteria. 



Storage of Water. — When water is kept in large reservoirs, 

 the solid particles in it, including bacteria, tend to fall to the 

 bottom. The number of bacteria in a water-supply may be 

 considerably reduced in this way. The use of large storage 

 reservoirs also provides for the dilution of any sudden excess of 

 pollution, and if the water is held in storage the pathogenic 

 germs present disappear for the most part in a few days or weeks. 



Filtration. — Water may be completely sterilized by passing 

 it through the Pasteur-Chamberland filters of unglazed porcelain, 

 or through the more porous Berkefeld filters. Such filters are 

 effective only when frequently cleaned and baked, and in practical 

 purification of water for household purposes they usually fail 

 because of the intelhgent care they require. Other types of 

 domestic filters are generally worse than useless. 



Filtration on a large scale is commonly employed in the purifi- 

 cation of water supplies of cities. By this method 98 per cent to 

 99 per cent of the bacteria in water may be removed. 



Slow Sand Filtration.''' — The filter consists of successive layers 

 of stones, coarse and fine gravel. The uppermost layers are 

 of fine sand. The whole filter is from i to 2 meters thick. The 

 sand should be 60 cm. to 1.2 meters in thickness. The accumu- 

 lated deposit from the water and a little of the fine sand must 

 be removed from time to time, but the layer of fine sand must 

 never be allowed to become less than 30 cm. in thickness. The 

 first water coming from the filter is discarded. The actual fil- 

 tration is done largely by the slimy sediment which collects 

 on the surface of the layer of fine sand. The filterbeds may 



' See Jordan, Journal of Experimental Medicine. Vol. V, p. 271. 

 2 For a fiill discussion see Journal American Medical Association. Oct. 3 to 

 31, 1903- 



