280 BACTERIOLOGICAL EXAMINATION OF WATER. 



which cannot be avoided if a fair delivery is to be obtained, 

 appears to militate against the immobilising and devitalising 

 influences which operate so strongly in filters made with very 

 narrow lacunar spaces. It is well known that the typhoid 

 bacillus can grow out into long thread-like forms, and if 

 this takes place in the lacunar spaces of the Berkefeld candle 

 the bacillus may find its way into the cavity in the interior 

 of the candle. It is also probable that, owing to deficient 

 immobilisation, the bacilli are gradually washed through the 

 filtering candles. This often takes place when filters are worked 

 with a pump bearing directly on the bougie ; if, however, an air 

 cavity is introduced between the pump and the filtering candle, 

 the pressure is equalised, and this source of danger is removed. 



(3) When a highly polluted liquid containing typhoid bacilli 

 is filtered through a Berkefeld candle the bacilli may appear in 

 the filtrate in four days. Consequently in order to obtain 

 protection from water-borne disease, when these filters are 

 employed, it is necessary to sterilise the candles every third day. 

 This cannot be done by simply brushing the candles, they must 

 be sterilised by boiling water or by exposure to saturated steam. 



METHODS OF TESTING WATER FILTERS. 



It is clear from what has been said, that when examining 

 filters as to their powers of removing bacteria from water, it is 

 necessary to distinguish between the direct passage of micro- 

 organisms through the filter, due to some imperfection in the 

 filtering material itself or in its connection with the receptacle for 

 filtered water, and the indirect passage caused by the growth of 

 the micro-organisms through the walls of the filtering material. 

 The direct passage of bacteria may be determined in the case of 

 non-pressure filters by placing a rich emulsion of some special 

 organism, such as the B. prodigiosus, in the water to be filtered 

 and examining the filtrate for this organism by means of plate 

 cultures. If the special organism be found in the filtrate, this 

 may be caused by some flaw in the filtering material or defects 

 in the fitting of the material to the delivery pipe. If the 

 filtering material be in the form of a bougie it should be placed 

 in a vessel containing an emulsion of the special organism and 

 then be connected to a sterile partially exhausted flask. It is 



