ACTION AND UTILITY OF FILTERS. 26<) 



candle. The direct passage points to some imperfection in the 

 filter, but the gradual growth through the material is common 

 to all filters, and does not indicate danger of infective disease. 

 He considers that the best way to test a filter is to pass through 

 it some easily recognisable special organism, and if this cannot 

 be detected in the filtrate, the filter may be considered to give 

 security against the passage of pathogenic organisms. 



Schofer's experiments, which were made in 1893, strongly 

 supported Gruber^s contention that typhoid bacilli, when 

 inoculated into pure and impure natural waters, will not grow 

 through the walls of a filtering candle owing to a 

 deficient supply of nutriment. Schofer employed a Berkefeld 

 bougie placed in a glass cylinder. Sterile distilled water and 

 sterile well water were filtered through the bougie for sixteen 

 and twelve days respectively without any sign of the B. typhosus 

 appearing in the filtrate, although a culture of this organism 

 was inoculated into the contents of the glass cylinder, not onlv 

 on the first day but about every second day throughout the 

 experiment. A considerable quantity of water was passed 

 through the filter, and the temperature during the experiment 

 varied between 19 and 26 C. A similar experiment was made 

 with the spirillum of cholera, but no traces of this vibrio 

 could be detected in the filtrate. Schofer then tried the effect of 

 filtering a polluted sterile well water inoculated with typhoid 

 bacilli. The well water contained in one litre : 1205 mgms. of 

 total solids, 38 mgms. of organic material, marked ammonia, 

 traces of nitrous acid, 251 mgrns. of nitric acid, 120 mgms. of 

 chlorine, 158 mgms. of sulphuric acid, 245 mgms. of calcium 

 oxide, and 28 mgms. of magnesium oxide. The experiment 

 was continued for twenty-four days ; during this time a culture 

 of B. typhosus was added twelve times to the contents of the 

 glass cylinder, and the temperature was maintained between 

 18 and 25'2 C. The B. typhosus, however, was never detected 

 in the filtrate. Other experiments were made with a canal 

 water, which were of special interest, as they appeared to show 

 that the only cause of the non-appearance of the B. typhosus 

 in the filtrate was a deficiency of nutrient material. The 

 canal water contained in one litre : 215 mgms. of solids, 47 

 mgms. of suspended material, 8 mgms. of chlorine, and 2 mgms. 



