ACTION AND UTILITY OF FILTERS. 271 



sterile filtrates were obtained from both filters, although inocu- 

 lations from the outer surface of the candles showed that living 

 pathogenic germs still remained. The addition of a little broth, 

 one-fiftieth of the bulk of the water, caused the pathogenic 

 organisms to appear in the filtrate. Typhoid and cholera bacilli 

 were then added to Thames water, taken near Waterloo bridge 

 at low tide, and sterilised before use. Negative results were 

 obtained with both filters, showing that these pathogenic 

 organisms, even when present in a highly polluted water, do not 

 pass through into the filtrate. "As a final and conclusive test 

 as to the possibility of disease germs traversing the walls of 

 these filters when fed with a water highly contaminated with 

 organic matter, an experiment was carried out with undiluted 

 sewage." A cholera culture was added to sterilised London 

 sewage, which was then filtered through a sterilised candle ; 

 spirilla could not be found in the filtrate although the experi- 

 ment was continued for eighteen days. A mixture of cholera 

 vibrios and sterile sewage was also passed through a sterilised 

 candle coated with slime resulting from previous filtration ; the 

 results in this experiment were negative. Lastly, a culture of 

 cholera was added to unsterilised sewage and then filtered 

 through an unsterilised candle coated with slime ; spirilla could 

 not be found in the filtrate. Experiments on similar lines were 

 also made with B. tvphosus, but these again gave an absolutely 

 negative result. All these results support the authors' contention 

 that waters naturally highly contaminated, and even sewage 

 itself, do not appear to supply pathogenic organisms with a 

 sufficiently favourable pabulum to enable them to traverse the 

 walls of Pasteur-Chamberland and Berkefeld candles. 



In 1897 Lunt made an investigation of the capabilities of the 

 Berkefeld filter, in order to obtain an independent confirmation 

 of the results obtained by Schofer and others, as to the power of 

 this filter to intercept pathogenic bacteria especially typhoid- 

 even when water bacteria are able to pass through. Lunfs- 

 experiments appeared to justify the following conclusions : 



(1) When the Berkefeld filters are efficiently sterilised and used 

 in an approved manner, they give on the first day of using an 

 absolutely sterile filtrate. This is the case with all the forms of 

 filtering candles examined, not only with natural water but 



