PURIFICATION OF WATER FOR DRINKING PURPOSES 153 



cholera, whilst that belonging to Altona remained free. 

 The remarkable fact was brought to light that, in those 

 houses supplied with the Hamburg water, cholera was 

 rampant, whilst in those on the Altona side and furnished 

 with the Altona water not one case occurred. 



The Hamburg water, however, to start with was, 

 as we have seen, relatively pure when compared with 

 the foul liquid abstracted from the Elbe by Altona ; but, 

 whereas in the one case the water was submitted to 

 exhaustive and careful filtration through sand before 

 delivery, in Hamburg the Elbe water was distributed 

 in its raw condition as taken from the river. But 

 further testimony was afforded later to the truth of this 

 interpretation ; for during the ensuing winter, when the 

 cases of cholera had almost completely died out in Ham- 

 burg, suddenly a most unexpected and unaccountable 

 recrudescence of the epidemic occurred, and this time in 

 Altona. This outbreak could not be traced to any direct 

 infection from Hamburg, but must have arisen in Altona 

 itself. In all, about forty-seven cases were recorded 

 between December 23, 1892, and February 12, 1893. 

 A searching inquiry was instituted, and it was ascer- 

 tained that the number of bacteria found in the filtered 

 water, usually about 50, had during these months 

 risen to as many as 1,000 and more in 1 c.c, clearly 

 indicating that the filtration of the water was not being 

 efficiently carried out. That this was actually the case 

 was proved by the fact that one of the sand filters, 

 which had been cleaned during the frost, had become 

 frozen over, and was in consequence not able to retain 

 the bacteria. That the outbreak did not become more 

 serious Koch ascribes to the fact that this imperfectly 

 filtered water was so largely diluted by that which 

 had been efficiently filtered. 



There cannot be any longer a doubt, therefore, as 



