164 BACTERIA IN WATER 



allowed to percolate slowly through a bed consisting of large 

 porous objects, such as broken bricks or large pieces of coke, and 

 here the percolation may be constant, no interval of rest being 

 given. The bacterial processes which take place in these two 

 methods are, however, probably closely similar. In the second, 

 the essential feature is a preliminary treatment of the 

 sewage in more or less closed tanks ("septic tanks"), where 

 the conditions are supposed to be largely anaerobic. This 

 method has been adopted at Exeter, Sutton, and Yeovil in this 

 country, and very fully worked at in America by the State 

 Board of Health of Massachusetts. In the explanation given 

 of the rationale of this process, sewage is looked on as exist- 

 ing in three stages. (1) First of all, fresh sewage the newly 

 mixed and very varied material as it enters the main sewers. 



(2) Secondly, stale seivage the ordinary contents of the main 

 sewers. Here there is abundant oxygen, and as the sewage flows 

 along there occurs by bacterial action a certain formation of 

 carbon dioxide and ammonia, which combine to form ammonium 

 carbonate. This is the sewage as it reaches the purification works. 

 Here a preliminary mechanical screening may be adopted, after 

 which it is run into an airtight tank the septic tank. 



(3) It remains there for from twenty-four to thirty-six hours, and 

 becomes a foul-smelling fluid the septic sewage. The chemical 

 changes which take place in the septic tank are of a most complex 

 nature. The sewage entering it contains little free oxygen, and 

 therefore the bacteria in the tank are probably largely anaerobic, 

 and the changes which they originate consist of the formation 

 of comparatively simple compounds of hydrogen with carbon, 

 sulphur, and phosphorus. As a result, there is a great reduction 

 in the amount of organic nitrogen, of albuminoid ammonia, and 

 of carbonaceous matter. The last is important, as the clogging 

 of ordinary filter-beds is largely due to the accumulation of such 

 material, and of matters generally consisting of cellulose. One 

 further important effect is that the size of the particles of the 

 deposited matter is decreased, and therefore it is more easily broken 

 up in the next stage of the process. This consists of running the 

 effluent from the septic tank on to filter-beds, preferably of coke, 

 where a further purification process takes place. By this method 

 there is first an anaerobic treatment, succeeded by an aerobic ; 

 in the latter the process of nitrification occurs by means of the 

 special bacteria concerned. The results are of a satisfactory 

 nature, there being often a marked diminution in the number of 

 coli organisms present. 



In the earlier stages of any sewage purification, there is little 



