26o 



SEWAGE AND ITS PURIFICATION 



then the determination of the volume and character of the 

 gases evolved would give an indication of the nature and energy 

 of the bacterial action. 



In the first Massachusetts experiments (Report State Board, 

 1899), the average of monthly analyses gave in parts per 

 100,000 : — 



This is in close accord with my results at Exeter in 1896 

 and later. At first the tank effluent, according to the usual 

 American practice, was sent over a sand filter, with, not un- 

 naturally, a bad effect, on account of the closeness and lack of 

 aeration ; afterwards a Dibdin coke bed with intermittent 

 filling was used with success. The report is in favour of '' a 

 much smaller septic tank than has been proposed abroad, 

 because the sewage has travelled far and has lost its oxygen," 

 and already undergone much of the anaerobic change. The 

 State Board of Health of Minnesota has prohibited the dis- 

 charge of sewage unless '* first passed through a septic tank or 

 filter-bed or both tank and filter-bed." 



The organic matter is now in a readily oxidizable state, and 

 passes on to the second or aerobic stage, in which it is dealt 

 with by the filters. A large amount of carbonic acid is pro- 

 duced in the filters by oxidation of the organic matter, and is 

 partly driven out in the stages of filling. Much of the combined 

 nitrogen now changes into nitrates ; the average at Exeter was 

 about one per 100,000 of nitric N. After a period of long rest, 

 the greatest amount of nitric N makes it appearance in the first 

 runnings, showing that it is during the period of rest that the 

 nitric acid is formed. I found in these cases 3 to 5 parts of 

 nitric N.^ 



^ J. R. San. Inst., xviii., i., p. 73. 



