1906.] Nitrification of Sewage. 69 



the superficial layers. A series of samples of air aspirated from different 

 depths, by means of iron tubes driven vertically into the body of the filter 

 while in continuous use, yielded the following mean results : 



Carbonic Acid in parts per 1000. 

 1 ft, 2 ft. 3 ft. 4 ft. 



19-5 21-5 20-0 20-0 



An interesting fact bearing upon the aeration of the filter was incidentally 

 made apparent by the method at first adopted of collecting the air. The 

 distributor passing backwards and forwards over the filter was found to 

 interfere considerably with the collection of the samples, and, to obviate this, 

 the apparatus was periodically stopped for varying periods during the 

 operation. Owing, however, to extraordinary discrepancies in the results 

 thus obtained, measures were taken to overcome this difficulty and enable the 

 air to be collected without interrupting the regular flow of sewage on to the 

 filter. The results, of which the figures just given are the means, were then 

 found to be remarkably uniform, whereas in the case of the earlier samples 

 the carbonic acid varied in amount according to the intervals which elapsed 

 between the stopping of the distributor and the collection, short though these 

 intervals were, from 2*8 to 26'1, showing how free was the current of air 

 through the filter. 



If, then, by using fine-grade particles the depth of filter may be greatly 

 reduced, the resulting economy would dictate such a course, but there is another 

 important consideration which, other things being equal, tells in favour of 

 shallow filters from the point of view of aeration. The air travels through a 

 filter from above downwards, the direction of the current probably being 

 mainly due to the percolation downwards of the sewage, and its more rapid 

 flow along the effluent drains. The air, therefore, as it passes downwards, 

 carries with it the products of the combustion which has taken place above, 

 and thus has an asphyxiating effect upon the organisms below, and it is 

 possible that even anaerobic fermentation may be revived in the bottom 

 layers. The sudden increase in the free ammonia figure, noted in the case of 

 the effluent from the lowest tray, may possibly be accounted for in this way, 

 because the albuminoid ammonia figure does not represent the total organic 

 nitrogen present, therefore there is an unrecorded margin of nitrogenous 

 organic matter available for the revival of the ammonia change should this 

 explanation of the phenomenon be the correct one. 



Be this as it may, however, the phenomenon does not appear to be 

 accounted for by fouling of the deeper strata. At the end of the 

 observations the filter was opened and carefully examined throughout its 



