E. J. Russell and E. H. Richards 527 



No account is taken here of the sohd matter deposited on the filter, 

 but over these long periods it probably has no great disturbing effect 

 on the figures. 



Variations in oxygen supply caused variations in the loss of nitrogen 

 but usually in an opposite direction : thus when the oxygen supply 

 was increased by diluting the sewage with an equal volume of tap- water 

 the loss was reduced to about 20 %, while breaks in the continuity of 

 aeration, such as are obtained when the liquid was run on to the contact 

 beds, caused the loss to increase to 50 %^. 



These experiments do not tell us what became of the nitrogen. But 

 in an earlier experiment by Letts^ it was shown that the amount of 

 gaseous nitrogen dissolved in a contact bed effluent was greater than 

 in the original sewage, the increase amounting to 16-20 % of the missing 

 nitrogen. There was therefore an actual production of gaseous nitrogen 

 during the process. The figures do not balance, but this could hardly 

 be expected: it is unlikely that the whole of the gaseous nitrogen 

 would remain dissolved iii the effluent, and in any case part of the loss 

 of nitrogen from the sewage must be attributed to the hosts of insects, 

 worms, etc., which breed in it, and then move off elsewhere. 



^ Thus the losses on the addition of tap-water were : 



Diluted sewage (equal volume of tap-water added) 



The figures are all shown as per 100,000 of sewage but as the volume of sewage did not 

 change they are proportional to the actual weights involved. 



Some of the missmg nitrogen in this case is to be found in the sludge which deposits 

 each time the contact bed is filled. 



2 Report to the Corporation of Belfast on the purification of the Belfast Sewage 

 (Baird, Belfast, 1908). Also, 5th Report of the Sewage Commission, Appendix \^. pp. 

 171-94. , - 



