PROCEEDINGS OF SECTION I. 671 



Once a month samples of the raw sewage, tank and final effluents, 

 are collected and submitted for analysis. The samples are gathered 

 over a period of 48 hours in order to ensure fair average specimens of 

 the daily flow. In the table on the preceding page I have set out the 

 results of these analyses. 



The figures show a marked uniformity from month to month, 

 which indicates that the system of sampling adopted can be depended 

 upon to give, if not absolute, at all events comparable results. It 

 will be observed that after the sewage has passed through the tanks 

 there is a slight increase in the free ammonia and a decrease of about 

 •one-quarter in the figures for organic matter. Oxidized nitrogen is 

 entirely absent. After treatment in the filter beds nearly nine-tenths 

 of the organic matter disappears, free ammonia diminishes by one- 

 Half, and a fair amount of oxidized nitrogen is found. The final effluents 

 were, as a rule, clear (rarely slightly turbid), bright, and free from any 

 smell, and withstood the incubator test (see post.). 



It will be seen from the table that in January the final effluent 

 contained organic matter much above the average. Up to this time 

 five or more tanks were being worked. This number was then reduced 

 to four, and coincidently the organic matter showed a steady decline 

 from month to month until June, when one tank was cut out. An 

 immediate rise was observed lasting for the next two months, and 

 followed by a fall to a point below any reached previously during the 

 series. Now, for the last four months three tanks only were employed, 

 and the final effluents obtained of a high grade. We may conclude 

 from this that, in dealing with this sewage, as high-class effluents can be 

 secured after a short passage through the tanks as after a more pro- 

 longed sojourn. 



In order to present the matter in a manner that does not involve 

 •columns of decimals, I have constructed the follo\ving table, which 

 gives the percentage reduction of the figures for albuminoid ammonia 

 -and oxygen absorbed from permanganate in four hours in both the 

 tank and final effluents, and also the sum of these which is the total 

 purification effected by the installation. I have further given in a 

 separate column the percentage reduction of total organic matter. 

 This I have calculated by expressing the difference between the sum 

 of the albuminoid ammonia and oxygen absorbed figures in the final 

 effluent, and the sum of the same constituents in the raw sewage as a 

 percentage of the latter. This is, of course, empirical, but it has a 

 practical application. 



On an average roughly about one-third of the total reduction 

 occur? ;r- +he tanks and the remainder in the filters. In the tank the 

 loss of organic nitrogen is slightly in excess of that observed for organic 

 carbon. It may further be observed in this connection that even 

 though the loss of organic matter in the tank, as indicated by the 

 ■chemical tests applied, be small (cf. records for November, June Sep- 

 tember, and October supra), still the tank effluent shows that it is in a 

 condition suitable for the second stage of purification by its response 

 to treatment by the filters. 



