678 PROCEEDINGS OP SECTION I. 



of bluestone shivers, and provided with suitable under-drains. The 

 filling and emptying of filters is automatically controlled by means 

 of the Cameron-Martin alternating gear of the latest type. The iwmber 

 of persons served by these works is about 1,500, and the average daily 

 flow is approximately 35,000galls. to 40,000galls, The system is 

 designed to receive domestic sewage only ; but, as at Chatswood, it 

 is found that the flow is afiected by rains. At the silt pit there is an 

 overflow, by which any quantity may be diverted from the tanks, the 

 amount being regulated by the height of a stopboard placed across 

 the outlet. No screening of any kind is attempted. 



Up to date two tanks only have been used, and these are used 

 one at a time only. The scum on the one at present in use is 19in. 

 thick at the inlet, thinning off gradually to 2in. at the outlet. On the 

 tank now resting the scum has gradually disappeared until a film of 

 about ^in. only remains. 



The results of analyses for the last 12 months are given in Table X., 

 and it will again be observed that the sewage is of essentially the same 

 character as is found in the parts already mentioned. On observing 

 the changes effected by the discharge of sewage through the installation, 

 it will be seen that there is practically no diminution, and in fact during 

 some months an actual increase, in the oxygen absorbed in four hours 

 in the tank effluent as compared with the raw sewage. The explana- 

 tion of this phenomenon in all probability lies in the fact that the 

 sewage takes approximately 48 hours to pass through the tank, and 

 doubtless during that time the process of disintegration is carried on 

 to a point where the organic carbon exists in combinations more sensi- 

 tive to the oxidising effect of the acid permanganate. It certainly 

 cannot be considered that there is any accession of organic carbon 

 compounds to the sewage during its discharge through the tanks. 

 The test indicates in reality molecular changes. There is also a loss 

 of organic nitrogen to the same extent as observed at Chatswood in 

 the final effluent. Whilst on the one hand the organic nitrogen figures 

 are somewhat higher than at Chatswood, on the other hand those for 

 oxidised nitrogen are higher also ; and in this connection it is interest- 

 ing to observe the effect of heavy rain upon the nitrifying processes 

 in the filters, as exemplified in the analyses for June and subsequent 

 months. Up to June the nitrates were abundant in the final effluent. 

 During that month heavy rain fell and nitrates correspondingly 

 diminished. Further, during this heavy rain, the stormwater overflow 

 mentioned above was not put into operation, so that the whole of the 

 flow passed through the tanks and filters. The result of this was a 

 complete cessation of nitrification in July. In the next month, namely 

 August, the nitrifying organisms must have recovered, at all events 

 partially, as evidenced by the very high figures for nitrites. During 

 the next two months the nitrifying process appears to be slowly reaching 

 its normal level. The percentage of reduction in organic matter is 

 set out in Table XL, which can, with advantage, be compared to 

 similar tables for North Sydney and Chatswood. 



