674 PROCEEDINGS OF SECTION I. 



of flow. The periods, of course, do not coincide in the different beds. 

 Mr. Cook has represented the working of the filters diagrammatically, 

 and has kindly permitted me to make use of a copy of his drawing, 

 which is attached. The diagram is very clear, and conveys more than 

 could be told in columns of writing. 



The scheme was designed to treat sewage on the separate system, 

 and as far as possible stormwater is prevented from entering the reticu- 

 lation ; but despite precautions it is found that during heavy rain a 

 considerable volume of excess water reaches the outflow, and that after 

 rain the flow is still maintained above the average for some time owing 

 to the entrance of subsoil water into the sewers. On account of the 

 large tank space at present available, several times the normal flow can 

 be run through the tanks ; but in very excessive storms the flume 

 leading to the main carrier overflows, and the many-times-diluted 

 sewage flows past the tanks and filters into the watercourse, where a 

 further dilution occurs by the water which naturally collects in this 

 part. No screening of any kind is carried out. About one-fifth of 

 the area of each tank is covered with floating scum of an average depth 

 of from Tin. to 8in., the remaining four-flfths being covered with a 

 close mass of water hyacinths. 



In February, 1896, that is about four years after the works were 

 first placed in operation, it was found that silt had accumulated in the 

 bottom of the tanks to an average depth of lOin., and up to this time 

 no attempt had been made to remove it. In addition to the silt the 

 tanks contained a considerable amount of fine suspended matter, 

 which might be considered as the accumulated residue of organic matter 

 incapable of undergoing further resolution in the tanks. This matter 

 was carried by the natural flow into the filter beds and appeared in the 

 final effluents, giving rise to slight turbidity. For this reason it was 

 considered advisable to attempt to remove this residue, and the work 

 was carried out during heavy rains by permitting the flow to scour 

 through the tanks and carry out as much as possible. The operation 

 was repeated several times during the winter rains, with the result 

 that in October, 1906, the average depth of the deposit had been re- 

 duced to about 6in., and a diminished amount of suspended matter 

 was observed in the tank effluent. 



In the following table the results of the analysis for 12 months 

 are set out. It will be seen that, speaking generally, the sewage is of 

 the same character as that seen at the North Sydney outfall, the 

 slightly lower figures for organic matter in the average being probably 

 due to the effect of heavy rain on the three occasions of sampling. It 

 will be noted that a far greater loss of organic nitrogen occurs in the 

 tanks than in North Sydney ; but it will be further observed that the 

 total percentage purification is not so great as at the latter place. The 

 effect of dilution of the sewage by subsoil water is well shown in the 

 analyses of June, September, and October, where it will be seen that 

 the sewage contained a very appreciable amount of oxidised nitrogen. 

 The comparatively high figures for albuminoid ammonia in the final 

 effluents may possibly be due to the washing through of the suspended 

 matter from the tanks mentioned above. 



