PROCEEDINGS OF SECTION J. 



667 



as was the case when the tanks were open — the lower parts are very 

 soft and sludgy, so that on withdrawing the measuring pole the opening 

 closes below by collapse of its sides, and no liquid is observed arising 

 therein. In the second place the scum does not present an even upper 

 surface. Bulgings and upheavals occur with corresponding valleys 

 and depressions. Consequently it may happen that varying readings 

 are secured from the one point at different periods, without any material 

 alteration in the average thickness of the scum. Despite these dis- 

 abilities I think the figures given below may be taken as a comparable 

 series, and as fairly representing what they claim. The observations 

 were made by Mr. Cook personally (and checked on the spot by Mr. T. 

 Griffiths, Superintendent Engineer for Sewerage), with a full apprecia- 

 tion of the above-mentioned possible sources of error. 



Thickness nf Scum in the several 2'anks for Each Month. 



..JV' 



were run for five days during heavy rain. 



* Denotes that the tank was in use during the month. 



