120 TYPICAL QUEENSLAND LAGOON 
coagulant is generally necessary in order to remove a portion 
of the polluting matter before filtration, and to quickly form 
on the surface of the sand in the filter a coagulum which 
takes the place of the natural film produced by subsidence 
of suspended matter and bacterial action on the surface of a 
plain sand filter. This form of filtration is only “ mechanical ” 
in connection with the arrangements for agitating the sand 
during the frequent cleansing necessary, and for the regular 
dosing of the applied water with the solutions of coagulant 
Experiments conducted with the water of our lagoon 
showed the best proportion of coalguants to be 1 grain per 
gallon of alum and 43 grains per gallon of hme. Four hours 
settlement after adding the coagulants but prior to filtra- 
tion was also found to be necessary. 
Time does not permit of a detailed description of the 
experiments conducted with this process, which extended 
over a period of six months, I must content myself with 
stating that very excellent results were obtained throughout 
the experiments. 
For purposes of comparison, I have constructed tables 
showing the average composition of the effluent of the 
mechanical filter for a period of three months, during which 
the best results were obtained at a rate of filtration equal 
to 104 million gallons per acre per day. For purposes 
of comparison, I have also calculated the average com-- 
position of the effluent from the intermittent plain 
sand filter before described during a _ similar period 
of three months, whilst operating at an average rate per day 
of 2.23 million gallons per acre, and a rest interval of 8 hours 
in every 24. I have also calculated in each case the reduces 
tion of organic matter effected in the same period. This shows 
the percentage reduction of organic matter, as indicated by 
the albuminoid ammonia figures to be roughly 58 in the 
case of the intermittent plain sand filter, against 56 in the 
case of the mechanical filter, although the average 
albuminoid ammonia contents of the plain sand _ filter 
effluent was .136 parts per million, as against .123 parts 
per million in the mechanical filter. It is true also 
that the best result obtained by mechanical filtration 
was better than the best result obtained from the plain 
intermittent sand filter, but in view of the decided 
