114 TYPICAL QUEENSLAND LAGOON 
consider in anything like detail the results obtained from 
but one of them, a filter whose section is composed entirely 
of sand. This filter has an area of 25 square feet, and a 
section as follows :— 
Fine sand at top of bed (effective size 335 m.ms.) 41 inches 
Underdrains composed of various grades of gravel 12 ,, 
Total depth of material Be exe jouer Oo mane 
The depth of water over the sand was 41 inches, whilst 
the maximum loss of head was 91 inches. The method of 
operating the filter is as follows :—It is fed from below with 
filtered water until it attains a depth of about 6 inches above 
the surface of the sand, unfiltered water is then fed on to 
the bed from above, and filtration commenced as soon 
as the water in the filter is 2 feet over the sand. For 
the greater portion of the time during which this filter was 
in operation, water was discharged on to the bed through 
an aerator, which divided it into a comparatively fine spray. 
The filter was allowed to run until it became so choked 
with organic matter that a “head” or pressure ot 7 feet 
10 inches was necessary to force the water through the sand 
at the required rate. The feed water was then turned off, the 
sand allowed to drain, and the surface removed for a depth 
of about half-an-inch. The filter was then restarted as before 
detailed. The period between the times when scraping the 
surface to remove clogging becomes necessary is termed 
a “run,” and varies according to the nature of the applied 
water. The average length of run for this filter during the 
time it was in operation, was over 70 days, showing clearly 
that the sand used for the filter bed was not by any means 
too fine, as this is a comparatively long run for a filter of this 
type. This filter remained in operation for 183 days, the 
average rate of filtration maintained throughout that period 
was exactly 2 million gallons per acre per 24 hours. 
Chemical and Physical analyses of the unfiltered water and 
effluent were made weekly, and a monthly average of the 
percentage reduction of organic matter and colour, is given 
in Appendix VIII. 
The average reduction of organic matter, as shown by the 
Albuminoid Ammonia determinations throughout the whole 
period of operation was more than one-half, whilst the reduc- 
