118 TYPICAL QUEENSLAND LAGOON 
an intermitter from which the water is discharged at intervals 
into the basin of a revolving sprinkler running on ball bear- 
ings, which distributes it reguiarly and evenly over the surface 
of the filter bed, through which it slowly percolates and flows 
into underdrains which lead into a small receiving vessel. 
The object of this method of discharging the effluent is to 
insure the thorough aeration of the water in its passage through 
the filters. That this has been effected is proved by the fact 
that even in the warmest weather the effluents from all but one 
of these filters contained over 95 per cent. of dissolved oxygen, 
and it has been proved that this aeration is chiefly produced 
during the passage of the water through the bed of the filter 
itself. Altogether 8 filters of this type were experimented 
with ; they may be divided into two classes, namely, those 
in which sand forms a portion of the filter bed and those in 
which the filter bed consists entirely of gravel. 
Two beds composed partly of sand were tried. In each 
of these the surface layer was composed of one foot in depth 
of sand, which was also of the same grade in each case. In one, 
the remainder of the filter bed was composed of coal in four 
layers of different grades, the coarsest with material 4 inch 
in diameter, being at the bottom, making the total depth 
of material 6 feet. In the other filter the remainder of the 
bed is composed of gravel in four layers, which are identical 
in grade and depth with those in the coal filter. The coal 
filter is fitted with means of securing both side and bottom 
aeration, whilst no provision for such aeration is made in the 
gravel bed. In consequence of this the amount of dissolved 
oxygen in the effluent from the coal bed is much greater 
than in the gravel bed effluent. The reduction of colour 
and organic matter however, was found to be much better 
in the gravel than in the.coal bed filter, whilst the bacterial 
purification was the same in each, the reduction effected 
being about 92 per cent. The effluent from this gravel and 
sand bed filter was in every way excellent, and the average 
results are equal to the best results obtained by sand 
filtration. 
The beds in which gravel alone was used, though effect- 
ing a remarkable reduction in the organic matter and colour 
of the water, are open to two objections, firstly—They take 
a considerable time to properly mature or get in working 
