View report page

174
ground collection of water on about 20 acres of land, and beingable,
during the time of flood, to raise large quantities of water
that has passed through the natural beds of gravel and sand,
which are characteristic of the neighbourhood. The freshets
and floods in the rivers were of average character in 1888, and
no special difficulties as regards filtration of the water supply
presented themselves. The results were on the whole satisfactory,
and generally speaking the filters were maintained in
a satisfactory condition. Turbid water is very difficult to filter.
The solid impurities in suspension, chiefly consisting of clay,
marl, and chalk, in a very finely divided state, can be got rid of
only by long subsidence previous to filtration. Turbid water
must of necessity be sometimes admitted, and filters are then
overtaxed.
The fact that the impurity of the water in the river is so
largely increased during periods of flood, renders it of great
importance that there should be sufficient provision cf subsidence
and storage reservoirs to enable the necessary daily supplies to
be delivered during such periods, while keeping the intakes
closed. The position of the several Companies varies in this
respect very considerably. The following statement shows the
number of days' supply, represented by the capacity of the
reservoirs for unfiltered water, behnging to each of the local
Companies:—Chelsea, 14; West Middlesex, 8 ; Grand Junction,
4. But the Grand Junction Company's resource during
floods, of pumping from the gravel beds adjoining the Thames,
is practically equivalent to an addition to their provision of
storage reservoirs. For sake of comparison it may be added
that the East London Company has capacity for 18 days' supply.
The same Company is conspicuous by its advance in work for
improving the water supply of its district in other respects.
Dr. Percy Frankland's biological experiments by the
process known as gelatine-plate cultivation, from which it
appeared that in 1887 the average reduct on in the number of