INLAND WATER SURVEY 421 
considerable period, the curve being flat-topped ; while after short periods 
of heavy rainfall the high water-level soon drops, giving pointed curves. 
(4) The average curve of pumping-levels follows closely the curve of 
rest-levels. In the present case a composite pumping-level curve is shown, 
stronger pumps having been installed in 1926. A measure of water 
pumped is shown by the difference in feet between the two water-level 
curves ; and the stronger pumps increase this difference without otherwise 
affecting the similarity of the two curves. From this illustration it is seen 
that in wells where it is not possible to measure rest-levels, except at very 
long intervals, some idea of the rest-level may be obtained if records of 
pumping-levels are kept, and the difference between pumping and rest- 
levels, as determined by occasional measurement, added. 
(5) Compare— 
(a) ‘ London Wells ’ (Mem. Geol. Surv.), Plate III. 
(6) £ Wells and Springs of Sussex ’ (Mem. Geol. Surv.), pp. 19-20. 
(c) ‘ Hydrogeological Conditions in the Chalk at Compton, Sussex,’ 
Inst. Water Engineers, 1921. 
(d) British Rainfall, 1919, pp. 257-262, for summary of details, and 
bibliography, of records from 1836 onwards, for Chilgrove. 
Matin MeEmMoranpumM F. 
RIVER GAUGING. 
By W. N. McC ean. 
1. CHOICE OF SITES. 
The sites generally chosen by river authorities for gauging rivers are at 
weirs which have been constructed for the purpose of abstracting water or 
controlling flow. ‘These sites are chosen because, by installing a recorder 
water level gauge in still water above the weir, the flow at any water level 
is given by some weir formula. That there is any resemblance to accuracy 
during floods seems to be unlikely. For low flow estimates they may be 
very convenient. 
In practice, the weir itself is not the weir of the laboratory experiment ; 
except at low flows, it is partially or wholly submerged and a varying formula 
has to be used according to the downstream water levels at some vague 
point. The velocity of approach is probably unmeasured. Records of 
the water abstracted, which may be the greater part of the low flow, have to 
be kept accurately. There are often many complications of sluice control. 
Generally speaking, the best site for river gauging, with the current meter, 
is a steady-running reach with a steeper water slope down stream, so that 
there may be no backing up due to the downstream flood conditions. 
These sites are not always obtainable, and often the water-slopes on rising 
and falling rivers will have to be very carefully measured during the gaugings 
and afterwards for the records. 
Sites should be chosen where the river is always within its banks or where 
the overflow is never of any considerable amount. 
The choice of a site for gauging is not often an easy matter and it may be 
necessary to use different sites for floods and low waters. There is much 
to be said for the reading of low flows at weirs or in narrow channels and the 
big wide reaches are best for floods. 
_ The main object of gauging a river should be to obtain continuous records 
of water level and of corresponding flow. 
