386 REPORTS ON THE STATE OF SCIENCE, ETC. 
* 
(a) Snow and ice, increasing the flows of the late spring and. causing, 
during a thaw, sudden heavy floods in the lowlands. 
(b) Ground saturation, causing quicker run-off in floods and a slower 
fall after the cessation of rain. 
(c) Tarns, lakes, reservoirs and canals, holding up and prolonging the 
floods and having varying effects on low flows. 
(7) Flooded areas, ditches, etc., in the lowlands, reducing the peak 
intensity of flood flows. 
(e) The capacity of the river channel and artificial channels in the low- 
lands, and of coastal embanked areas which keep out tidal flow. 
All of these affect the water levels and water slopes. 
The measurements of evaporation and condensation, etc., are subsidiary 
to the main measurements and, in fact, cannot be ascertained except through 
analysis of the completed records. 
The capacity for storage of different portions of a catchment is one of 
the principal factors governing intensity and duration of flood flows. When 
rainfall ceases, there are all descriptions of storage in action and, as the 
rainless period extends, the aggregate flow since rain ceased comes nearer 
and nearer to the measure of total area storage. An area will drain off in 
very similar fashion after each flood, subject to certain variants such as 
saturation or melting snow, etc. 
Therefore, after rain ceases, the aggregate flow measures, firstly, the 
flood capacity of the river and some flooded areas: a most valuable 
measurement combined with land survey and river-bed survey and water 
levels. Later, the aggregate flow measures the de-saturation of the area. 
The rains which fall in summer and do not run off are also a measure of this 
de-saturation volume. 
The continuous record of water level at carefully selected sites is the 
foundation of all problems connected with the various forms of storage. 
When the areas of lakes and reservoirs are known at different water 
levels and the outflow is known, the rate of change in water level of such 
areas yields the inflow value. 
When, on a river, the inflow to and outflow from a flooded area has 
been measured, the capacity of such flooded area can be estimated. 
The times of concentration at various points on streams and rivers are 
a measure of area storage. This is particularly applicable to reservoired 
areas, when the interposition of a reservoir lengthens the period of concen- 
tration according to its volume and its weir dimensions. 
In the control of floods, water level information on all these storage areas 
throughout the catchment and the knowledge of the times of filling and 
emptying such storages are of the first importance. 
The bearing of storage on the maintenance of good low flows is obvious 
and probably requires no further comment here. 
(C) Flow.—The River Basin is the obvious unit of area for the records 
of flow, and the general characteristics of the area will indicate the sites 
at which flow measurements should be made and at which continuously 
recorded water levels should be maintained. Gauge post water levels at 
numerous other sites will complete the picture. 
The principal object should always be the measure of aggregate flow as 
obtained by continuity of water level records. The actual measurements 
of flow, for all water levels, at a selected site should be done once and for 
all with the utmost accuracy. 
The aggregate flow which passes a gauging site during a flood or series 
of floods is only in a secondary manner affected by the configuration of 
