170 
The N.Z. Journal of Science and Technology. 
[Aug. 
gauging. (It is perhaps hardly necessary to give a word of caution against 
accepting any gauging as being the minimum stream-flow unless it is known 
# to have been taken during a period of extreme drought; the low-water 
flow of an average summer may be more than double the true minimum.) 
An attempt to make the curve of fig. 2 generally applicable leads to the 
expression 
Y = 
R - 15 
70 _ 
\/ x + a 
Where Y = perennial yield (cusecs per square mile of catchment) ; 
R - mean annual rainfall (inches) ; 
x — storage provided by dam (inches)— i.e., 
storage capacity (cubic feet) 
2,323,200 X catchment (square miles) ’ 
a = natural storage, computed by putting x zero and Y the 
minimum gauging. 
As this formula has not yet been tested on any other watershed no 
accuracy is claimed for it, but it may be useful for a preliminary estimate to 
indicate whether a proposal is worth further investigation. It must not be 
used for values of x + a exceeding 20. 
When finally ascertained the “ perennial yield ” represents the rate 
at which water may be drawn continuously from the reservoir without risk 
of failure in a dry season, and would be adopted for practical operation of 
a scheme devised in the joint interests of power and navigation. A certain 
amount of relief from flooding would also result as a “ by-product,” 
becoming more complete as the capacity of the reservoir approached 
the ideal. 
Actually the rate of draw-off will not be uniform, but will vary during 
the day in accordance with the load curve on the power plant. For a load 
factor of 50 per cent, the maximum flow of the stream will be double, and 
the mean rate during working-hours about one and three-quarter times 
the■“ perennial yield,” but it will drop between midnight and early morning 
to perhaps a fifth of the mean value. These variations require the pro¬ 
vision of a storage, for typical load curves, of seven or eight hours’ mean 
discharge, which must be deducted from the reservoir-capacity before com¬ 
puting the seasonal regulation. Usually the loss is scarcely appreciable, 
and of more importance is the effect on navigation below the power-station. 
As river traffic usually prefers daylight, the variation in flow should be 
favourable—at any rate, for a moderate distance down-stream—by pro¬ 
viding greater depth of water during working-hours; but the storage action 
of the pools, bays, and backwaters of a natural stream will probably wipe 
out the “ load wave ” before it travels many miles. 
In conjunction with river-works the preservation of natural storage is 
highly important, both to maintain the yield and to prevent the filling-in 
of reservoirs by the debris of erosion. Afforestation of mountainous areas 
and deep cultivation and terracing of farm lands are to be encouraged, 
and the draining of swamps and lakes of the upper watershed treated with 
suspicion. It is occasionally proposed to “ regulate ” lakes by works designed 
to prevent their seasonal changes of levels. Such a policy would be dis¬ 
astrous as regards the regulation of the outflowing streams, as it is the 
variation in lake-level which acts so beneficially in equalizing the stream- 
flow. 
