168 The N.Z. Journal of Science and Technology. [Aug. 
pay to develop the most favourable site to its practical limit before beginning 
work on the next one. The provision of storage will, moreover, modify 
the site of river-works, as by its augmentation of the summer flow it may 
render unnecessary the locking of the lower portion of the stream, and in 
general it may be said that the most suitable position for a multiple - 
function dam is likely to be towards the upper reaches of a river, so that 
it will extend the navigable length to the greatest extent. 
The theory of the regulation of stream-discharge by reservoirs may be 
approached thus : Suppose that a stream has been gauged continuously 
for a period of many years and its mean flow ascertained, and that a dam 
can be erected of sufficient storage capacity completely to balance the 
variations of inflow. The dam may be assumed to have a self-adjusting 
orifice set to discharge at the rate of the mean stream-flow. The action 
of such a dam would be perfect as regards the down-stream course of the 
river in respect to both navigation and flood-prevention, but the power 
available from the dam will vary owing to the drawing-down of the water- 
level in a dry season, and the navigation of the upper reaches of the reservoir 
would suffer from the same cause. In short, ideal perfection is unattain¬ 
able, though the loss of a certain amount of power in late summer is partly 
offset by the reduced lighting and heating load on electrical circuits at 
this time. 
Practically, the cost of the large reservoirs required to effect complete 
regulation would be prohibitive, even if sites having the necessary storage 
could be found, and it is necessary to ascertain the best method of utilizing 
a reservoir of much smaller capacity. Suppose that the orifice were set 
at the same discharge-rate as before (the mean stream-flow), this smaller 
reservoir will fail both in times of heavy rain and of drought—that is, 
excessive floods will overtop the weir—and the reservoir will run dry 
in every abnormally dry season, allowing the discharge to fall to the 
unregulated flow of the stream. According to the main purpose in view, 
the orifice area may be changed in either of two ways. It may be set at 
an ascertained higher discharge-rate, which will just prevent the tnaximum 
flood from overtopping the weir, and the dam will then hdve its greatest 
value for flood-prevention, but it is easily seen that in this chse its value 
for power and navigation is much reduced, as it will be efnffly during 
considerable periods in dry weather; or the orifice may be feet at a lower 
discharge-rate, which the reservoir is just capable of mairltJifiing through 
the driest season on record. This is the value of most interest to power and 
navigation schemes—here called the “ perennial }deld.” 
The relation of the two discharge-rates to the reseTVoir-bapacity is 
illustrated in fig. 2, which is calculated from the gaugings of a small stream 
near Auckland* having a catchment area of 3f square miles with a mean 
annual rainfall of about 65 in. Although these curves would not be directly 
applicable to watersheds of other areas or conditions, their general form 
should be similar in all cases. There are certain points of particular interest 
on the perennial-yield curve. It is probable that this curve is asymptotic, 
never quite reaching the mean stream-flow, owing to long-period variations 
of climate, but for practical purposes it may be said that full regulation 
is obtained with a reservoir of about 30 in. capacity, as defined on page 165. 
A capacity of 15 in. will make available the whole stream-flow during the 
three driest years in succession, approximately three-quarters of the mean 
* The gaugings of this stream (the Nihotupu) are used only to illustrate the 
method of calculation. It is not suggested that this stream is suitable for purposes 
other than city water-supply. 
