134 Failure of the Yan Yean Reservoir. 
ment of 5,000 gallons per minute in‘December as a reliable 
average. 
My own estimate, on the same ‘principle, is 5,833 gallons 
per minute, allowing two-thirds of increase for the three 
winter months; but this again is reduced by 900 gallons, 
which I allow as a minimum, and very scanty supply for the 
inhabitants of the district. The available discharge is, there- 
fore, 4,933 gallons per minute, independent of the floods, or 
six feet seven inches in the reservoir. 
I do not say that the discharge does not frequently exceed 
this; but I am strongly of opinion that in some seasons it 
does not do so. 
I shall now, however, consider what may be regarded as 
the highest average, and I shall deduce the amount from the 
sectional measurements of the river. 
It may be considered as an axiom, that when a river has 
defined banks, these indicate its ordinary limits, which it only 
exceeds in time of floods. In other words, every river ma 
be regarded as having excavated for itself a bed sufficiently 
large to hold its ordinary stream. The ordinary stream, 
therefore; will be confined within the ordinary banks, and the 
highest average in the winter, unless in floods, will not over- 
flow the banks. 
Let us examine the sections of the bed of the river at the 
entrance of the aqueduct. The mean of the sections gives 
28 feet within the banks, and 13-2 feet under the 
water line. Therefore, with the same velocity of half a mile 
per hour, the section could contain no more than 5,381 
gallons per minute, without flooding the right bank. With 
double the volume, the velocity would not be increased, 
according to the usual formule, by one half. Therefore, I con- 
sider it to be demonstrated, that 8,071 gallons per minute, or 
three times the January measurement, is the highest discharge 
of the river, except in floods; and-it is exceedingly rare to find 
any river in Australia level with its banks for six months in 
the year. 
This discharge will give five feet four and a half inches in 
the reservoir for the six winter months; but it will be 
observed that it does not include the amount at present lost 
in the swamps, which I have calculated at 1,830 gallons per 
minute for the whole year, or two feet five inches in the 
reservoir. 
The loss in January, as we have seen, is 3,253 gallons per 
minute, or at the rate of four feet four inches in the reservoir. 
