XXXIV. J. HAYDON CARDEW. 



ation, percolation and efflux would be an invaluable aid for 

 the prediction of floods ; when the nature of a tributary's 

 basin is known, it is possible to forecast the effect of any 

 storm upon the main river, and by constant observation of 

 all the tributaries of the river, to arrive at a perfect system 

 of prediction of floods, so that the settlers and towns along 

 the banks may be warned in time to make preparation 

 against loss of life and property. 



In this connection Mr. C. J. R. Williams, Assoc, m. inst. c.e., 

 in an admirable paper read before the Institution of Civil 

 Engineers in 1899 gives a very interesting account of the 

 manner in which floods are predicted in the Brisbane River 

 by the aid of hydrography; he describes how by flood gauges 

 in the main stream and its branches, all connected by 

 levelling, with accurate cross sections of the channel, the 

 gradient of the stream when normal and in flood, and 

 with the observed current velocities at different heights, 

 it is quite possible to give an accurate prediction of the 

 height and time of the flood well in advance of its arrival; 

 a comparison of the calculated and observed heights of the 

 actual flood levels show marvellously accurate results, but 

 of course an immense amount of information, involving per- 

 haps years of observation is necessary for such a purpose, 

 however the cost and labour is amply compensated for by 

 the insurance which it provides against all kinds of loss. 

 In the same way we can acquire knowledge which will 

 enable us by the construction of balance reservoirs so to 

 regulate floods as to greatly ameliorate the effects of 

 drought. 



Again the results of hydrological study would be very 

 beneficial for the adequate provision of water supply to 

 large towns, and if proper attention had been paid to this 

 science the Prospect Dam would not be in the very serious 

 condition it is to day, and the lamentable troubles and 



