152 Transactions of the Royal Society of Victoria. 



The first two are my own, the last was obtained from a 

 report on the Coliban works, which also mentioned a reported 

 discharge of 32,132, but this I think doubtfril. I should 

 like, if I could obtain the necessary data, to obtain co- 

 efficients for various average chai'acter of watersheds. I 

 need hardly remark that this formula would be inapplicable 

 for very absorbent ground or sandy areas. 



We may also estimate the discharge in a more direct 

 manner. 



Let MP = rainfall for any period, 77 = rate of fall, for 



L 



instance so many inches per hour, where t is the time the 

 water takes to travel the length of the watershed. If we 

 put S for hourly soakage, as it will take less than t for the 

 average time for the whole of the water to reach the outlet, 



t 

 it may be t^ or -, and knowing that 1 inch rain per hour 



represents very nearly a discharge of 1 foot per second per 

 acre, we arrive at the following : — 



(acres; / __ § ^ =: discharge per second. 



area \ P x / 



As time and the length depend on each other, substitute 



V 



we obtain 











a ( ■ 



R 



(1)^ 



s 1 



vx / 



= discharge. 



Again, 



assuming s v x 



to be constants, we obtain 





/ 



X 



, X 



/ X , \ 



(|-?2/) or a (^- ly) 

 Which corresponds very nearly with Mr. Surge's No. 3 



« ( I) 

 The effect of soakage being omitted, hence I have adopted 



181 



discharge = a 7123 , iqaq ^ ^^^ ^ ^ chains. 



