30 EAINFALL AND FLOODS. 



a figure, but it can scarcely be rash to say that it is difficult 

 to believe in the necessity, on those occasions, of a provision 

 more than 50 per cent, beyond that v^hich sufficed in 

 October last. Provision for the discharge of 4'3 million 

 cubic feet of water per hour sufficed in October last. Pro- 

 vision of 6| million cubic feet per hour would, I imagine, 

 have sufficed to prevent the two greatest floods that have 

 occurred in forty years, and, a fortiori, all the smaller 

 floods. 



The problem may be attacked in another way. A certain 

 proportion exists in every case between the maximum rate 

 of discharge and the average rate of fall. This proportion 

 cannot be constant, but, with the exception of certain cases 

 to be noticed immediately, I am disposed to think it makes 

 a sufficient approach to constancy to serve as a rough 

 guide. Let us see what the proportion was in October 

 last. 



The fall of rain in 65 hours on that occasion was 2*8 

 inches. The volume of water precipitated over the 68 

 square miles of the Frome basin was therefore about 442 

 million cubic feet. The average rate of the fall per hour 

 was 6*8 million cubic feet. We have seen that the maximum 

 rate of discharge was 43 million cubic feet per hour. The 

 ratio of these two numbers is as 1"58 to 1. 



Applying this ratio to the rainfall of March 6th to 8th, 1889, 

 we get for the maximum rate of discharge 6' 7 million cubic 

 feet per hour. 



Treating in the same manner the rainfall of October 

 22nd to 24th, 1882, we get for the maximum rate of dis- 

 charge 6'2 million cubic feet per hour. 



The calculation may be otherwise expressed in the forms 

 following : — 



