26 KAINFALL AND FLOODS. 



fresh after heavy rains is much shorter than in the Avon. 

 Still, I think it will be admitted that these observations 

 show that the question of what proportion of a particular 

 rainfall will find its way to the mouth of the river over 

 the basin of which it falls is one which, if it could be 

 answered (which I believe it cannot), has little or no bearing 

 upon the question of flood. 



I repeat that what we want to know is one thing only — 

 it is the maximum rate at which the flood- water will require 

 to be discharged. At the time when this maximum rate — this 

 moment of danger — occurs, it is likely enough that not more 

 than half of the entire rainfall will have been dischargfed. 



So far, my remarks have been open to the charge of being 

 little more than a kind of " destructive criticism." I have 

 shown " how not to do it." I am not sanguine that I have 

 much to offer in the way of positive help towards the 

 solution of the problem which now presses for solution. 

 But considering the extreme uncertainty of a priori con- 

 clusions — an uncertainty inherent, as I believe, in the 

 nature of the case, and abundantly exemplified by the 

 divergence of opinion amongst experts — 1 propose to con- 

 sider whether some more definite clue may not be gathered 

 from actual observations. 



Obviously, in a case of flood, it is not possible to 

 ascertain by actual observation what the maximum rate 

 of discharge of water by the river would have been if the 

 discharge had been sufficiently free to prevent flood. But 

 there are cases in which, as the result of heavy rain, the 

 river almost overflows, or, it may be, just overflows ; and 

 if we know what the maximum rate of discharge has been 

 in such cases, and compared the conditions of the rainfall 

 then with the coDditions of the rainfall at times when 



