RAINFALL AND FLOODS. 21 



plying tliis by the number of square miles (68) we get 156, 

 or, more accurately, 158 million cubic feet as the quantity 

 of water which will fall over the Frome basin for every inch 

 of rain. 



This is all simple enough, but we are only at present on 

 the threshold of the inquiry. The later stages of it bristle 

 with clifiBculties — difficulties which I venture to think have 

 hardly been sufficiently considered by some of the eminent 

 engineers whose opinions have been sought. I am em- 

 boldened to say so much as this by observing the very wide 

 divergences of opinion expressed by engineers with regard 

 to the provision which should be made for carrying off the 

 flood-water of the Frome after heavy rains. 



A former engineer of the Bristol Docks, Mr. Howard, 

 held that 8 million cubic feet per hour might be taken 

 as the maximum discharge for which provision should be 

 made. A more recent opinion has been given, I am told, 

 that safety will not be secured until provision is made for 

 the discharge of from 18 to 21 million cubic feet per 

 hour — an estimate which appears the more remarkable 

 when contrasted with an important observation of Mr. 

 McCurrick, the present engineer of the Docks, on the 19th 

 of October last, showing that a discharge of less than 

 4|- million cubic feet per hour sufficed on that occasion to 

 prevent flood, — the occasion being one which, if not strictly 

 crucial as an experiment, was at least very exceptional in 

 character. 



It is well known that the volume of water discharged by 

 a river in a given interval of time — say a year — is very much 

 less than the volume of rain which falls in the same in- 

 terval over the basin of that river. A similar result is 

 found in what are known as percolation experiments. A 

 cylinder or a tank from 2 to 5 feet in depth is filled with 



