20 EAINFALL AND FLOODS. 



rated. This object might be gained very rouglilj by 

 exposing any flat-bottomed vessel with, vertical sides and 

 simply measuring the depth of water tliat had collected 

 within it. But much more accurate results may be obtained 

 by collecting the rain which falls over an area of know^n 

 dimensions (for example, a circle of eight inches diameter), 

 and measuring it in a vessel of much smaller diameter, 

 graduated in accordance with its relation to the area of the 

 receiving surface. In this way we may easily measure the 

 depth to the thousandth of an inch. 



The principal floods which are liable to occur in Bristol 

 are those which are due to an overflow of the river Frome ; 

 and before an engineer can have the necessary data for 

 determining the provision that should be made for prevent- 

 ing such overflow, he must ascertain the maximum quantity 

 of water which the channel of the Frome may under any 

 circumstances be called upon to carry and discharge. 



This, then, is the problem to be solved. Put in a precise 

 form, what we want to know is the maximum number of 

 cubic feet of water per hour that the Frome may be re- 

 quired to discharge after the heaviest rains, if the city is 

 to be saved from flood. The first step in this inquiry is to 

 find the extent of country which the river Frome drains, 

 or, in technical language, the area of the Frome basin. 



The area of the Frome basin is considered to be 68 square 

 miles. This being known, it is easy to calculate the 

 number of cubic feet of water which will fall over the 

 Frome basin for every inch depth of rain. In a square 

 mile there are (roughly) 28 million square feet. Twelve 

 inches of rain over a square mile would therefore mean 

 28 million cubic feet of water, and one inch of rain would 

 be a twelfth part of that, — namely, about 2,300,000, — or 

 (to put it shortly) 2'3 million cubic feet of water. Multi- 



