and the Storage of Water. 
19 
100 cubic feet per minute at the junction with the Thames at 
Cricklade, 22 miles from the source. Allowing for the in- 
creased drainage area, this showed a loss of 340 cubic feet per 
minute. To remedy this, Mr. Simpson advised the mill-owners 
that the bottom of the stream and the fish-ponds through which 
the river passed should be puddled, at an estimated cost of 
3000/.* 
Subsequentlv, at a public meeting held to consider this subject, 
arrangements were entered into by which the millers and others 
interested agreed to contribute to a common fund for the pay- 
ment of men to be regularly employed to inspect the stream, 
and to puddle the bottom and sides wherever leaks were dis- 
covered, f 
In like manner streams may be abnormally increased in 
volume by springs which are fed by rain falling outside the 
watershed of the river in the manner shown in the illustra- 
tion (Fig. 1, p. 11). The River Frome, in the neighbourhood of 
Chalford, when gauged in its ordinary condition by Mr. Taunton, 
was found to yield, about 7 miles from its source, in a dry season, 
321 cubic feet a minute. This quantity was increased, within 
three-quarters of a mile, to 1(305 cubic feet, by very strong 
springs, one of which alone was found to yield 64 cubic feet a 
minute. The ordinary summer flow of the river in dry weather 
was 481 cubic feet per minute, equal to about 28 per cent, of 
the total fall of rain on an area of about 25 square miles, but for 
a few springs it represents only 17 per cent, of the total quantity.^ 
Owing to the various causes already described, it will be seen 
how difficult, in fact impossible, it is to lay down any rule or 
formula which can be applied generally to the proportion of 
rainfall which is discharged by the river or stream draining the 
district. It is only by carefully noting the statistics of rain for 
each particular district — examining the geological condition of 
the district in which the river has its source and through which 
it flows, with the nature of the vegetation with which the surface 
is covered — that even an approximate estimate can be formed 
of the quantity of water for which proper provision should be 
made. 
DiSCHAEGE OF WaTER BY NATURAL CHANNELS. 
Having settled the quantity of rainfall to be drained off anv 
given district, the next point for consideration is the best form 
* " The Perennial and Flood Waters of the Upper Thames," ' Trans. Instit. 
Civil Engineers,' vol. xxii. 
t House of Lords' Committee on Conservancy Boards; Evidence of Mr. Taunton. 
QQ. 2180, 2200, 2230. 
I " Rainfall and Evaporation," ' Trans. Instif. Civil Engineers,' vol. xlv. 
c 2 
