44 
make out either. We can easily calculate, with the aid of a map 
and a table of rainfall, how much water is landed in any particular 
river basin in the course of a year. Then, if we make a series of 
estimates of the quantity of water annually discharged into the sea 
by the river itself, we can get a fair approximation to the quantity 
of water that is being returned to the sea by that particular 
channel. Theoretically this quantity should be the remaining 
two-thirds of the total quantity of rainfall, after deducting the 
one-third dissipated by evaporation. Practically the quantity is 
less than two-thirds; but with that particular deficit we are not 
now concerned. 
The history of the quantity discharged by rivers is somewhat 
complex, but, at the same time, it is a matter of considerable 
practical importance in many ways. What happens is something 
of this kind. After heavy rain, when the ground has become 
thoroughly wet, some of the surplus water begins to flow down the 
nearest slopes in the direction of the lowest level it can find; and, 
if nothing happens to prevent its doing so, it will flow on, receiving 
additions right and left as it goes, until it reaches some small sike 
or other established water-channel. This, in its turn, flows on 
until it reaches a beck, the beck until it reaches a river, and the 
river until it joins the sea, and carries the water back to where it 
came from originally. The sun supplies the force that lifts the 
water in the first instance, gravity takes it back. 
But it very rarely happens that rain water goes back to the sea 
by this direct route—certainly not more than half of it flows off 
the surface. Notice the difference in the quantity of water that 
finds its way into the side gutters of a street where the roadway is, 
say, asphalted, and where it is paved with material of a more open 
and porous kind. The gutters in the one case rapidly become 
the channels of small torrents, and as rapidly become dry again ; 
while in the other the quantity running off is comparatively small, 
and is both slower in rising to fulness, and equally slow in drying 
up. These two cases well illustrate on a small scale some phe- 
nomena that occur on scales of almost every magnitude. 
In nature the quantity of surface-water flowing off is in like 
