DRAINAGE OF FOREST-SOIL. 241 
and melting snow. Nothing more effectually disintegrates a 
cohesive soil than freezing and thawing, and the surface of 
earth which has just undergone those processes is more subject 
to erosion by running water than under any other circumstances. 
Hence more vegetable mould is washed away from cultivated 
grounds in such climates by the spring floods than by the heaviest 
rain at other seasons. 
In the warm climates of Southern Europe, as I have already 
said, the functions of the forest, so far as the disposal of the 
water of precipitation is concerned, are essentially the same at 
all seasons, and are analogous to those which it performs in the 
Northern United States in summer. Hence, in the former 
countries, the winter floods have not the characteristics which 
mark them in the latter, nor is the conservative influence of 
the woods in winter relatively so important, though it is equally 
unquestionable. 
Tf the summer floods in the United States are attended with 
less pecuniary damage than those of the Loire and other rivers 
of France, the Po and its tributaries in Italy, the Emme and 
her sister torrents which devastate the valleys of Switzerland, 
it is partly because the banks of American rivers are not yet 
lined with towns, their shores and the bottoms which skirt them 
not yet covered with improvements whose cost is counted by 
millions, and, consequently, a smaller amount of property is ex- 
posed to injury by inundation. But the comparative exemp- 
tion of the American people from the terrible calamities which 
the overflow of rivers has brought on some of the fairest por- 
tions of the Old World, is, in a still greater degree, to be 
ascribed to the fact that, with all our thoughtless improvidence, 
we have not yet bared all the sources of our streams, not yet 
overthrown all the barriers which nature has erected to restrain 
her own destructive energies. Let us be wise in time, and 
profit by the errors of our older brethren! 
The influence of the forest in preventing inundations has 
been very generally recognized, both as a theoretical inference 
and as a fact of observation ; but the eminent engineer Belgrand 
