"218 ACTION or FORKS^TS ON THE SURFACK-FT.t i\V i.F RAIN WATER. 



effect of modulating the properties of rocks, giving to them what they 

 have not ; and he alleges it is in this way, more especially, that their 

 action is salutary in the control of waters on the mountains. Of this 

 view of the subject he gives the following illustration : — " When the 

 rain falls on a denuded brow of a hill, composed of argillaceous earth 

 the water moistens the surface — this absorbs a great quantity of it, 

 through its hydroscopicity and capillarity — but when once this surface 

 is moistened, the transmission of water goes on only very slowly from 

 particle to particle, for the permeability is almost nothing, in 

 consequence of the minute subdivision of the molecules which are 

 brought into the closest contact ; and that which is absorbed remains 

 on the surface and dilutes the superficial layer, which is soon thus trans- 

 formed into a thin clay devoid of cohesion. A layer more or less deep 

 will then detach itself from the surface of the mass, and will flow to the 

 bottom as mud more or less fluid, according as the rain may have 

 been more or less violent. By a very gentle rain, a superficial layer 

 is moistened ; but the water falling slowly may be able to penetrate 

 it completely, in virtue of its hydroscopicity and capillarity. In this 

 cas.^ there will be only a superficial flow, for the greater portion of the 

 water will penetrate the soil. 



" But suppose that this same argillaceous land, or other unstable 

 ground, were wooded, the trees in spreading the fall of the rain over 

 an expanded surface, that of their foliage, would moderate the rain- 

 fall, and would at the same time augment the absorbent power of the 

 soil, as well as its permeability, and as a consequence augment the 

 quantity of water retained superficially. The mobility of the surface 

 thus softened would, undoubtedly, be increased, but the roots im- 

 prisoning it would retain the softened ground to such a degree that 

 no amount of water falling upon them from the heavens alone could 

 cause it to slip away. Wherever landslips occur on wooded grounds 

 they can be otherwis9 accounted for. 



" If, in conclusion," says he, " we examine a forest situated on a 

 land permeable en masse, as are the plateaux of Jurassic limestone, the 

 first effect of the forest would be to cover the soil with a thick layer 

 of humus and of moss, which combine in a very high degree 

 hydroscopicity and capillarity. The quantity of water retained thus 

 in the upper layer of the soil will be much greater than it would have 

 been had there been no forest there, for on the rocks referred to the 

 destruction of woods is alnaost immediately followed by a denudation 

 of these rocks of soil. 



" It follows from this that on these lands the forest arrests the 



