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 and his commentator Valles have deduced an opposite result from various facts of experience and from scientific considerations. They contend that the superficial drainage is more regular from cleared than from wooded ground, and that clearing diminishes rather than augments the intensity of inundations. Neither of these conclusions appears to be warranted by their data or their reasoning, and they rest partly upon facts, which, truly interpreted, are not inconsistent with the received opinions on these subjects, partly upon assumptions which are contradicted by experience. Two of these latter are, first, that the fallen leaves in the forest constitute an impermeable covering of the soil over, not through, which the water of rains and of melting snows flows off, and secondly, that the roots of trees penetrate and choke up the fissures in the rocks, so as to impede the passage of water through channels which nature has provided for its descent to lower strata. As to the first of those, we may appeal to familiar facts within the personal knowledge of every man acquainted with the operations of sylvan nature. Rain-water never, except in very trifling quantities, flows over the leaves in the woods in summer or autumn. Water runs over them only in the spring, in the rare cases when they have been pressed down smoothly and compactly by the weight of the snow--a state in which they remain only until they are dry, when shrinkage and the action of the wind soon roughen the surface so as effectually to stop, by absorption, all flow of water. I have observed that when a sudden frost succeeds a thaw at the close of the winter, after the snow has principally disappeared, the water in and between the layers of leaves sometimes freezes into a solid crust, which allows the flow of water over it. But this occurs only in depressions and on a very small scale; and the ice thus formed is so soon dissolved that no sensible effect is produced on the escape of water from the general surface. As to the influence of roots upon drainage, we have seen that there is no doubt that they, independently of their action as absorbents, mechanically promote it. Not only does the water of the soil follow them downwards, but their swelling growth powerfully tends to enlarge, not to obstruct, the crevices of rock into which they enter; and as the fissures in rocks are longitudinal, not mere circular orifices, every line of additional width gained by the growth of roots within them increases the area of the crevice in proportion to its length. Consequently, the widening of a fissure to the extent of one inch might give an additional drainage equal to a square foot of open tubing.