Babinet quotes a French proverb, "Summer rain wets nothing," and explains it as meaning that the water of such rains is "almost totally taken up by evaporation." "The rains of summer," he adds, "however abundant they may be, do not penetrate the soil to a greater depth than 15 or 20 centimetres. In summer the evaporating power of the heat is five or six times as great as in winter, and this power is exerted by an atmosphere capable of containing five times as much vapor as in winter." "A stratum of snow which prevents evaporation [from the soil] causes almost all the water that composes it to filter down into the earth, and form a reserve for springs, wells, and rivers which could not be supplied by any amount of summer rain." "This latter--useful, indeed like dew, to vegetation--does not penetrate the soil and accumulate a store to feed springs and to be brought up by them to the open air."[397] This conclusion, however applicable it may be to the climate and soil of France, is too broadly stated to be accepted as a general truth, and in countries where the precipitation is small in the winter months, familiar observation shows that the quantity of water yielded by deep wells and natural springs depends not less on the rains of summer than on those of the rest of the year, and, consequently, that much of the precipitation of that season must find its way to strata too deep to lose water by evaporation. The supply of subterranean reservoirs and currents, as well as of springs, is undoubtedly derived chiefly from infiltration, and hence it must be affected by all changes of the natural surface that accelerate or retard the drainage of the soil, or that either promote or obstruct evaporation from it. It has sufficiently appeared from what has gone before, that the spontaneous drainage of cleared ground is more rapid than that of the forest, and consequently, that the felling of the woods, as well as the draining of swamps, deprives the subterranean waters of accessions which would otherwise be conveyed to them by infiltration. The same effect is produced by artificial contrivances for drying the soil either by open ditches or by underground pipes or channels, and in proportion as the sphere of these operations is extended, the effect of them cannot fail to make itself more and more sensibly felt in the diminished supply of water furnished by wells and running springs.[398]