"To continue the comparison," observes D'Hericourt, "roofs are smooth and impermeable, and the rain-water pours rapidly off from their surfaces; but this rapidity of flow would be greatly diminished if the roofs were carpeted with mosses and grasses; more still, if they were covered with dry leaves, little shrubs, strewn branches, and other impediments—in short, if they were wooded."—Annales Forestieres, Dec. 1857, p. 311.
The mosses and fungi play a more important part in regulating the humidity of the air and of the soil than writers on the forest have usually assigned to them. They perish with the trees they grow on; but, in many situations, nature provides a compensation for the tree-mosses and fungi in ground species, which, on cold soils, especially those with a northern exposure, spring up abundantly both before the woods are felled, and when the land is cleared and employed for pasturage, or deserted. These humble plants discharge a portion of the functions appropriated to the wood, and while they render the soil of improved lands much less fit for agricultural use, they, at the same time, prepare it for the growth of a new harvest of trees, when the infertility they produce shall have driven man to abandon it and suffer it to relapse into the hands of nature.
In primitive forests, when the ground is not too moist to admit of a dense growth of trees, the soil is generally so thickly covered with leaves that there is little room for ground mosses and mushrooms. In the more open artificial woods of Europe these forms of vegetation, as well as many more attractive plants, are more frequent than in the native groves of America. See, on cryptogamic and other wood plants, Rossmassler, Der Wald, pp. 82 et seqq., and on the importance of such vegetables in checking the flow of water, Mengotti, Idraulica Fisica e Sperimentale, chapters xvi. and xvii. No writer known to me has so well illustrated this function of forest vegetation as Mengotti, though both he and Rossmassler ascribe to plants a power of absorbing water from the atmosphere which they do not possess, or rather can only rarely exercise.]
The value of the forest as a mechanical check to a too rapid discharge of rain-water was exemplified in numerous instances in the great floods of 1866 and 1868, in France and Switzerland, and I refer to the observations made on those occasions as of special importance because no previous inundations in those countries had been so carefully watched and so well described by competent investigators. In the French Department of Lozere, which was among those most severely injured by the inundation of 1866—an inundation caused by diluvial rains, not by melted snow—it was everywhere remarked that "grounds covered with wood sustained no damage even on the steepest slopes, while in cleared and cultivated fields the very soil was washed away and the rocks laid bare by the pouring rain." [Footnote: See, for other like observations, an article entitled Le Reboisement et les Inondations, in the Revue des Eaux et Forets of September, 1868]
The Italian journals of the day state that the province of Brescia and a part of that of Bergamo, which have heretofore been exposed to enormous injury, after every heavy rain, from floods of the four principal streams which traverse them, in a great degree escaped damage in the terrible inundation of October, 1872, and their immunity is ascribed to the forestal improvements executed by the former province, within ten or twelve years, in the Val Camonica and in the upper basins of the other rivers which drain that territory. Similar facts were noticed in the extraordinary floods of September and October, 1868, in the valley of the Upper Rhine, and Coaz makes the interesting observation that not even dense greensward was so efficient a protection to the earth as trees, because the water soaked through the sod and burst it up by hydrostatic pressure. [Footnote: Die Hochwasser in 1868 im Bandnerischen Rheingebiet, pp. 12, 68.
Observations of Forster, cited by Cezanne from the Annales Forestieres for 1859, p. 358, are not less important than those adduced in the text. The field of these observations was a slope of 45 degrees divided into three sections, one luxuriantly wooded from summit to base with oak and beech, one completely cleared through its whole extent, and one cleared in its upper portion, but retaining a wooded belt for a quarter of the height of the slope, which was from 1,360 to 1,800 feet above the brook at its foot.
In the first section, comprising six-sevenths of the whole surface, the rains had not produced a single ravine; in the second, occupying about a tenth of the ground, were three ravines, increasing in width from the summit to the valley beneath, where they had, all together, a cross-section of 600 square feet; in the third section, of about the same extent as the second, four ravines had been formed, widening from the crest of the slope to the belt of wood, where they gradually narrowed and finally disappeared.
For important observations to the same purpose, see Marchand, Les
Torrents des Alpes, in Revue des Eaux et Forets for September, 1871.]
The importance of the mechanical resistance of the wood to the flow of water OVER THE SURFACE has, however, been exaggerated by some writers. Rain-water is generally absorbed by the forest-soil as fast as it falls, and it is only in extreme cases that it gathers itself into a superficial sheet or current overflowing the ground. There is, nevertheless, besides the absorbent power of the soil, a very considerable mechanical resistance to the transmission of water BENEATH the surface through and along the superior strata of the ground. This resistance is exerted by the roots, which both convey the water along their surface downwards, and oppose a closely wattled barrier to its descent along the slope of the permeable strata which have absorbed it. [Footnote: In a valuable report on a bill for compelling the sale of waste communal lands, now pending in the Parliament of Italy, Senator Torelli, an eminent man of science, calculates that four-fifths of the precipitation in the forest are absorbed by the soil, or detained by the obstructions of the surface, only one-fifth being delivered to the rivers rapidly enough to create danger of floods, while in open grounds, in heavy rains, the proportions are reversed. Supposing a rain-fall of four inches, an area measuring 100,000 acres, or a little more than four American townships, would receive 53,777,777 cubic yards of water. Of this quantity it would retain, or rather detain, if wooded, 41,000,000 yards, if bare, only 11,000,000. The difference of discharge from wooded and unwooded soils is perhaps exaggerated in Col. Torelli's report, but there is no doubt that in very many cases it is great enough to prevent, or to cause, destructive inundations.] Rivers fed by springs and shaded by woods are comparatively uniform in volume, in temperature, and in chemical composition. [Footnote: Dumont gives an interesting extract from the Misopogon of the Emperor Julian, showing that, in the fourth century, the Seine—the level of which now varies to the extent of thirty feet between extreme high and extreme low water mark—was almost wholly exempt from inundations, and flowed with a uniform current through the whole year. "Ego olim eram in hibernis apud curam Lutetiam, [sic] enim Galli Parisiorum oppidum appellant, quae insula est non magna, in fluvio sita, qui eam omni ex parte cingit. Pontes sublicii utrinque ad eam ferunt, raroque fluvius minuitur ac crescit; sed qualis aestate talis esse solet hyeme."—Des Travaux Publics dans leur Rapports avec l'Agriculture, p. 361, note.
As Julian was six years in Gaul, and his principal residence was at Paris, his testimony as to the habitual condition of the Seine, at a period when the provinces where its sources originate were well wooded, is very valuable.] Their banks are little abraded, nor are their courses much obstructed by fallen timber, or by earth and gravel washed down from the highlands. Their channels are subject only to slow and gradual changes, and they carry down to the lakes and the sea no accumulation of sand or silt to fill up their outlets, and, by raising their beds, to force them to spread over the low grounds near their mouth. [Footnote: Forest rivers seldom if ever form large sedimentary deposits at their points of discharge into lakes or larger streams, such accumulations beginning or at least advancing far more rapidly, after the valleys are cleared.]