Upon the whole, we shall not greatly err if we assume that, for a period of not less than two thousand years, the walls of the basin of the Po--the Italian slope of the Alps, and the northern and north-eastern declivities of the Apennines--have annually sent down into the lakes, the plains, and the Adriatic, not less than 375,000,000 cubic yards of earth and disintegrated rock. We have, then, an aggregate of 750,000,000,000 cubic yards of such material, which, allowing to the mountain surface in question an area of 50,000,000,000 square yards, would cover the whole to the depth of fifteen yards. [Footnote: The total superficies of the basin of the Po, down to Ponte Lagoscuro [Ferrara]--a point where it has received all its affluents--is 6,938,200 hectares, that is, 4,105,600 in mountain lands, 2,832,600 in plain lands.--Dumont, Travaux Publics, etc., p. 272. These latter two quantities are equal respectively to 10,145,348, and 6,999,638 acres, or 15,852 and 10,937 square miles.] There are very large portions of this area, where, as we know from ancient remains--roads, bridges, and the like--from other direct testimony, and from geological considerations, very little degradation has taken place within twenty centuries, and hence the quantity to be assigned to localities where the destructive causes have been most active is increased in proportion. If this vast mass of pulverized rock and earth were restored to the localities from which it was derived, it certainly would not obliterate valleys and gorges hollowed out by great geological causes, but it would reduce the length and diminish the depth of ravines of later formation, modify the inclination of their walls, reclothe with earth many bare mountain ridges, essentially change the line of junction between plain and mountain, and carry back a long reach of the Adriatic coast many miles to the west. [Footnote: I do not use these quantities as factors the value of which is precisely ascertained; nor, for the purposes of the present argument, is quantitative exactness important. I employ numerical statements simply as a means of aiding the imagination to form a general and certainly not extravagant idea of the extent of geographical revolutions which man has done much to accelerate, if not, strictly speaking, to produce. There is an old proverb, Dolus latet in generalibus, and Arthur Young in not the only public economist who has warned his readers against the deceitfulness of round numbers. I think, on the contrary, that vastly more error has been produced by the affectation of precision in cases where precision is impossible.