The same dissipation of vegetable matter takes place when a ditch has been dug in clay ground, and {322} the excavated earth thrown up to form a dike on one side. On removal of the dike, the original surface, which no doubt, at the time the dike was formed, consisted of vegetable clay mould similar to the surface around, is always found to be close, heavy, poor clay, containing little or no carbonaceous or vegetable matter. In this case, from the draining effect of the ditch, the original surface under the dike must have been drier than the subsoil of the crowns of the ridges. The difference of depth and richness of vegetable mould, may nearly always be referred to existing causes, such as the original surface (diluvium, or decayed rock), being a combination of earths favourable to vegetation; occupying a genial situation; being favourably placed with regard to moisture, that is, less or more moist, according as the original surface has been clayey or sandy, or open or close bottomed; and is in no way connected with those flood torrents to which we owe the diluvium deposits themselves—tills, sand and gravel, in which we have never found any vegetable matter, excepting in the coaly or mineralized state. Unless in the case of alluvium, or of drift sand, or where surface earth has been rolled down from {323} heights, or been _forced_ by man[62], soil is seldom found to exceed 6 feet in depth, and that only in warm moist situations, propitious to vegetation. In Scotland we never have seen it exceed 3 or 4 feet in depth where its accumulation had not been aided by the above causes. The most common depth is from 6 inches to 2 feet; but, in many of our sterile districts, the surface hardly deserves the name of mould, containing very little vegetable matter, or that matter being unavailable from the presence of tannin.