309. In some other places of the same mountains, the granite is more abundant. A line from the lake of Thun, along the course of the Aar, and over the mountains to the upper end of Lago Maggiore, crosses a very elevated tract, and passes by the sources of the Rhone, the Rhine, and the Tessino, which last runs into the Po. A good deal of granite is discovered here, in the mountains of Grimsel and St Gothard; but by far the greater part of it is the veined granite, the granite in mass being confined chiefly to the north side of the Grimsel. Both together do not occupy more than one-third of the line, and therefore the latter less than one-sixth. 310. The essay on the mineralogy of the Pyrenees, by the Abbé PALASSO, contains a mineralogical chart of those mountains. From this chart I have found, by computation, that the granite does not occupy one-fifth of the horizontal surface on the north side of the ridge, reckoning from one end of it to the other. Indeed, many great tracts, even of the central parts of the Pyrenees, contain no granite whatsoever; and not a few of the highest mountains consist entirely of calcareous schistus. A large deduction should be made from the fraction 1/5 on account of the substances unknown, which, from the construction of the chart, are often confounded with the granitic tract. 311. I might add other estimations of the same kind, all confessedly rude and imperfect, but still conveying, by means of numbers, a better idea of the limit to which our knowledge approximates, than could be done simply by words; and, on the whole, it would appear, that if we state the proportion of granite to schistus to be that of one to four, we shall certainly do no injustice to the extent of the former. It remains to form a rough estimate from maps, and from the accounts of travellers, of what proportion of the earth's surface consists of primary, and what of secondary rocks. After supplying the want of accurate measurement by what appeared to me the most probable suppositions, I have found, that about 1/18 of the surface of the old continent may be conceived to be occupied by primitive mountains; of which, if we take one-fifth, we have 1/90 for the part of the surface occupied by granite rocks, which differs not greatly from the least of the two limits assigned by Dr Hutton. 312. In estimating the granite of Scotland, Dr Hutton has certainly erred considerably in defect,[160] and Mr Kirwan, who always differs from him, is here nearest the truth; though he is right purely by accident, as the information on which he proceeds is vague and erroneous.