These inconsiderable hills being passed, the surface again subsides nearly to a plain. The fine and comparatively fertile sand which prevailed to the east of the ranges, is exchanged for a gravel made up of rounded granitic fragments, varying in dimensions from the size of a six-pound shot to finish sand. This great mass of granitic fragments, evidently brought down by the agency of water from the sides and summits of the Andes, slopes gradually from their base, appearing, as far as examinations have extended, to correspond in some measure, in magnitude, to the elevation and extent of that part of the mountains opposite which it is placed. The minute particles derived from the quartzy portions of the primitive aggregates, being least liable to decomposition, have been carried to the greatest distance, {279} and now form the almost unmixed soil of the eastern margin of the great sandy desert. The central portions are of a coarser sand, with which some particles of felspar and mica are intermixed: nearer the mountains, pebbles and boulders become frequent, and at length almost cover the surface of the country. The opinion above advanced, that the great sandy desert has resulted from the wearing down of the mountains, both before and since the retiring of the ocean, should, perhaps, be received with some caution. We have no foundation for the belief, but in the examinations which enabled us to discover that the materials composing both regions are similar in kind; that the granitic soils of the plain are precisely such as would result from the disintegration of the rocks now existing in the mountains; and that the numerous deep ravines and water-worn valleys traversing the mountains in various directions, indicate the change here supposed to have happened. It is probable many parts of this extensive desert may differ from that traversed by the Platte, in having the surface more or less covered with horizontal strata of sandstone and conglomerate, instead of loose sand and pebbles. Indeed, there are many appearances indicating that a formation of this kind formerly extended down the Platte much farther than at present. From the minute account given in the narrative of the expedition, of the particular features of this region, it will be perceived that its eastern portions bear a manifest resemblance to the deserts of Siberia. The soils, and I believe the rocks, wherever any occur, are saline: plants allied to chenopodium and salsola are peculiarly abundant, as are the astragali and other herbaceous leguminæ; while trees and forests are almost unknown.