672 AMERICA [S. AMERICA. to "be absent on the east side of the Andes in Columbia, as also along the coast of Venezuela, In the mountain range south of the Orinoco, hornblende, talcose, and mica schists again appear on a large scale, more especially in the Caratal district, where auriferous veins occur. In Brazil the mi caceous and talcose schists enter into the composition of the Mantiqueira chain and of the uplands to the west; they probably pass beneath the valley of the Panama, since they occur to the west of it, and extend through the pro vinces of Goyaz and Matto Grosso, so as to approach within a moderate distance of the similar strata in Bolivia. They are associated with talc-bearing quartzites, which are famous for the diamonds and auriferous particles they have yielded in the district around Mount Itacolumi. The soil is usually a clay, such as would result from the decomposition of talcose schist; but we shall have occasion to refer to the soil of South America subsequently. Some of these rocks in various parts of the continent have yielded Silurian fossils. Facts are not sufficient yet to warrant the correlation of these strata with those of other countries, or to settle how far they belong to distinct geological periods. The pre valent strike of the rocks is about east and west, but some times the strike approaches to a north and south direction. At any rate, the rocks which overlie them do so uncon- formably, indicating that, prior to the deposition of these newer rocks, land had appeared at least once on areas now constituting part of South America. At this early date the continent was represented by a few islands only; one corresponded with part of Brazil, another with parts of (Venezuela and Columbia; perhaps a third more or less with Peru, Bolivia, Ecuador, and Chili; while a few small islands appeared where now we have the Pampas. These .were the nuclei around which the present land has accumu lated, and already we see faint indications of the existing outline and broad geographic features of the future con tinent. The next group of rocks are always in stratigraphical discordance with those beneath them ; and, in consequence of the highly metamorphosed condition of those on the west side of the continent, it is difficult to correlate them with the rocks of Brazil. In the Andes of Chili they are represented by enormous stratified masses of quartzose porphyries, which there is good reason to believe are metamorphosed argillaceous schists and felspathic sand stones, into which rocks they have been seen gradually to pass. These porphyries not only form the great bulk of the principal chain of the Andes, but also the smaller chain on the west, the interval between them being formed by the longitudinal valley of Chili. Further west they rest on syenitic rocks, beyond which come the older rocks already noticed. On passing into Bolivia, we find that to the west of the great fault developed there, the beds con sist of micaceous sandstones and dark bituminous schists, which are believed to be the equivalents of the porphyries of Chili. Such formations constitute the west slope of the Andes from Sorata to Illimani, and also form two bands, one stretching from Illimani to Cochabamba, the other between Calamarca and Chayanta. Towards the west they dipbeneath black bituminous and siliceous limestones, which are well developed near Tiahuanaco. Carboniferous strata with seams of coal occur near Pisco and Arequipa in Peru. In Brazil the beds which succeed those previously men tioned are quartzites, rich in mica and magnetic oxide of iron; talcose schists; and crystalline limestones, containing a great deal of talc. These rocks form the highest regions and loftiest peaks in Brazil. Unconformably upon these rest micaceous sandstones and argillaceous schists, which occupy the western part of St Paul province between Ytu and the banks of the Parana. In Mount Arasoyaba and eome other places, a carboniferous limestone succeeds and is overlain by a thick white or yellowish sandstone, siliceoua limestones, and bituminous schists. The limestones occupy most of the area between the Uruguay and the Parana. In the Diamantina district the carboniferous limestone is over lain by red sandstone, which belongs to the succeeding group of beds. The beds in Brazil appear to be fuller than those on the west side of the continent, and represent pro bably both the Devonian and Carboniferous periods. Rocks of this age are also exposed in a narrow band round por tions of the basin of the river Amazon. The strata next in succession are of secondary age. The lowest formation is a red sandstone, which is spread over a very extensive area. It is remarkably well developed in Chili, where, together with more recent beds, it is a marked feature in the crest of the Andes. In the small chain to the west it is associated with conglomerates. In the small chain of Colorado, near Tiahuanaco, a thick conglomerate rests on the older limestones, and supports red sandstones and con glomerates dipping west beneath marls. The red sand stone extends across the province of Carangas, and unin terruptedly over both slopes of the western Cordillera. In the desert of Atacama the red sandstone, with the over lying marls, forms a number of parallel chains directed north and south. Red sandstones and conglomerates form the base of the Cordilleras of Quito; they stretch north into the basins of the Magdalena and Cauca rivers, and east over the basin of the Orinoco. In the province of Bahia, and far away to the north, there is a great development of red sandstone. The age of these sandstones has been variously stated ; their stratigraphical position would indicate a secondary age, and possibly they may be Triassic. After their deposition, and prior to that of the marls, syenitic rocks were introduced amongst the strata in the Andes, causing the red sandstones, as also the older sandstones and schists, to be converted into porphyries. This erup tion was also accompanied by, and probably connected with, the formation of auriferous veins, the elevation of the strata, and the faulting of the rocks. The strike of the strata, as also of the faults, was about 8 to the E. of N., but subsequent movements have modified the direction in places. The red sandstone has a similar strike from Venezuela in the north to Magalhaens Strait on the south, and this favours the idea that all are of the same age. Several of the ranges in the Andes have a corresponding direction. As the sandstone is believed to underlie the basins of the great rivers, it appears that during its deposi tion South America was still represented by a few large islands only. Its elevation gave rise to north and south trending mountains, whereby these scattered portions were connected, and the Andes received their first development. The great features of the continent were then first dis tinctly marked out, and only a few gaps remained to be filled up. The next succeeding period, represented by strata, is characterised by saliferous and gypseous marls, which rest unconformably on the rocks beneath. In Chili they occur in the lower plains, or abut against the western spurs of the Andes ; but they have been largely denuded, so that they now occur in isolated plateaux or basins, and there is a patch capping the lofty Aconcagua. It is stated that in Chili the marls, or at least some of them, are Liassic. In Bolivia, also, they form plateaux between San Andres and the mountains of Tarija. Here they consist of alter nations of greenish marls and wine-coloured beds support ing limestones, with interst ratified beds of gypsum and saliferous clay. There are beds of gypsum and limestone in the Orinoco plains. Maris are associated with marly sandstones in the Gulf of Bahia and in the plains of Reconcavo. These beds are arranged along strikes which are approximately east and west, and their elevation is
apparently connected with the formation of chains running
