vince of Chiquitos, porphyritic or granitoidal gneisses, which are superimposed on granite, and form the support of fine grained gneisses, or of mica-slates containing garnets and gnenatites; at Monte-Video, and at Maldonado, there are blackish and finely laminated gneisses; and at Tandil, according to the determination of M. Cordier, there are tabular varieties of petrosilex.

In Brazil, and in the east of the province of Chiquitos, the gneiss everywhere supports transition clay-slates. But when the latter are awanting, the gneiss is succeeded by formations of a much more modern date; for at Conception, at San Ignacio, and at Santa Anna de Chiquitos, portions of the Patagonian territory formation repose on the gneiss. M. Pissio has pointed out tertiary deposits, resembling the molasse of Europe, reposing on the gneiss of the environs of Bahia. At Monte-Video, and in the Pampas, the gneiss is surrounded by the Pampian tertiary formation; and, lastly, at Chiquitos, it is covered by the recent alluvia.

Silurian System.-The most ancient beds which M. d'Orbigny has found superimposed on the rocks of a decidedly crystalline character in South America, present, wherever he has seen them, a very uniform composition. They are, in their lower portions, coarse clay-slates (des phyllades schistoides), having a blue colour, and often containing chiastolite; and they pass in their middle portions into fine grained clayslates (des phyllades satinés), having a rose colour. These two series of beds, which are the most developed, possessing frequently a thickness of several hundred yards, do not contain any traces of organized beings. Above these there are sandstone flags (des phyllades grésiformes), or very micaceous slaty sandstones (grès phylladifères très micacés), whose thickness is upwards of 160 feet.

In these last mentioned beds, M. d'Orbigny has collected fossils, which, however, are rare, and which belong to the genera Cruziana, Orthis, Lingula, Calymene, Asaphus, and Graptolithus. Of ten species of these genera, eight are extremely analogous to the species of the silurian strata of Europe, and three, viz., Calymene macrophthalma, Cruziana rugosa, and Graptolithus dentatus, are even identical with

them. We may say, therefore, that these fossils, as a whole, have the same aspect, the same facies, as those of the Silurian formations of Europe. Here is an identical zoological physiognomy transported thousands of leagues. The rocks also present mineralogically many analogies to those of the Silurian formations of Europe. This double relation, conjoined with the position these formations occupy below all the other fossiliferous deposits of South America, has naturally led M. d'Orbigny to refer them to the Silurian system, as established by Mr Murchison; and it is probable that they are at least very nearly connected with it.

These Silurian formations of South America occupy spaces of considerable magnitude, and occur at points very remote from one another. They present themselves along nearly the whole eastern border of the Bolivian plateau, forming a band which follows the Andes, properly so called, or the eastern Cordillera, and is parallel to the granitic rocks, from Sorata to Illimani, a distance of more than 300 English miles. To the east of the eastern Cordillera they are still more developed, and form a band of nearly 40 miles in breadth, and upwards of 600 miles in length, included between the plains of Santa Cruz de la Sierra on the east, and the 72d degree on the west. They thus form an immense band on the east, as well as on the west of the eastern chain, and which runs from N.W. to S.E., but is much more developed to the east than to the west of the chain. In the region included between the Andes and Brazil, we again find the Silurian series in the south of the province of Chiquitos, near Tapera, near San-Juan, to the north of the Sierra of Santiago, and to the south of that of Sunsas. They there constitute a band running from E.S,E. to W.N.W., and having a length of upwards of 150 miles. There, as in the Andes, they present, at their lower portion, a blue coarse slate, supporting fine rose-coloured clay-slates, on which repose yellowish slates. M. d'Orbigny was, however, unable to find any trace of organized bodies in these beds, of which the first has a thickness of at least 600 feet, while the others are not more than 50 or 60 feet thick.

In Bolivia, the Silurian strata possess an interest of a very

direct kind, inasmuch as they contain the richest of the Bolivian gold mines, and also some mines of silver. Wherever gold is met with in situ, it occurs in veins of milky quartz, which traverse the lower beds, viz. the coarse clay-slates. It is in such a geological position, that the ore is mined on the slopes of Illimani, at Oruro, at Potosi, &c. If we consider that all the localities where gold is obtained by washing, are situated in valleys where the clay-slates have been much dislocated and denuded, as, for example, at the Rio de la Paz, at Tipoani, at the Rio de Suri, at the Rio de Choquecamata, &c., we must naturally conclude that the ore is derived from the same slates.

Devonian System.-Wherever M. d'Orbigny observed the Silurian series, it was succeeded by an enormous mass of hard quartzose sandstones or quartzites, which, from their position and fossils, he regards as the representative of the Devonian formation, or the old red sandstone. This extensively distributed system is composed of whitish or yellowish compact quartzose sandstones, without traces of fossils, passing, in their lower portions, into blackish or ferruginous, very micaceous, slaty sandstones, and only then containing the remains of organized bodies, sometimes in large beds, at other times disseminated through the strata of rock. These rocks are almost always superimposed on the Silurian series, and most frequently in a conformable manner. They are succeeded, but unconformably, by carboniferous strata, which are characterized, in a particular manner, by the fossils they contain.

This great quartzose deposit presents itself over as large spaces as the Silurian system, which it everywhere accompanies; and it is distributed in the same way. On each side of the band of Silurian rocks of the eastern chain of the Andes, it forms, for about 450 miles, another great parallel band, independently of the detached portions scattered through the interior of the Silurian band. There is also a great development of these quartzose sandstones on the Silurian formation of the eastern part of the province of Chiquitos.

Apart from his own personal observations, M. d'Orbigny has learned that these same quartzose formations abound in Brazil, in the chain of Parecys, in that of Diamantino, to the west of Motogrosso, and in those chains which are to the east

of Cuyaba, mountains which follow the same direction as those of Chiquitos, and which, according to M. d'Orbigny, belong to the same system. Perhaps, he adds, the same rocks are to be found more to the east, in the province of Minas-Geraës, a supposition which seems to be confirmed by the meritorious labours' of M. Pissio, which were lately laid before the Academy of Sciences.

MM. Humboldt and Eschwege have, for a long time, fixed the attention of observers on the rocks of stratified quartz which occupy vast areas in South America, to the south of the Equator, in Peru as well as in Brazil.* These stratified quartzes were divided among various primitive and secondary formations; perhaps a judicious application of the principle of metamorphism, such as was lately suggested by the memoir of M. Pissio,+ may enable us to include all under one and the same formation, the Devonian formation of M. d'Orbigny. The exact determination of the age of the quartzose sandstones of Bolivia is thus an important question for the geology of South America, and even, we may say, of the greater part of the Southern Hemisphere, if, as we may presume, the quartzose sandstones of Table Mountain, near the Cape of Good Hope, belong to the same formation.

In the Devonian formation of the province of Chiquitos, M. d'Orbigny did not observe a single trace of fossils; whereas he observed fossils several times in the lower parts of the sandstones of the same system in Bolivia, especially at Achacaché, near the lake of Titicaca, in the environs of Cochabamba, near Totora, and at Challuani, in the province of Mizque, in the provinces of Tocopaya and Yamparaes, in the department of Chuquisaca. These fossils, which belong to the genera Spirifer, Orthis, and Terebratula, are always in the state of impressions, and occur in widely extending but very thin layers, between the lamina of the rocks. Of seven species of these different genera which M. d'Orbigny brought from Bolivia, four have the greatest resemblance to fossils of the Devonian system of Europe. Some of the others ap

* Humboldt, Essai géognostique sur le gisement des roches dans les deux hémisphères, p. 91, 96.

† Comptes Rendus, vol. xvii., p. 34. Meeting of 3d July 1843.

proach fossils, which, in Europe, occur in the Silurian rocks. Everything, then, leads us to refer this great quartzose deposit to the palæozoic formations. Its connection with the slate formation which it covers, does not allow of its being removed far from it; and as the union of these slates with the Silurian system seems to us reasonable, the reference of the great quartzose deposit to the Devonian system or old red sandstone, appears to us the most judicious that can be made in the present state of our knowledge.

Carboniferous System.-The system of Devonian quartzose sandstones is succeeded in Bolivia, and in some other parts of South America, by another series of beds which M. d'Orbigny refers to the carboniferous system. This new series of beds consists in its lower part of compact grey limestone containing siliceous nodules, analogous to the carboniferous limestone of English authors, and perfectly similar to that of Visé near Liège, and to the limestones of many parts of the British Islands. This limestone has been more particularly noticed in the islands of Quebaya (Lake of Titicaca.) At other points (at Yarbichambi), the lower parts of the same system present compact, yellow, or rose-coloured calcareous sandstones. These beds contain numerous fossils. In the islands of Quebaya, and at Yarbichambi, they are succeeded in conformable stratification by red non-argillaceous quartzose sandstones, which are friable and destitute of fossils. It is from observations made on these two points that M. d'Orbigny has concluded, that he may refer to the carboniferous series all the friable rose-coloured argillaceous sandstones which repose on the Devonian system, and which are inferior to the presumed Triassic variegated clays.

The system of beds, of which we have just indicated the composition, presents itself at a great number of points, and is distributed throughout nearly the whole breadth of the American continent. The Morro of Arica, washed by the waves of the Pacific Ocean, is composed, at its base, of a limestone, which seems to belong to the carboniferous epoch, judging from the impressions of a productus contained in one of the specimens collected by M. d'Orbigny, and from numerous organic remains observed by him at the locality. This