I had already been struck by what little I had seen of the plateau region of the province of San Paulo. Beneath the superficial crust of vegetable soil, the plateau appears to be formed of more or less red arenaceous deposits, such as would result from the erosion and decomposition of the gneiss or granite which is the only rock I had seen in the country. In the valley of the Parahyba, the connection was unmistakable. Every section in the valley showed thick beds of the same coarse-grained, red arenaceous deposits, and on the slopes the same material lay at the base of whatever masses of granite we approached. But what especially struck me were the forms and appearance of the mountains on either hand, if that designation could properly be given to them. I saw nothing that would elsewhere be called a mountain range. The outlines were in most places rounded and covered with vegetation, but at intervals occurred steep conical masses, of the same general type as the sugar-loaf peaks surrounding the Bay of Rio Janeiro. However steep, the rocks nowhere showed angular peaks or edges, these being always more or less rounded.
It would be rash to generalize from the partial observations of a passing traveller; but the broad outlines of the geology of Brazil, or, at least, of the eastern provinces, have now been well traced,[44] and some general conclusions may safely be drawn. It is true that large districts of the interior have been but partially explored, and remain blanks on the geological map; but the eastern half of Brazil is undoubtedly ancient land; presenting no trace of secondary strata except in small detached areas near the coast, and where more recent tertiary deposits are to be found only in a portion of the great valley of the Amazons. A mountain range, having various local designations, but which may best be called the Serra da Mantiqueira, extends from the neighbourhood of San Paulo to the lower course of the Rio San Francisco, for a distance of twelve hundred miles, and this is mainly composed of gneiss, sometimes passing into true granite, syenite, or mica schist; and the same may be said of the Serra do Mar, a less considerable range lying between the main chain and the coast. The southern limits of the Serra do Mar do not appear to be well-defined, but we may estimate its length at from five to six hundred miles. The other mountain systems of the empire are less well known; but I believe that the ranges dividing the province of Minas Geraes from Goyaz, and the so-called Cordillera Grande of the province of Goyaz, lying between the two main branches of the great river Tocantins, are largely formed of ancient sedimentary rocks of the Laurentian and Huronian groups.
DISINTEGRATION OF GRANITE.
The granite of the Serra da Mantiqueira and Serra do Mar is coarse-grained, with large crystals of felspar, and is therefore much exposed to disintegration. So far as I know, the vast masses of detritus forming the plateaux of this region show no other materials than such as would be produced by the disintegration of the crystalline rocks, and there is strong reason to believe that these have never been overlaid by sedimentary deposits.
Let us now consider what must have been the past history of a region formed of such materials, exposed, during a large part of the past history of the earth, to the action of the elements. In such an inquiry one of the chief points for consideration is the amount of rainfall. The direct effect, both mechanical and chemical, of rain falling on a rock surface is perhaps not the most important. Still more essential is its action in removing the disintegrated matter, and thereby exposing a fresh surface to renewed action. The difference in the absolute result due to abundant or deficient rainfall would be found, if we could calculate it accurately, to be enormous. In a nearly rainless country, such as Egypt or Peru, we see a slope covered with débris, and are apt to conclude that the rock is being rapidly disintegrated; but, in truth, what we see is the work of many, perhaps many hundred, centuries, which remains in situ because there is no agency to remove it. In a land of heavy rainfall the débris is speedily carried to lower levels, and the work of destruction is constantly renewed.
We have scarcely any observations of rainfall in the mountain districts of Brazil. The only reliable return that I have seen is that of one year’s rainfall at Gongo Seco, in Goyaz, which amounted to more than a hundred and thirty inches; but we may safely conclude that it is everywhere very great. It is also important to note that if, as most geologists now believe, the Atlantic valley has existed since an early period of the earth’s history, Eastern Brazil must always have been a land of heavy rainfall. A great mountain range on the eastern side of the continent might have created a desert region in the interior, but would have received in the past as much aqueous precipitation as it does at the present time.
We have, therefore, to consider what must have been the ancient condition of a region subjected throughout vast periods of geological time to the utmost force of disintegrating agencies applied to a rock very liable to yield to them, and where, without reckoning the large proportion which must have been carried by rivers to the sea, we see such vast deposits of the disintegrated materials formed out of the same matrix. To my mind the conclusion is irresistible that ancient Brazil was one of the greatest mountain regions of the earth, and that its summits may very probably have exceeded in height any now existing in the world. What we now behold are the ruins of the ancient mountains, and the singular conical peaks are, as Liais has explained, the remains of some harder masses of metamorphic gneiss, of which the strata were tilted at a high angle. As the same writer has remarked, although the crystalline rocks are for the most part easily disintegrated, some portions are formed of much more resisting materials, and these have to some extent survived the incessant action of destructive forces.
RUINS OF THE ANCIENT MOUNTAINS.
We are far from possessing the materials for a rational estimate of the probable extent and elevation of the ancient mountain ranges of Brazil. In the first place, we have a plateau region occupying a large part of the upper basin of the Paranà, with an area of fully 100,000 square miles, covered with detritus to an unknown, but certainly considerable, depth. In addition to this, it cannot be doubted that the finer constituents carried down by that river, and its tributary the Paraguay, from the same original home, have largely contributed to the formation of the Argentine pampas and Paraguay, including the northern portion of the Gran Chaco. Borings and chemical analysis of the soil may hereafter give us reliable data; but in the mean time we may safely reckon that an area of 200,000 square miles has been mainly formed from the materials derived from the ancient mountains whose importance I endeavour to point out. In addition to all this, we should further reckon the soluble matter and fine silt carried to the ocean during the long course of geological history, and take into account that the same great mountain region also furnished materials to streams which flowed northward and eastward.
In attempting to speculate on the past history of this region it is important to remark that, so far as evidence is available, there is reason to believe that Brazil has undergone less considerable changes of level than most other parts of the earth’s surface. Even if we go back to the period of the earlier secondary rocks, there is no evidence to show that movements of elevation or depression have exceeded a few hundred feet.