SciELO - Scientific Electronic Library Online

vol.3 issue9Um nível para o lavrador author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand




Related links



Print version ISSN 0006-8705

Bragantia vol.3 no.9 Campinas Sept. 1943 

Rochas-mater da "terra roxa"



M. Gutmans




Physico-chemical and mineralogical studies of the soils of the São Paulo State by the Soil Division of The Agronomical Institute proved the existence of different violet soils in South-Brasil and its origin from basaltic rocks.
The Brasilian denomination "terra roxa" is already many times translated to "red soil", which is inexact, because the colour "roxa" corresponds to "violet" in English.
We must insist on the perfection of Brasilian expression, which gives the shortest and the best characteristic of the true violet soil, derived through the decomposition of basalts and diabases in the São Paulo State. The term "red soil" originated much confusion, because there are in this state many "red soils" of different origins, but the true violet soil is quite unique.
The violet colour of this soil appears very beautifully on the clean fields above the diabase hills at the distance of some hundreds of meters.
In the state of complete dryness the violet soil becomes coffee brown, but never gets a red colour.
The violet soil is the best soil of South America, on the contrary the red soils, which are mostly lateritic, are bad soils.
Some exceptions, do exist, of course.
The basalts produce laterites and other red soils of better qualities.
With the purpose to contribute to the study of the violet and the red soils the present essay was made, describing the basic rocks of South-Brasil, as mother rocks of soils.
The basic monograph of Djalma Guimarães "Magmatic Province of South-Brasil" (5), many times mentioned in the petrographic literature, definitively established the principal types and the mineral-components of triassic basic rocks in South-Brasil.
We have found some varieties of basic rocks, not yet known, but considered important for the question of violet and red soils.
To soil science we consider of importance the discovery of acid melaphyres without labradors, which are essential for all other basaltic rocks, with exception of most basic and extreme types.
These melaphyres have oligoclases and andesines, as its principal components, and are outcropping in many points of the Botucatú sandstone zone, from Franca up to Pirajú in São Paulo State.
The melaphyres of Franca and Pedregulhos appear on a high "plateau" with the orientation NNW.
In this direction the plateau is about 50 km long and in WSW direction about 25 km wide.
The slopes are very dissected, showing numerous big outcrops of melaphyrs, which rarely outcrop on the surface of the plateau, because the plateau is covered by variegated sands and sandstones.
Above all, in isolated patches, appear loose, conglomeratic beds, never above 1 - 2 mt; with decomposed basaltic pebbles, and eolic sands.
These eolic sands are of much later age than the S. Bento sandstone below, which contains the melaphyres, because between the eolic sands above and the latest melaphyres the variegated sandstones and sands were deposited.
These sandstones have argillaceous cement and are very similar to Baurú sandstones in Rio Preto district.
Our melaphyres were not yet analyzed chemically, but the predominance of more acid plagioclases in relation to the diabases and the augite-porphyrites indicates more acid general composition.
We explain it by the separation of more basic part of the primitive basaltic magma in the depth, where the already formed augites and basic plagioclases remained.
Thus, the plutonitic phase of the magma, which originated the melaphyres, must have more basic plagioclases.
The search for these plutonites was directed to the cristalline zone of gneisses and schists, where the magmatic channels were discovered to the depth of several kilometers by the erosion.
In the Bocayuva district of the Paraná State several big outcrops of gabbroid rocks were found.
By microscopical examination we discovered the same absence of dinamometamorphic stresses which is the best characteristics of all triassic basaltic rocks in our region.
In some sections we found some bending of augites and plagioclases, which are common in many diabases and augite-porphyrites. Evidently these bendings happened during the eruption. In no case could these bendings be the result of dinamometamorphism.
The mineralogical composition of these gabbroid rocks is very singular.
They have the same augites which are found in diabases, augite-porphyrites and melaphyres.
The characteristic angle of these augites c ^ Ng is about 42° - 44°26', as determined by D. Guimarães (5) in different samples gathered in the States of S. Paulo, Paraná, Santa Catarina and Rio Grande do Sul.
For the determination of this angle we used the Fedorow stage and we found that in all our rocks the augites form twins, with Nm common for both parts and Ng of one half coinciding with Np of other half.
In such a case c ^ Ng = 45°.
However there are many instances of big variations in optical properties of augites in the same sections.
The biggest variations happen with the angle 2V, which gives the angles 0° - 50° in one section of the same rock.
But these variations happen in all basaltic rocks, from plutonites to extrusives.
The plagioclases, with An47 - An60, give poikilitic texture, being included in bigger crystals of anorthose, with 2V = - 54°.
The olivine in big rounded grains, only slightly altered, is always present.
As accessory minerals are big well formed prisms of apatite, ilmenitic magnetite and brown biotite.
We discovered big outcrops of the same gabbroid e rocks near São Bento do Sapucaí in the cristalline belt.
The only difference is in somewhat slighter basidity of plagioclases, which have An44 - An6o.
By mineralogical composition these rocks are between gabbros and essexites. The next rocks would be shonkinites, which differ by the presence of orthoclase and nepheline, and small amount of plagioclase.
Thus we must introduce a new term, calling this rock bocaiuvite by the place where are the biggest and the most characteristic outcrops.
It seems difficult to put the bocaiuvites, with its sodic tendency, in the basaltic family of rocks, but there are no other plutonic phases in the crystalline belt, which would be nearer to our diabases.
We must have in view the anortosic rims of plagioclases in our augite-porphyrites, which have also sodic tendencies.
The essexites are common in the South-Namib (12), appearing as the central parts of biggest dikes of monchiquites.
Unhappily the petrographic description of these essexites does not give the optical constants of its ortoclase, and we do not know if this ortoclase is sodic. Very probably they are sodic by its genetic relations to monchiquites.
Little can be said about the part of basic rocks in the development of the orography and hidrography of crystalline zones, but in the rest of territory the basic rocks give to the surface its main characteristics.
It is well seen in the São Paulo State. There the basic rocks reached the surface by two systems of faults.
One system runs parallel to the limit of the crystalline belt, changing its direction from East-West in the South of the State to North-South in the North. The biggest diabasic dikes belong to this system.
The other system has its faults directed normally to the first.
These faults show very clearly that the Paraná basin is the result of regional subsidence by step-faults.
The similar step-faults can be observed on Mato-Grosso side of the Paraná river and we have little doubt about the extension of this system to the basin of Paraguay river, more to the West.
The step-faults of bigger system, which run parallel to the limit of crystalline belt, have generally its downthrow on the side contrary to the crystalline formations.
There are some exceptions of this rule.
The Botucatú sandstone blocks are separated by faults, which have its downthrow on the side of older Passa Dois formation, but these faults have always little vertical displacements.
The biggest rivers, Paraná and Paranapanema, flow parallely to biggest faults but the affluents, as Tietê, cross these faults normally.
The eruptions occurred in extreme plains, through faults and fissures. The elimination cf lava was difficult and the channels of eruption were closed by hardened lava after the first outbreak.
The magma of following eruptions penetrated mostly in the São Bento sandstone, forming extensive sills.
Obviously the sills in lower strata must be rare, as the magma had much more resistance there, in view of the pression of overlaying formations.
Almost in all borings of S. Paulo State, which reached in some cases the considerable depth of 1600 - 1800 mts, the diabase was encountered in the whole extension of borings, but no sill of bigger dimensions was met below the S. Bento sandstone.
Mostly the diabase in biggest depths was in form of dikes.
The diabasic intrusions and lavas strengthened the S. Bento sandstone, which resisted the following erosion much better as older formations.
Most of high plateaus in South-Brasil with abrupt sides are formed by S. Bento sand stone with diabasic or silicified layer.
Excepting the predevonian metamorphic belt all narrow valleys with abrupt sides are located in S. Bento sandstone region.
By the different resistance against the erosion is easily explained the broad lower plains of 500 - 550 mt altitude, on Passa Dois, Tubarão and Itararé formations, between high plateaus in the west and the crystalline coastal elevations in the East.
It is generally admitted that all differences of chemical and mineralogical composition of eruptive rocks come from chemical variations of the magma, which produced these rocks.
There are now established as many kinds of magma as magmatic rocks.
The differences of structure are mostly explained by modern thermal diagrams, based on enormous number of experiments.
To apply these diagrams for our rocks we must assume that in the nature the elements which form different combinations known as rocks are kept together during the whole forming process as they are kept in the furnace of the laboratory.
It would be the case of a quite isolated subterranean chamber, with no connections either with lower reservoirs or earth surface.
On the contrary, the basaltic eruptions were the product of an enormous network of aults and fissures, which the magma crossed during the process of crystallisation.
We mentioned already melaphyres, which were the product of the liquid part of the magma, after the separation of more basic crystals.
By the use of the Fedorow stage it was discovered that in most samples of diabase or augite-porphyrites the plagioclases of the same generation and dimensions have different chemical composition.
The differences in the case of plagioclases reach sometimes 10 - 20% An, which is much above the errors of the Fedorow method.
Evidently in diabases, augite-porphyrites and melaphyres we have mechanical mixures.
In other words many times the crystals were separated from the mother solution and mixed with a solution of other composition.
The Fedorow stage permits easily observe that most pyroxenes and plagioclases have rounded edges.
Generally such rounded edges are explained by magmatic reabsorption, and in some instances it is quite true.
The diabasic magma during the eruption was heavily charged with crystals and its liquid part was very near the solid state.
It is proved by no chemical action or very slight chemical action on the sides of transporting channels.
Only rarely the contact zone is several meters thick.
In every such case it is a contact of a big reservoir, where the magma remained till the solidification, or we have an instance of hydrothermal postvolcanic action.
In such a half-solid magma no chemical differentiation was possible.
In biggest sills and dikes of S. Paulo State we collected numerous samples at regular intervals normally to the sides of the intrusions.
In many cases we discovered evident changes in the structure of the rock, and in the mineralogical and chemical composition of the samples of the same intrusion, but the detailed geological investigation always showed that the changes come from different phases of the eruption, and are not the result of differentiation.
Different may be the case of bocaiuvites, which have consolidated as plutonitic masses far below the surface of the earth.
The eruption began in the depth of many thousands of meters. In this depth the magma was more liquid and gaseous and could absorb completely the fragments of crossed rocks.
The deepest strata, crossed by basaltic magma, were glass of acid composition, in half solid state, which readily was absorbed by moving basaltic magma, giving the more acid andesitic augite-porphyrites which in all diabasic dikes occupy the central parts, or cross the previously consolidated diabasic sills.
We admit the existence of many magmas of different chemical composition, but most of these secondary magmas were the products of mechanical mixing during the eruption of one magma, contaminated by one or two glasses in half liquid state; once consolidated as eruptive rocks the magmatic mixtures left no other traces.
The basaltic permotriassic eruptions by its extraordinary development supply the best evidence for the problem of the causes of volcanic action.
The accumulation of glacial-lacustrine and eolic deposits during the permo-triassic time reached a total thickness of thousands of meters and was the cause of a subsidence of corresponding dimensions.
There was one important factor which transformed this subsidence in a catastrophe of exceptional grandeur.
Generally the accumulation occurs in geo-sinclinals between parallel ridges, where the pressed magma reaches the surface through channels in the form of volcans, which are the safety valves of magmatic kettle.
There were no safety valves in the enormous Gondwana Continent, inaccessible for tectonic disturbances during the long permotriassic period, with exception of some little parts at its borders.
The loading of this continent occurred in very different manner, when compared to deltaic or sinclinal sedimentation, which generally have big changes of thickness on the belts of several kilometers broad.
The water-currents, maritime or continental, are steady.
On the contrary, the eolic deposits, which were the last load on the Gondwana Continent, covered it in large overlapping zones, with many changes of direction of wind during every year.
There were no mountains which could accumulate blown sands in belts. Such belts by subsidence and consequent lateral movement of magma, as estipulated by isostatic theory, would gradually reestablish the balance of crustal forces.
The magma found the escape in vertical faults and fissures almost simultaneously over the shole continent.
We had not sinking and rising blocks of isostatic mutual displacement. The sinking and rising movements happened in the whole continent. The sinking part were the Gondwana sediments and the rising the basic magmas. At first appeared melaphyric lavas, leaving behind heavier magma. There was no lacking of volcanic gases, which opened the fissures for the oncoming Java.
When the eruptions of gas ceased and the expanding lavas closed the issues, as it was explained before, the subsidence of the continent continued some time more by the impulse taken.
In this last period of subsidence the lighter sediments sank deeper in he avier magma and the diabasic sills were intruded.
The difference of load on continental blocks explains the beginning of continental subsidence but cannot explain all phenomena correlated.
The vertical displacement of magma instead of lateral, explains naturally the positive gravity anomalies over deltas, which the isostatic theory cannot explain.
The oscillations of the level of the sea are well understood as the result of the sinking of lighter mass in heavier magma by inertia and later regaining of balance.
Also many if not most part of volcanic outbreaks have its cause in the gravity pressure of the crust of the earth on the magma.
The magnetometric crossection of S. Paulo and Mato Grosso States (18) shows the gradual increase of the vertical component from the crystalline belts of both states toward the axis of Paraná basin.
The subsidence near this axis was the greatest and consequently the biggest are there basic intrusions.
The base of S. Bento sandstone comes there down to the 80 mts level above the sea.
In the North of the S. Paulo State the total thickness of eruptive sills is about 125-150 mts and the base of the sandstone rises to the 700-750 mts level.
In the North-Uruguay the borings discovered a diabasic sill of 360 mts thickness and here the base of the S. Bento sandstone goes down to 500 mts below the sea-level.
The same gradual descending of the Gondwana System toward the South is observed in South Africa.
The Gondwana System has the maximum thickness of 27800 feet in the Cape Provinces simultaneously with maximum total thickness of diabasic sills and lavas of 4500 mts, and the lowest situation of Dwyka Tillite.
In Central Transvaal the Gondwana System is olny 2430 feet thick according to A.du-Toit and the base of Dwyka Tillite rises the 1400 mts level.
Farther to the North in the Nyasaland border the Gondwana System is 18000 feet thick and we have there a group of lavas up to 4500 feet thick.



Texto completo disponível apenas em PDF.

Full text available only in PDF format.




O autor deixa aquí consignados seus agradecimentos ao sr. J. E. de Paiva Neto pelas sugestões apresentadas e pela crítica do original.



1.  Paiva Neto, J. E. de. A "fração argila" dos solos do Estado de São Paulo e seu estudo roentgenográfico. Bragantia 2:355-432. Fig. 1-4. Chapas 1-79. 1942.         [ Links ]

2.  Paiva Neto, J. E. de. "Perda ao rubro" em análises sumárias de terra. Bragantia 1:433-440. 1941.         [ Links ]

3.  Walther, K. Contribución al conocimiento de Ias rocas "basálticas" de la formación de Gondwana em la América del Sud. Fig. 1-4. Lâminas 1-6. Montevideo. Imprensa Nacional. 1927.         [ Links ]

4.  Cunha, J. da, e João Bruno Lobo. As rochas nefelínicas como matéria prima para a fabricação do vidro de garrafa. Mineração e Metalurgia. 2:200-204. 1937.         [ Links ]

5.  Guimarães, Djalma. A província magmática do Brasil Meridional. Publ. Serviço Geológico do Estado de Minas Gerais. Belo Horizonte, pág. 1-64, 1933.         [ Links ]

6.  Johannsen, Albert. Em A descriptive petrography of the igneous rocks, vol. 3, pág. 1-360. Fig. 1-178. The University of Chicago Press, Chicago, Illinois, 1937.         [ Links ]

7.  Paiva, Glycon de. Condições dos jazigos dos diabásios Myrmekiticos. Boi. Serviço Geológico e Mineralógico do Brasil 45:1-6. 1930.         [ Links ]

8.  Wagner, P. Differenüaüonsercheinungen in Diabasen des Ostthueringischen Hauptsattels. Mineralogische und Petrographische Mitteilungen. 50: 107-180, fig. 1-27, diagr, 1-6, Leipzig, 1938.         [ Links ]

9.  Branner, John C. Em Geologia Elementar, 2.ª ed., pág. 1-396, fig. 1-174, Rio de Janeiro, 1915.         [ Links ]

10.  Winchell, Alexander N, Em Elements of Optical Mineralogy. 3.ª ed., parte 2, pág. 1-459, fig. 1-361. New York, 1933.         [ Links ]

11.  Walther, K. Erupticos afro-brasileño-argentinos de edad permo-triasoeojurásica. La supuesta uniformidad del magma «melafídico» y sus relaciones com la parentela alcalinítica. Sep. Bol. Instituto Geológico Montevideo (Uruguay) 24:117-167. Lâm. I-II. Fig.a. Montevideo, 1938.         [ Links ]

12.  Kaiser, E. Em Die Diamantenwueste Suedwest-Afrikas, vol. 1, pág. 1-321, fig. 1-58, Berlin, 1926.         [ Links ]

13.  Leinz, V. Ein Beitrag zur Kenntnis der Gesteine von Angola. Zentralblatt fur Mineralogie und Geologie 1935:97-102. 1935.         [ Links ]

14.  Monta, V. e H. O. Donnell. Carte géologique de l'Angola. Lisboa, 1933.         [ Links ]

15.  Bodenbender, Guillermo. Triásico y terciario en la Falda Oriental. Bol. de la Academia Nacional de Ciencias (Cordoba) 31:73-139. 1929.         [ Links ]

16.  Derwin, Ch. R. Em Geological observations on coral reefs, volcanic islands and on South America, being the geology of the voyage of the Beagle. London, 1846.         [ Links ]

17.  Windhausen, Anselmo. Em Geologia Argentina, vol. 2, pág. 1-645, Lam. 1-58, fig. 1-214, Buenos Aires, 1931.         [ Links ]

18.  Gutmans, Marger e Paulo Vageler. Contribuição para o conhecimento das anomalias magnetométricas do Brasil Central. Bragantia 1:441-451, perf. I-II, 1941.         [ Links ]

19.  Scorza, E. P. Eruptivos básicos do Piauí. Mineração e Metalurgia 5:261-268. 1941.         [ Links ]

20.  Wegener, A. Em Die Entstehung der Kontinente und Ozeane, pág. 1-242, Braunschweig, 5.ª edição, 1936.         [ Links ]

21.  Toit, Alex L. Em The geology of South Africa, pág. 1-539, Oliver & Boyd, 2.ª edição, 1939.         [ Links ]

22.  Machatschek, F. Em Das Relief der Erde, vol. 1, pág. 1-545. 1938.         [ Links ]

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License