RECORD: Darwin, C. R. Geological diary: Provinicia do Rio de Janeiro. (4-6.1832) CUL-DAR32.51-60 Transcribed by Kees Rookmaaker, edited by John van Wyhe (Darwin Online,

REVISION HISTORY: Transcribed from the microfilm by Kees Rookmaaker, corrections and editing by John van Wyhe 5.2010, corrected against the manuscript by van Wyhe 7.2010, corrections by Gordon Chancellor. RN5

NOTE: This document, part of the largest scientific document composed by Darwin during the voyage of the Beagle, is written mostly in ink. Marginal notes are here integrated into the text.

Editorial symbols used in the transcription:
[some text] 'some text' is an editorial insertion
[some text] 'some text' is the conjectured reading of an ambiguous word or passage
[some text] 'some text' is a description of a word or passage that cannot be transcribed
< > word(s) destroyed
<some text> 'some text' is a description of a destroyed word or passage
Text in small red font is a hyperlink or notes added by the editors.

Reproduced with the permission of the Syndics of Cambridge University Library and William Huxley Darwin.

See the introduction to the Geological Diary by Gordon Chancellor.



1832 April

Provincia do Rio de Janeiro


Whilst riding to the Rio Macaè. I passed through an extensive tract of country lying NE of Janeiro & collected the following scanty geological notes. —

The formation is entirely either granite or gneiss: it generally contains much feldspar becoming porphyritic from large lamellar crystals: Near the Rio St. Joao x 1/2 way Between this & the R. Macae, the rock gneiss likewise contains crystals of Hornblende. in one place I found a bed. of that rock. — Near to Rio the granite gneiss was crossed by veins of a finer kind, & in many places by large ones of quartz. — The hills assumed very various forms being either rounded & massive or peaked. — generally there was an apparent entire absence of stratification or cleavage. —

Where the Hornblende rock occurred the formation dipped to the South. & to my surprise proceeding a few miles. apparently it dipped in a contrary direction the North. — In the maps a chain of hills is figured as terminating here. — can this be an anticlinal line. — The gneiss-granite is in many decomposed in a like manner as at Bahia; the soil is generally sandy; but in many places consists of a bright red clay, which is very fertile. — There was on our whole route a most curious absence of pebbles. — occasionally the ground was strewed with fragments of

51 verso

(a) In the interior, such quartz veins occasionally contain gold. — Mr Aston1 has a fine specimen. containing a considerable quantity.

(b) Prince Maximilian (P 144)2 mentions near the Rio Itapemirim the occurrence of beds of ferruginous sandstones. — and that the coast is generally composed of cliffs of red or yellow Lithomarge: which evidently is the decomposed (or rather uncrystallized) gneiss. — as at Bahia. —

(c) (June) The Lagoa (near Botofago) is another instance of this; its shore like the interior ones is characterized by the pureness of the sand. — Annually a communication (V 64 gen: observ:) is artificially made to the sea. — but which is soon again closed up. — The valley of Botofago from its flatness & from the hard mud injected into crags in the gneiss. has evidently been leveled by water. — Having one sufficient cause, it is unphilosophical to theorize on another. — But have not reasons been advanced to show the Atlantic is falling? Some geological facts would tend to render it more probable. —

In geology of Brazil discuss theorize extreme frequency of dike in granitic countries.
Ask Lyell in [illeg] R.N. p. 563
V. Sydney C. of Good Hope papers

(a) Vide P 36 in gen: journal4

M. Lesson5 talks of St. Catherine & coast all granitic

1 Arthur Ingram Aston (1798-1859), Secretary of Legation at Rio de Janeiro 1826-1833.

2 Maximilian 1820.

3 Red notebook, p. 56.

4 Journal of researches, p. 36.

5 Duperrey 1826-30.



1832 April

Provincia do Rio de Janeiro


quartz, which owe their origin to the veins of that rock. —


I saw a piece of sandstone containing marine shells lying on the beach. — There may therefore be some Tertiary formation like that at Bahia. — (b)

On looking at a map. a chain of lakes & swamps may be seen running parallel. to the line of the shore. — The intervening tract consists of white sand & broken shells being chiefly bivalves. — It appears to be gaining in width, (c) & as the beach is very steep, a gale of wind must throw up sand out of reach of the ordinary tides: Anyhow it is clear that the process has gone on to a great extent in places more than a league from the shore, & where the size of the trees. attest how long things have remained their present sate, yet the soil consists of sand & the same broken shell, which are seen close to the sea. — The lagoons are shallow, & are either salt, brackish. or fresh-water. — The salt ones produce a most extraordinary abundance of a bivalve. resembling a Mactia, & on the Stones Balani were adhering. — The fresh water likewise produced in like numbers of Limnaeae (a), & even in to this one the sea, periodically flowed. — So that a most trifling change. would determine. whether one of

52 verso

(a) After observing immediate neighbourhead of the city. I am of the same opinion. — The ravines at base of Corcovado are the most abrupt I have seen; but I suspect these were formed at some former period (when the hills of decomposed gneiss assumed their present form?) Old decaying trees. entwining with their roots blocks of gneiss stand in situations, which they could not do if the water had great violence. — These narrow vallies must have been formed. when the hills were not clothed as at present. by a thick forest which both breaks the rain & protects the soil. —

In Freycinets Voyage Vol I M. Gandichand1 doubts, whether the decomposing granite rocks ever were in state of perfection.
General result from his & Eschwege observation is that gneiss is generally common. — few instances of greenstone, laminated porphyry & Home stone, being unusual exceptions. Granitic formations appearing to extend so far as matty gneiss.

Granitic ruins discussed. RN 882

Daubuisson Vol. II p. 47-51 (& p. 72)3
decomposition of Granite

Fragments in gneiss p. 75

Gold being found chiefly on surface of granites agrees with metamorphic action, surface at bottom of sea: vapours condensed

1 Gaudichaud 1826.

2 Red notebook, p. 56.

3 Aubuisson de Voisins 1819.



1832 April

Provincia do Rio de Janeiro


these lagoons should be filled with salt or fresh water. — Whilst passing through leagues of this country. it most forcibly struck me how small a revolution would produce strata. resembling the Tertiary ones. — Extensive tracts of sand. with shells ready to be cemented, others of a clayey mud; large masses of trees & luxuriant vegetation which might form lignites & probably near to the mountains beds of pebbles. — Again how certainly alleviations of salt & fresh water deposits might be expected & as in case of the Limnaea a mixture would not be improbable: — Also a small change of level would immerse many terrestrial animals either in fresh or salt water deposits. —

In De la Beches geology1 a question is discussed. whether the in the Tropics, the degrading effects of rain are greater than in a more temperate zone? — that is. what ratio. the protecting influence of the vegetation bears to the greater increased one of the rain? As far as I am enabled to judge (a) I differ from Mr D.L. Beche & think that the alteration produced on the surface of the ground by the rain is even not greater than in England. — Tropical regions may be divided into Savannahs & forests, I speak merely of the latter: my opportunities for observing were in the newly cleared spaces, & here the hills were covered with mould, & had a rounded

1 De la Beche 1831.

53 verso

(a) These large crystals sometimes contain garnets & small plates of mica, therefore may be said to be doubly porphyritic.! 542



1832 April

Provincia do Rio de Janeiro


& unbroken outline & the valleys were singularly flat. — I could perceive no signs of violence; indeed the entire absence of pebbles goes towards proving the same fact. — At Bahia I came to the same conclusion. — But at St. Jago. where from the barrenness of the ground there is little or no vegetation, the surface evidently was much furrowed.

May & June

In the immediate environs of the city the geology is uninteresting. 468 ... 470 — The prevailing rock is gneiss, containing large crystals 541 (even 3 or 4 inches long) of feldspar. (divide into two Hemitropes?) (a) & garnets. 603 ... 605 — Mica generally dark coloured & not very abundant. — The slaty cleavage is only perceptible in hand specimens where the mica & crystals of feldspar are arrayed in planes. — When seen in the rock it appears compact. — The hills are of a rounded precipitous shape & occasionally much peaked & steep. — Their bare sides, being coated with lichens present the same tint, which heath does in England. & thus forms a beautiful contrast to the luxuriant forest at the base. — This rock. seems to have a tendency to break at right angles. to the plane of cleavage. Hence enormous concentric layers fall off & the angle of cleavage dip being high it results that the hills assume the rounded conical form. — The sugar-loaf

54 verso


(a) That is Quartz with Mica

page written in pencil.



1832 May, June

Rio de Janeiro


on its SW side is a good example of this fact. — I cannot attribute this appearance to common decomposition; then the layers would be at parallel to the surface, which is not the case as may be seen, where the direction of the beds off of the hills. is not the same. — How also can could it be it explained, that a regular layer at the depth of some yards within the solid rock should undergo decomposition, so that this great exfoliation should take place? — These fallen blocks are of great size, & being covered with succulent plants & creepers they form in miniature beautiful scenes. — The porphyritic gneiss generally rests on Mica slate. 467. 539 540 — this does not contain much mica. but some garnets: it is of a grey colour & not hard. — In most cases junction of two rocks will be defined 523. 524 543. 556 557. 558 559. 560 561, in others it is gradual; they occasionally alternate. — in one instance I found a bed of gneiss in the mica slate, only 6 inches thick & abounding with garnets & mica. — It would appear that the general inferiority of the slate is not a common occurrence (V. Daubisson). — In quantity it bears but a very small proportion to the gneiss. — Both these are frequently crossed by veins of quartz. Granite & Mica slate. (a) latter generally very siliceous. — in some cases. veins which appears quartz are in reality granitic 628. 562 the Feldspar & Mica more frequently crystallizing

55 verso

(a) In Villegagnon the cleavage of Mica slate dipped to SE at an angle of 41° — (reg NE)

(b) Spix. Vol I P 290.1 near the borders of the Capita. de St Paulo. Gneiss has a direction in 3d hour of German compass. — (NE ?)

Rio de Janeiro — Geological transactions Mr Caldcleugh2 some curious facts. —

[sketch] NE E SE
considerable angle

1 Spix 1824.

2 Caldcleugh 1836.

[sketch] NE E SE considerable angle] in pencil.



1832 May. June

Rio de Janeiro


on the sides. — In this formation the strike of the cleavage & beds is the same. it varies one point on each side of NE & SW (b) dipping to SE. — but in few places it is exactly the same: in one place it even dipped due E in another a bed of quartz dipped to S by E. —

The general range of hills is the same but as might be expected each particular point of elevation is a little thrown out. — From the Corcovado the direction was NE by E 1/2 E & SW by W 1/2 W. — (c) The dip of bedding and cleavage was often the same (a) & the angle (when I could observe) it between 50 & 60 degrees. —

Ascending the Corcovado by the Aqueduct, I observed a fr an oblong pap (about 10 yards) 495 ... 500 of rock composed of glossy feldspar, quartz, garnets mica in small black plates, with a tough fracture & no cleavage in general appearance resembling greenstone; it was surrounded on all sides by decomposing gneiss. cleaving in usual direction. — The long axis of pap was NW by W & SE by E. so entirely differing from the gneiss. — I am at a loss to understand this, but must presume it to be a sort of dyke exposed on the side of the hill. —

On the road to the Gavia, I found layered blocks 526 ... 528 with similar fracture. consisting of glassy feldspar. mica in small plate & I think perhaps Hornblende. — I have no doubt hidden in the forest. the rock was not far off. —

56 verso

(a) The rock was divided into spheroidal masses on by an angular cleavage

(b) decomposing into concentric layers. —

(c) March 1833 Upon thinking on this subject. — I am doubtful the greenstone is different from what is found in the dykes; & Spix mentions greenstone boulders on the road to santa Cruz. Vol I. P 272 I well remember the boulders ver were very round & of a large size (4 feet diameter?) of & generally of the same sort. — Mr Fox.1 found boulders of greenstone in the islands of St Sebastian & St Catherines, (though not on the main-land.) & at Port Alagre. — (1479.80.81). 1477. 1478 In St. Sebastian there was a dyke of greenstone

1 Henry Stephen Fox (1791-1846), British diplomat, Minister Plenipotentiary, Buenos Aires, 1831-1832; Rio de Janeiro, 1833-1836. These notes were later published in Darwin 1842, p. 427.



1832 May, June

Rio de Janeiro


I have already said the porphyritic gneiss is by far the most abundant. — At Tijeuka the hills are not so much rounded & bare. 576 ... 578 — & here there is a considerable quantity of a moderately fine grained Granite: I observed this in only one other place, excepting of course the dykes, many of which are very broard. —

In very many places I found large boulders of greenstone. 471 ... 475 525 579 (crystals & both ingredients well marked (b) & some Iron pyrites). — From the local situation & height I have no doubt the rock was in the neighbourhead & that they were not carried by the action of water. (c) from any distance. — In only two places I found the rock, in situ. — At a village called Pedro Argulheri (spelling?) or the rock of needles the road passes through a broad dyke of fine grained decomposing greenstone. (a) — it cuts through & partly overlies the gneiss. —

At the island of Villegagnon, there is a vertical dike cutting through the mica slate: the junction is exceedingly well defined & neither rock is altered by it. — This dyke must is at least (the sea hiding termination) 20 yards wide & 100 long. — It runs N 1/2 W & S 1/2 E. whilst the cleavage of Mica slate strikes NE & SW. — The rock of which this dyke is composed varies in character. — sometimes fine grained black; crossed by many lines of cleavage. 620. 621 — at other much paler-coloured & layered crystals 622 ... 624

57 verso

(a) The famous peak of Itaculumi is composed according to Eschwege1 of Quartz rock & overlies Clay-Slate: Here we have a great bed, also abounding in Iron in the Porph: gneiss: Humboldt has observations on the Quartz beds of S. America P 116 "Superposition of Rocks."2

1 Eschwege 1832.

2 Humboldt 1823.



1832 May June

Rio de Janeiro


forming balls. which are singularly difficult to break. — . Again it passes into a rock allied to those of Serpentine formation (?). this latter sends off small veins into the Mica slate. — 622 ... 623 724 625 ... 627 From its position & mineralogical character this dyke evidently at some considerably posterior time has cut through the mica slate. — This is singular Must we suppose this is formed by some chemical alteration of the primitive crust of the globe. or by some internal revolution. by which this more modern rock is brought under the antient Mica slate. — probably this is one of the layers or veins of diorite mentioned by Mr Eschwege & quotes Humboldt. — I could no where find the Phonolite of the same author. —

I ought to have mentioned the occurrence of large siliceous beds. 611. 612 549 547. 548. 613 544 — 546 608-610 — in places nearly pure; in others jaspery & abounding with Iron. — this latter & a rock in which the quartz is in line (specimen 544 &c & 608 &c) is found on the Corcovado & East of the Lagoa. — (a) where from its withstanding weathering & cavernous appearance. looks like an old Castle wall. —

From what I have stated I should suppose the formation of this country ranks amongst the more modern of the primitive class. — In confirmation of this I will describe an appearance in the common Porphyritic gneiss at Botofogo bay. — An oblong mass of siliceous rock 606. 607 in which mica & garnets are arranged in parallel lines is surrounded on all sides by the

58 verso [blank]



1832 May, June

Rio de Janeiro


usual gneiss. — The line of cleavage in both rocks is nearly the same, but the mica slate dips vertically the other at small angle & therefore rather overlaps the slate. — The lateral junction is well defined. — The block is 7 yards long & 2 wide. — Subsequently it has been penetrated & slightly dislocated by veins of granite in large crystals. — I can only account for these appearances. by supposing mica s slate to be entangled during the formation of Gneiss: — invariably the mica slate is inferior in the present state; but from the alternation &c they must have been formed about same time — it hence results these blocks of mica slate belongs to some older formation. —

Ferruginous springs are not very uncommon. they so abound with Iron. that the bottom of channel is lined with the oxide. — On road to Tigenka one is covered in & used medicinally. — I may here observe — that I believe Turf does is not grow formed here: & there were several situations well adapted for it. — Is not Turf generally absent in Tertiary formations? —

In Geolog: notes at Bahia (P 32) I mentioned the occurrence of hillocks of decomposing gneiss. — the same phenomenon is largely visible here: & my opinion about is altered. Many of the hillocks are composed of it & the beds at least 100 feet thick. — It is clear, they have remained in same state

59 verso

(a) I have also traced veins of quartz. — In Spix there are proofs for state of constituent minerals, that the lithomarge is not an independent formation. — (Notes Vol II P 168 & 198)

Analogy with Wacke in the Trap Formations

(b) contain Feldspar. Mica. quartz in variable proportions. Neither can I understand the universal high angle of dip. — In volcanic rocks this is not the case. —



1832 May, June

Rio de Janeiro


for some long time, they existed previous to the formation of the great vallies — As at Bahia (1), the lines of bedding cleavage remain perfectly distinct. — the clay varies much in colour being red. white. grey or blue. 521. 522 — but bright red is by far the prevailing tint. — I cannot attribute these facts to decomposition it would appear more to be owing. to the crystallization having only proceeded so far as to group homogenous particles. without fairly solidifying them. — My reasons. are the impossibility of water or air penetrating so deeply into stratified clay. — from observing that the upper parts do not always appear the most decomposed. — & from not seeing on the surface, where the gneiss is now weathering any tendency to form the described red clay. — I came to this opinion before reading Daubuisson. who states somewhat similar facts. — I should imagine that where the mineral was not solidified, the constituent parts of Mica. were very abundant. it would explain the quantity of red oxide of Iron. as Mica contains — 22. — I also observe that the Porph: gneiss weathers chiefly where it contains most mica. — in others it long resists the action of air 7 water, as is shown by the unaltered curve of the hills, when dipping into the sea. —

I must observe before finishing. that studying a Gneiss district (b) is very startling to the igneous doctrine of the formation of those rocks. which

60 verso

Boulders of greenstone. not in situ on road to Villa Rico. P 195 — Vol II Caldcleugh.1

From Mr Fox observation on whole coast the South of Rio, there are Greenstone Boulders. —

1 Caldcleugh 1825, vol. 2, p. 195.

From Mr Fox...Boulders. —] added pencil.

[N.B. Page 60 ends with 'which' - apparently an unfinished sentence which is not continued on 60 verso or 61 (next document).]

This document has been accessed 2750 times

Return to homepage

Citation: John van Wyhe, editor. 2002-. The Complete Work of Charles Darwin Online. (

File last updated 22 March, 2013