RECORD: Darwin, C. R. [Notes on the geology and corals of Keeling Islands]. [4.1836] CUL-DAR41.40-57 Transcribed and edited by Alistair Sponsel (Darwin Online,

REVISION HISTORY: Transcribed and edited from the manuscript by Alistair Sponsel, transcription checked and amended by Kees Rookmaaker, further editing by John van Wyhe (JvW). 8.2010, 5.2011. RN2

NOTE: Darwin's time in the Cocos Keeling Islands is recounted in the Beagle diary, pp. 413-419. See the chart of the Keeling Islands from Narrative 3.

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
«some text» 'some text' is a later insertion by Darwin
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.


[Sketch of a reef island in profile]
E P D H C B A Islands L
generally much smaller More more

Beginning with the Lagoon: the shores are formed of fine Calcareous sand. with a few fragments of Coral which more frequently is of the «delicate» branched kinds.

The shore has either an entirely uniform slope. or has a small beach. which latter circumstance generally occurs at points where the tide sweeps round «or where to leeward. the waters of the Lagoon have some power». — So gradual is the slope, that generally when wells have been dug within a hundred yards of the beach at spring tides. the fresh water rises within 18 «or 12» inches of the general level: the water ebbs & flows in these wells. about 4 ft. in which thickness. the whole is Calcareous sand. partly «slightly» consolidated agglutinated. This formation appears to extend some way inland: & walking across these strips of land. (which appear only a few hundred yards wide) first we meet on the surface branched & lighter Coral «caused by such always been. thrown. as the beach advances, furthest back», & close to the outer beach the solid Astreas. — The level of the strip appears perfect. but. as the external beach is 5 or 6 feet above the wash of high water in ordinary weather: & the lagoon side has its edge on a level with the high water there: there must be the rise in the island. equal to the height of the heavy weather beach. This few number of ft. is of course quite imperceptible: On the lagoon side. inland. in parts. a beach of sand. may be discovered. which of course has been formed. as beaches. sometimes are now formed on that coast. —

page lightly crossed out in pencil.

1 verso

The stratification in whole mass of Lagoon rocks on the main horizontal, in detail very irregular.

Bones of collar of Huthuriae1 ought to be found. — & adhering oysters; in outer reefs

All Lamelliform Corals.2

Astrea does grow in the Lagoon

Young Islands. appear always to assume a convex side to seaward:

Some of the included fragments are 1-2 ft in diameter

The series of specimens is very interesting.

1 Possibly Holuthuriae.

2 written at the point of a bracket encompassing "far prevalent genus...these (3575)." on CUL-DAR41.41.

The stratification ... irregular] pencil.

Bones of collar ... oysters] pencil.

All Lamelliform ... on p.2] pencil.


on the corner of the Islands. bordering the passages between the reefs. there sometimes is a hillock. which appears composed of blown up sand. (1) (probably frequently possibly in some cases» consolidated as at K.G.s Sound).

The beach A is composed of well rounded fragments of solid Coral. — these are generally. about 1 to 2 ft in diameter. & rather thinner in depth: the far prevalent genus is Astrea: perhaps more than one species. (but species 3568 ... 72) is the common one. — The next kind in abundance is (3573 3574 «3585») & then (3575). — all these are either in a tolerably perfect state. or much petrified by Carb of Lime «& fragments of this Breccia». — The smaller fragments are the same. or a few btanched kind «& millepora». — This

A succession of beaches. have formed all the outer parts. of the strips of land: The sun & atmosphere. acting on the larger blocks causes them to split. & their angular fragments. fill up the inequalities: the branched kinds are abundant on the surface because being lighter. they are thrown further inland, as the beach progresses outwards. — This beach is thrown up. by gales of wind, a heavy surf. & spring tides. — Beneath this beach, there is an extent (at this particular section) of only 10 yards. (generally more «?») of a breccia. the rock is very solid. white colored & sonorous beneath the hammer: this breccia consist of Calcareous sand. cemented & the fragments blending into each other: generally it contains or

2 verso

NB The action «of» gales on the outer parts must be to grind & form Calcareous powder. —

recoil carry it to Seaward

N.B. Several Earthquakes & the Greenstone Rocks on the Northern Islands in the Breccia – it is sand

several small pebbles have been found. of the same rock.

NB. Greenstone Rock. brought by roots of trees

direction of longer axis? breadth of strips of land; line of Northern Isle = Has before been a lagoon. =

One single tree of a particular sort on one old Isd close to the Beach

The most surprising thing is the quantity of Lime! in the whole intertropical seas!

We know that sediment at K. Georges Sound will accumulate at ∠ (30°?). —

V. Humboldt1 for inclination of Mountains.

The number of worms tend to destroy the Corallina living

Limestone strata, if formed as reefs ought to perforated with Cyclindrical holes.

N.B. Crystal like pieces. broken spines of Echinus

1 Humboldt 1819-1829. JvW

«of»] pencil.

recoil carry it to Seaward] pencil.

N.B. Several ... Echinus] pencil.

V. Humboldt for inclination of Mountains.] lightly marked out in pencil.

The number of worms tend to destroy the Corallina living] lightly marked out in pencil.


is composed of any fragments of all sizes of the Corals. which are nearly perfectly petrified: there are large masses. in which scarcely a trace of Organic structures can be discovered ∴ the rock would be described as a yellowish or white. Calcareous rock. with an almost Crystalline fracture: containing a few minute cavities. — The horizontal extent of this vary varies much, in thickness it can rarely exceed. 3 ft. — it is a very curious & interesting rock. — It is generally covered by «near the limits» at of high water. — this mass. is irregularly stratified & slopes. just perceptibly to seaward. owes its origin clearly to fragments heaped up. & cemented by the action of atmosphere & tides:

its quantity is constantly varying. as shown by the water worn state of the ledges: the «heaviest gales» tides sometimes «heap it up &» consolidated it – & sometimes «common action of tides remove it» remove them: & gales tear off fragments to «help to» form the high beach. — To seaward of the Breccia

There is a flat. which in common language would be called absolutely level. but certainly «in most places» there is a trifling slope to seawards. — About 60 yards is quite smooth «(These numbers are all small) especially as compared to West side». then it becomes traversed by shallow channels & holes. «NB Sulivan1 would include all this in reef. —» which over a space of 30-40 yards increase in width & depth. It being low water & spring tides. «& extraordinary smooth water» by the aid of a leaping pole. we reached very far into the breakers. — Here we found great masses. «rounded» with a curvilinear outline, up to 8 ft in diameter

1 Bartholemew James Sulivan (1810-1890), second Lieutenant on the Beagle. JvW

page is lightly crossed out in pencil.

seaward of the Breccia] pencil line separates the previous writing from the following.

3 verso

N.B. the greater «less» slope «of bottom» to windward owing to greater accumulation of Detritus : if such is the case? —

It is certain that the kind sub-branched. of which there are so many fragments: must grow at a little depth in greater quantity. — For structure of reef outside V. soundings.

N.B. the greater ... detritus] pencil.



living Astrea. This species ([blank space]) formed the chief masses. — There was also of a yellowish green. Millepora. which grows in thick vertical plates. which intersect each other at various angles. & form an exceedingly strong net work. «V. note Book for species: the branched Millepora: &c &c» In the interstices flourish strong branching Corals. —

The central dead parts of the masses of «of the Millepora» net works «& the towers of masses of Astrea» are coated & raised by a Corallina. — The channels are coated with the same substance & have a little sand. at the bottom. to seaward. similar lumps. for a distance of 10-20 yards might be seen beneath the water. beyond this for another 20 yards there was green water «notable» water & then came the Blue. — Where we stood. the masses of Astrea were from 4 — 8 ft in diameter, but irregular in figure. & the channels about 6 ft deep: The Astrea, was on its surface to a depth of 3 or 4 inches was dead. further inland a greater depth was dead. to seaward. without doubt the whole surface. was living. – It is a beautiful instance how short a time is sufficient to kill them: at this time under such favourable circumstances the dead parts were awash. The Astrea. extends laterally & must soon fill up the channels. — Specimens. ([blank space]) will show. a layer. additional on the sides as compared to the top. — The Millepora

entire page is lightly crossed out in pencil.

«notable»] pencil.

Where we stood ... to the top] text lightly crossed out.

4 verso

NB. absence of Coral Isds in the Atlantic. — Coast of Peru. which perhaps are exceptions to the whole rest of the Intertropical world. —

NB. Absence ... of Peru] pencil.

All writing on the page lightly crossed out in pencil.


equally requires constant immersion. — The interstices must «are» soon to be filled up & form solid masses. — I was most surprised to see. the enormous quantity of matter. which the succesive paper like layers of Corallina

have accumulated, instantly the This substance evidently bears exposure. for some time to the air «Yet. which is very curious. must require a breaking sea:» — instantly the surface of the Astrea dies. it is occupied by Corallina — The summit & sides of the dead Millepora. appear to offer «most» favourable situations to its growth. as do the sides & bottoms of the channels. This Corallina raised lumps 8 & 10 ft long about two feet above the level. the living solid Corals. — The absence of all points of resistance must allow this substance to flourish in the break of the sea «There is another species of Corallina. [sketch] V. Specimen», although firm it is not hard. — (V 3580 specimen) — Further from the lowest break of the sea. the channels. becomes shallow & narrower; The Astrea, when it grows. is in the form of thin plates, with curvilinear edges — Here a great number of species of Algonium flourish & some stony branching Corals. — In one part, over the extent of a mile an enormous quantity of a Clam «Holuthuriae» lived. partly sunk in the Coral. — «NB.» These are placed vertically might afford useful indication. — Worms were in Milliporae. —

page except for bottom lines (from "a Clam" onwards) lightly crossed out in pencil; a pen line separates "...level. the living solid Corals. –" from the section starting "The absence of all points of resistance..." and they were marked over by different pencil deletions.

have accumulated ... Milliporae] pencil, with insertions in ink.

5 verso [blank]


In this part Corallina likewise flourishes, but does not form elevated knobs. — In this part, they would remain long without water, although the same height. from being further from the break of the sea. — When dead probably « the perhaps» they would exfoliate & their fragments would help to fill up the smaller channels. — Perhaps Flinders Negro heads «are of this formation». color pinkish, Flinders black. —

When we reach, the level floor. we find it quite smooth. & covered with very minute Fuci. — The channels have been entirely filled up. & being long exposed no Corals lives there (I exclude a few hollows). — The whole is composed of a very hard solid rock. – which is petrified Coral & hard Calcareous sandstone (3578 & 3579) — It is manifest, that the few remaining channels between the living Coral. have gradually been filled with detritus the whole petrified & smoothed by the action of the tides. —

This flat slopes at a very trifling angle to the Breccia. — On its surface are scattered a few blocks « (but not very large) 1 to 2. & sometimes 3 ft:» which evidently have remained there sometime. & are waiting till an unusually heavy gale shall carry them to compose a Breccia or high beach — where a few channels yet remain, if a

oage in pencil; whole page is lightly crossed out in pencil. Insertions and underlining in pen.

6 verso

NB. Alcyonium includes Sigillina &c &c

Cocos Isd. connected with volcanic force of Sumatra. that rises, this falls.

page in pencil; writing is also crossed out lightly in pencil.


a fragments of Astrea is cast, the into a cavity the lower parts of still living, it adheres & so fills up the Cavity. —

This section I understand, gives a good * general idea of the reef. the proportionate width constantly varys., the breakers being either «much near» close to Beach or four «(?)» times as far as I have stated — the inclination of the bottom «& other local causes» probably determines this «I cannot find out any certain relation to Windward or Leeward sides» —

This is a section of a reef forming on to land; on the reefs which connect the Islands. are very broard, are nearly low dry at «dead» low water. are intersected with «numerous» channels «they are nearly level or rather slope to lagoon, cause as described at Tahiti» their outer parts are as described & within these whole is no> no part quite so smooth as "the flat." probably because not so high, the sand & other loose materials constantly being removed & carried into the lagoon. — From the width of this reef it is manifest, that such is the base of all the Islands. — Where a Minute Island was formed on the reef. the whole process was evident: — (Omitted, slope imperceptibly «universally this structure» to lagoon, resembles Tahiti. hence have no doubt description is right); there can be no doubt the Coral extends along outwards to sea, fragments broken off & falling on the sloping

page] pencil: top is lightly crossed out in pencil. insertions in ink.

This is a section of a reef forming on to land... sloping] each line is lightly crossed out in pencil.

7 verso

Section of Island

[sketch]1 Sea 3 ft 3 or 4 fm
Loose fragments
Sand on Lagoon side branching Coral Fragments & Sand
Corals in situ interstices filled up

Breccia Coral Breccia Coral &c
Fragments. sand —
Branched growing coral
1000 fathoms
line of possible degardation if upraised into dry land

The period of time between each elevation «depression»2 would determine how far the slope, extends outside?

NB. If tides had been very small. when such a mass had accumulated, there would be very little Breccia & [illeg]

1 This diagram is discussed in Stoddart, David R., 1995. Darwin and the seeing eye: iconography and meaning in the Beagle years. Earth Sciences History, v. 14, p. 3-22.

2 Note this slip.

Section of Island] pencil.

diagrams and text] pencil.

The period of time ... slope extends outside] pencil.


talus of sand would form foundation, as the rounded lumps of Astrea & network of Millepora extended to Seaward, the broad channels between constant break of the sea, would be lessened behind, the surface of the Astrea would die, laterally they would extend & fill up the channels — & Corallina would form «[illeg]» lumps — «During the times» as the front extended to seaward, so would the channels control, the upper parts of Coral «would all» die further down, the lumps of Corallines crumble down & th «but the» branching Corals «may still» flourish more. (There «in front of this» & sand would fill up all the «remaining» channels & the Coral be petrified — as soon as this was effected & the break of the sea still further removed. no Corallines or Corals would grow. & the weaning of the sea.) In this state if, the cross action of the tides «& gales of wind in unusual direction» «(or other causes» would «should» accumulate on the less exposed parts of the reef a heap of fragments there might be cemented; Then & form a barrier. Then the remaining channels & holes would be filled up with sand & fragments, be petrified & the surface worn smooth by the tides, & we should have the "solid field" — Upon the Breccia in the little Isd. loose (because dry) fragments would be thrown; which extending outwards would form the island, & on the inside «would from same process» the sand of lagoon.

& sand would fill up all ... weaning of the sea.)] crossed out.

8 verso

Small Pumice pebbles «abundant» on beach of Lagoon – from Sumatra, like the seeds

Small Pumice ... seeds] pencil.

«abundant»] ink.


Wednesday [6 April 1836]

Visited the SE Isd. — By an old chart said to consist of several Isds the junctions are still visible by being narrower, lower & «even» the trees younger. — In a well near the houses. we found strata 3-4 ft thick either of slightly indurated «not very hard» Calcareous sandstone, a breccia of rounded fragments, which in one part was excessively compact. — At spring tides, the water within well rises within 2 ft. of general surface, & probably this height does not equal the rise in the lagoon. therefore the land is so much less higher than the 2 ft. —

Whole surface [illeg] Islands strewed over with the (2d common kind) & some Astrea, well rounded. «& some shells which Mr Lesk1 says only are found outside» — the cause must lie in these being always tossed farthest as the coast travels seaward «This quite agrees with reef being covered from seaward». = On the sea-beach, the reef from Breakers to coast about 200 yd.; of this the structure is identical with what has been described. excepting that the Breccia was in larger proportion. — The beach was here about 9 ft above spring tides. & therefore the land at this spot & there are hillocks of blown up sand about 14 ft: It is close to the beach where this 9 ft is found: which must make the land 8 ft higher than at the wells. — Making allowances – for the wear & tear «splitting» & settling of the fragments. & some of the sand. being thrown away. Yet, the 9 ft land would probably. always remain higher than that near the lagoon, which has been accumulated in a similar manner. This may be owing possibly to the inclination of the land, at an earlier period, causing the sea to break further off than at present, & so to heap the fragments less high. — If however, this outward increased

1 William C. Liesk (sometimes spelled Leisk), an English resident on Keeling. See Armstrong 1991. JvW

9 verso

X. Considering the breadth of reef between the 2 this must always happen. —

As the reef is broarder at SE. point, than behind Houses = How are the soundings. =

NB. Yet Sulivan makes the reef. very narrow. on some of the point on SE part (owing to greater accumulation?)

The four Bulwark agents
1 Astrea
2 [sketch]1 }
3 Millepora } ? which most effective
4 Corallina

The width of [illeg] discoloured water beyond the Breakers varys much. — I could discover no law. – V. chart excepting on West Isd. where the narrower the reef. the wider the pale water & V. & Versa. Sulivan.

1 Little diagram, as on p. 5, of unnamed species of "corallina".

X. Considering ... always happen. —] pencil.

As the reef ... soundings. =] pencil.

The four Bulwark agents ... Corallina] pencil.


height was general – I should suspect. that at the violent Earthquakes, the land may subside a trifle, as it rises at Chiloe, without being noticed; the outer parts near the Beach would constantly be repaired, but not so the interior. —

The upper parts of the Lagoon are much filled up with Coral. extensive flats are nearly awash at low water, & only here & there a circular hole of 10 or 12 fathoms deep is left, & these are daily being filled up. The more common depth is about 6 fathoms, so that this is a shallow lagoon. & here perhaps we may infer that all «subsiding movements on any great scale» here has long been quiet.

The commonest species are the (Seriatopora. .. Crown Coral, (a yellow sort [blank]), ([blank]) Fungia. Escara «Chama»:– Meandrina in great loose balls. — Astrea, (the bulwark species) infrequent. as are the two other kinds which are found outside «& Corallina not abundant». – These Corals are brittle & soft, & on standing on them a person breaks through them to some depth. —

For a space of some miles, near the head of Bay, there are extensive flats of dead Corals, such as are found a mile further down alive, these are remain in their proper position, but are quite rotten & old, so that a person sinks as in walking through straw. — I was much puzzled. what could have killed them, without a general change of level. — I look at the cause. as owing to the strong wind causing very low tides, which before the SE Island was united would not happen, but rather. as happens.

& Corallina not abundant] pencil.

10 verso

All the Sand is purely Calcareous

All the Sand ... Calcareous] pencil.


in the Pacifick, a strong breeze would pump in quantities of water. — From this change of circumstances. the part which might be exposed for an hour to the sun, might well be 2 — or even 3 ft lower than formerly & this would quite suffice to kill a Coral flat, which had attained its extreme height. —

I observed this dead [brittle] Coral, showed no signs of adhering & forming as rocks. but rather of wasting —

The shores at the head of the lagoon, have a different character from any other part; here is an extensive bed of mud which reaches far into the bay, at a very small angle. — Here flourishes the Turtle grass – This mud is entirely Calcareous. — I can only account for its presence by the trade wind drifting across the finer particles of sand which is accumulating on the Weather coast.

If we imagine the future state of the Lagoon. we shall see a perfect ring of land formed. & with. new tidal opening complete excepting by one orifice. — The Corals will have grown up over whole extent. to their highest limits «that is to «lowest» low water» & will exist in the state, in which they occur at head of bay. that; their comminuted fragments will form a fine bed, on which Turtle grass will grow. — Which will be left nearly dry at low water. — hence the opening will be merely tidal. — This is the present state of Keeling Isds —

If we except the fine sand, which (where many trees were not present) might be drifted from the one or windward side. It is difficult to imagine how it would

11 verso

Cocoa nut trees & old store encroached on by the lagoon.

There are on Water Isd. several. stores & other buildings. built. 6- to 8 years ago. which were built on the beach. above «above common» highest water mark: now these are surrounded by water, during every tide: fact agrees with cocoa nut trees

Well in «nearer to the lagoon. than outside» centre of Isd. sand to high water mark. beneath which loose (outer — fragments well rounded [ ) ? ]

«therefore here there must have been considerable increase from Lagoon side +++»

Depression corresponds to Sumatra line of Elevation as Pacifick to Cordilleras.

+++. It is remarkable. that at high water mark in the well (therefore certainly a little below same in Lagoon) that there should be loose small pebbles from outside (NB since seeing spit in leeward part of lagoon. I can distinguish the two): at same height outside. there would be Breccia or base of high beach of large fragments. This explicable from extreme width of reef. before land: but it is one out of many arguments helped by small depression.=

Cocoa nut trees & old store encroached on by the lagoon.] pencil overwritten in ink.

Well in] pencil.

entre of Isd. sand ... rounded] pencil.

Depression corresponds ... Cordilleras.] pencil.


& in proportion as Lagoon nearly filled up, fresh water & impurities would be more ingenious

ever entirely be silted up. — We must not allow false appearances to misguide & to compare a lagoon with a lake surrounded by higher land. from which detritus might be carried. — These facts may account for infrequency of Lagoons, being converted into «absolutely» dry land ++ — moreover. when the lagoon was nearly filled up the impurity of the water might add. much to the slow growth of the Corals —

++ Independent of repeated depressions. —

[small diagram of two reef lines] 1 2

These two kinds of reef are equally common: even perhaps. (2) is most common:

These two kinds. although at first appearing. very different. then are the same: the lumps & long continued smooth breakwater «very curious» are raised about 2 ft above the level of the flat: they are each composed of superficial Corallina. — We might suppose in number one. if the lateral & outward growth of was checked: the Astreas &c would close up the channels & the Corallina knobs unite into a breakwaer. Perhaps. the species may [be] slightly different in the two cases?? not essentially: yet in «the» one case Petrifaction appears to take place & not in the other. — I could pick the knobs. but not the Breakwater. —

It is hard to believe that the solid breakwater is cut down. as the breakers recede outwards. yet such almost must be the case: as soon as the Corallina dies, wear & tear commences. & we know in solid rock. the upper part of tidal surface. suffers most degradation. — If the breakwater is not cut smooth the reef cannot increase laterally for if so such mounds would appear on the flat. —

& in proportion as Lagoon ... repeated depressions. —] pencil.

12 verso

The Windward side «with respect to land» thought oldest. most perfect

Mem. Keeling Isd. opens to windward

Mr. Sulivan states that the SE decidedly in the windward side & SW leeward — Curious facts of blocks thrown up, – doubtless when on outlying Islet; sand dunes protected by Vegetation: no ridge on strips of land; «∴ not essential to a perfect lagoon Isd.–» no ledges in lagoon: —

I believe the stone crops out in several places on sides of Lagoon, which looks like a trifling subsidence —

[7-11 April 1836] Horsburgh Isd. on the Lagoon side «in one spot has accidentally» fine calcareous sand is turned into hard rock is stratified [small sketch] (a) (a) «i.e., much smaller are» like Coquimbo — it must be the lower part of sand, «hardened» intermed[obscured] beds happening to be softer. — The Island. being to leeward appears to have increased largely on lagoon side by sand heaped up hence low. levels — In the Breccia outside. rounded pieces of Breccia again become cemented. — Certainly the lowest parts of the coarse beach: = (NB a note has been entered in Coquimbo paper)

The Breccia. beach being much nearer to Breakers than on Leeward may be owing to surf — & not to comparative growth of Corals — The strips of land. may be owing to increase from Lagoon side. — «It is hard to judge from their compounded causes»

The Windward side] pencil.

thought oldest. most perfect] pencil.

Mr. Sulivan states ... strips of land;] pencil.

no ledges in lagoon: —] pencil.

I believe ... subsidence — ] pencil.

Horsburgh Isd. on ... increase from Lagoon side] pencil.


(1) These hillocks «are chiefly» on the Windward shore: there was one which Mr Sulivan measured estimates at 24 ft. & another even considerably higher: even 35 ft !

The North point of Horsburg; a sandy beach with shoal water extending a long way out. quite exception to general structure of reef «produced by set of tides»

figure of the Lagoon. quite irregular.

In general discussion: paucity of plants a difficulty in the depression theory. – «Mem. Lesson»1

The structure of these reefs. strictly agree with Flinders.

Barrier: in the Pacifick. from different tides. perhaps not. Breccia in Beechey.2 There is nothing distinct about

Brecciea: but simply. about. the Coral itself being petrified.

NB. The reef ought to be described first & then the structure of Islands «: Islands far back on the reef»

Mr Sulivan says there is a larger proportion of sand than Coral in the lagoon: describe species

Certainly taking the whole Isd. together the Breccia. is wearing away. in a large proportion: I have been inclined to think. that all the elements of the reef. were crawling outwards. but in some places it would certainly appear (as in Breccia on West Isd «& block on windward. sand-dune») that it was deposited when the sea broke much nearer: — Perhaps the Island ought to be considered as at rest after changes. consequent on some change of level. = It is scarcely possible to follow out all the consequences on slow or sudden depressions: & outward growth, too many uncertain elements. —

The Breccia. (chiefly from Sulivan, certainly does in places slope «gradually» to seaward. in a few spots. on to the Flat

When land & water are opposed. such equal & antagonist powers. a depression of an inch or two would turn the Balance of degradation. The depression must have been triflings. from fields of dead coral

1 Duperrey1826-30. JvW

2 Beechey 1831. JvW

Note that this page has been torn and used to be (from the remaining fold) in the position of an even numbered page.

even 35 ft !] pencil.

produced by set of tides] pencil.

Mem. Lesson] pencil.

: Islands far back on the reef] pencil.

13 verso

This may be considered as a very perfect lagoon Isd. — long quiescent. shallow lagoon & much proportional land: figure open to leeward. – «Above half the lagoon near dry at «spring tides» low water»

Where there are layers of solid stone on agglutinated sand. F. W. where loose stones. salt water. – Liesk

Temperature at 770 Fathoms. 45°. — surface 78°. —

at Diego Garcia (where in Indian ocean ?) in 1812 severe Earthquake. which rent a Islet. – Capt. Ross1

Capt Ross suggests heaping up of fragments by an oblique swell. to the usual direction of gales. —

The action of worms & shells in eaten coral. added to the Fish in formation of sand: but sand probably carried in gales. through the openings. – NB. on outside all [injurious] sand. removed. in lagoon accumulated.

N: Keeling Isd. has same identical structure of reef. Liesk.

1 John Clunies Ross (1796-1854), Scottish sea captain who settled Keeling. See Armstrong 1991. JvW

«Above half the lagoon near dry at «spring tides» low water»] pencil.


(3565) Yellowish white. vesicular stone. soft. consisting of small particles of shells & Corals. intimately united & blended together

(3566) do. containing «angular irregular» patches of compact yellowish Carb of Lime. which apparently have been a lamelliform Coral with the cells. completely filled up. —

(3567) compact yellowish almost Crystalline fracture. Cast of Lime. probably Coral petrified.

(3568) Astrea. the commonest Coral. block on the outer coast

(3569) do. more compact. perhaps other species.

(3570: 71: 72.) do more or less altered & [illeg] to the character of the yellow Carb: of Lime

(3573. 3574.) The next most abundant kind; the latter partially petrified; sub-branched. —

(3575) Perhaps a distinct species. not uncommon but more so than the foregoing one. (.3574). Probably same species. partially petrified. —

Mem. the pink encrusting Corallina.

All the Corals. Lamelliform: «No»

A Lagoon Island a map of Carb: of Lime. – rock. shell. Corals & sand. —

[sketch] a b c

This sheet is a totally different kind of paper from DAR 41.40-46, which are all watermarked "W. Fincher 1835"; this paper is watermarked "Ben Picardo 4" and is torn in half in the same manner as DAR 41.46.

14 verso

(b) To irregularity of [recoil] tide cause?1

Beach of lagoon sand. —

Mem: anecdote of heavy rain. killing all the Fish

Mem: Capt. Fitz Roys idea of Fish causing Coral sand.

Largest fragments about 2 ft. square. by 18 inches deep: —

Mem: all Islands in Atlantic Volcanic: therefore we must expect. that such would happen here. —

The Breccia. very solid. – shell fragments. a good many branching Coral. from. reef. —

(3577) The hard. sonorous cemented petrified Breccia quite white. —

3578 — white petrified Coral. from superficial solid floor

3579. together with compact calcareous Sandstone

3580. Mass composed of layers of a pinkish cellaria (NB. there is a pretty scarlet Corallina there)

Capt Ross

3581. From «large» rounded fragment in Breccia of a compact solid Greenstone

3582 Astrea converted into «snow» white Calcareous rock glittering Crystall. fracture & part converted into a ferruginous — Calcareous stones

(3585) The 2d «(or equal)» most abundant Coral same as (3573) Its large proportionate quantity is perhaps partly owing to being thrown furthest. —

1 (b) probably refers to the sketch on 14.

Capt Ross ... thrown furthest. —] pencil.


Monday [4 April 1836]

Crossed over to West Isld. — Found on one part of Beach, exposed beds of Calc Sandstone precisely resembling those of Horsburgh Isd.: Here then the Lagoon eats its own formations: I several other places old trees have been undermined & fallen:

In another part, a large lagoon covered with the finest white sand, is only covered at Spring tides;

It is formed of spits of bits of branched Coral thrown up by gales of wind, into which the tides carry very fine sand, no doubt in these parts the general action is to add to the land «Where I have seen such addition is off Horsburgh & West Isd: at SE Isd. there was shoal but no addition.». & in the whole upper part of lagoon. — On outer coast, which is the Leeward coast, the reef is very broard; the Breccia slopes down to it, is only covered by very few loose fragments or by a sand beach — This is evidently owing to flat having totally destroyed the effect of breakers. — There is however in part a little way inland, a high beach & some sand hillocks. — That the surf once broke here is certain; because in an oblique line, (from the curl of the windward swell) on a point there are heavy breaker, on the shore in this line is the thickest & largest mass of Breccia I have ever seen, stratified in a curve conformably to the coast — Ancient trees grow on the few lones trees stones on the Breccia, & indeed in a great part of this coast, the sea is rather removing the loose sand covering & trees above the Breccia, small subsidence:

This is another half sheet of the "W. Fincher 1835" paper; not an apparent mate of DAR 41.46. Page written in pencil.

«Where I have seen such addition is off Horsburgh ... addition.»] ink.

15 verso

The reef here is precisely similar to Tahiti1 only rather higher, covered & superficially composed of the three species of Corallina «(the pretty species I saw no where else)». — (Mem. how strange it is Corallina, although tidal flourish only in the Breakers & a little (but I suspect a different species) in the lagoon). About 3 inches beneath the general level of the [live] Corallina, the breakwater is excessively hard, by chissel, pixaxe & I at last attained a fragment & strongly suspect it is Corallina «([blank])» petrified

Different circumstances determine waste & petrifaction

Everywhere I hear of water pumping over the reefs.

The lateral advance of Coral is certain. from the state of this reef. —

The slope of reef is to the Isld, in these distant reefs — the growing part being highest. —

The great inequality here subsidence. —

The strip of outside land is here narrow. but exists in full character: — Tahiti excellent description. Corallina on the Breakwater. chiefly the sub-branched kinds —

Few fragments on this reef. — sea no powers «: this perhaps is cause of width of reef: not having made a beach & Breccia.»

Sounding irregular. according to Manner & rapidity of depressions. —

Sand in lagoon not cemented.

1 Darwin uses Tahiti, rather than Otaheite.

The reef here ... three species of Corallina] pencil.

«(the pretty species I saw no where else)».] inl.

— (Mem. how strange ... sea no powers] pencil.

«this perhaps is cause ... beach & Breccia.»] ink.

Sounding irregular. ... not cemented.] pencil.


Mem: very great inclination between the 2 soundings on the SE side, so steep that it must be rock

12/1000 6/500 10./10000 10/8.0

500/6.000 (12, 0021

[sketch] 4 fathoms
2500 Fa2

1 Darwin appears to be trying to calculate the slope of the reef offshore.

2 There is what looks like it might be the beginnings of a diagram: a long vertical line, carefully ruled, with a mark in the middle saying "4 Fathoms" and another toward the bottom saying "2500 Fa".

this is another half sheet of "W. Fincher 1835" opposite side (but not a clear match) from DAR 41.46 and DAR 41.48.

Mem: very great ... must be rock] pencil.

16 verso [blank]


Sections at N Island.

[sketch] No. 1
11 m / 15 yds / 5 yds / [ten] yds / reef /light water / seven yds

[sketch] No. 2
20 yds / 100 yds / 200 yds

[sketch] No. 3
5 yds / 5 yds / 300 yds / 30 yds

[sketch] S / H Blende / No. 3 / no. 1 / no. 2 / N

at No. 3 the reef projects further than in any other part of the Islands, & the surface very smooth. scattered blocks from 1 to 3 feet through, on the flat inside the reef.

off No. 1. the Breccia is broken. and in some places worn level with the beach the land appeared to be gradually wearing away —

outside House

[sketch] 10 yds / 5 yds / 200 yds


The beach steeper & much Higher than on the N Island
the Breccia appeared to be in layers dipping to the sea — the lower, layers where the top ones were washed away much more compact.

another half sheet of "W. Fincher 1835" again opposite orientation from DAR 41.46 and DAR 41.48. Page written in a fair hand, likely to belong to B.J. Sulivan.

17 verso

At SE Point. about one mile NE of last section

[sketch] 8 yds / 8 yds / 100 yds

[sketch] Breccia.

at this point the top of the Breccia has been gradually carried away by the sea. & thrown further up the beach the larger pieces moved about 5 yards. the were from 3 to four feet long and about 18 inches thick apparently torn off in flakes The inner part of the reef (the flat) shewed rock blocks from two to four feet.

other SE pt about half a mile north of last

[sketch] 30 yds / 20 yds / 50 yds

[sketch] 4 ft / 6 ft

at this point two blocks of the Breccia were laying nearly on the top of the sand hill 23 feet above the H Water Mark. several other pieces not quite so large were laying near the top of the Bank. but not so high as these the beach broader than many other parts & composed of fine sand. The reef very close to the Beach.


In every case first inclination (blue water) to about 20 «-30» fathoms, then suddenly inclines — (at mean angle of 45°. ?

from land to 1200. ∠ 36° — from the 20 fathom. 39° —

from 270 to do 48° —

at mouth from the 5 fathom sounds to the 550 ∠ 24° – from the 20 to 550 21°

In the 1200 section 20 fathom 200 yd from shore

In other or SW section 20 — 100 — — — [20 fathom 100 yd from shore]

On scale of 1. inch mile (∴ sect 9. inch long) two line can barely be ruled so as to be distinguished apart if head a few feet distant

Cocos Isd. 6 or 7. larger Isd — 23 or 24 of all sorts beach with trees

Mem growing points with Cocoa nuts — «young trees. —»

Keeling Isd. 1. 400 yds «without reef. 1. 300?» width 1200 yd including reef

reef 100 yd – Blue water broadest on West side; on the northern part of which no reef «but» sand, & there 400 yd wide. — case like Horsburgh Isd. – one single slip of land. opening dead to SE. —

Same struct of reef. Liesk. —

Line cut by rocky cliffs

DAR 41.51, a small scrap of "Fincher 1835" paper.

In every case first inclination ... Cocoa nuts —] pencil.

«young trees. —»] ink.

Keeling Isd ... Line cut by rocky cliffs] pencil.

18 verso

Keeling Isd. little to W of North of 14 miles distant

«Cocos Isd.» While opposite side or northern. very imperfect. «especially west. part. breakers. —»

The deeper part of northern basin, which does not dry, 4-6. fathoms. with numerous patches flush. — & deeper holes

Longer axis NNW. 9 & ½ miles: shorter 6 & 1/2 ; —

widest average parts 800 yd. – leeward side berhaps broadest

Blue water 200 yd wide: Is of Horsburg 850 yd «(with 6 or 7 fathoms)» —

On East side. 100 yd reef from shore

On West. 200 yd —. —

On Southern end reef joins on to lagoon

Whole northern end. shoal. connected with mouth.

([sketch] form of growing Isld. —

3 & 4 fathoms at entrance, 9 at anchorage

Reef at S.E. houses. from do to outside («which is» all outside reef.«)» (550 yd) : at houses & near anchorage 500 yd: islets at anchorage begin to appear 250 yd from outer [break].

page in pencil.


Mr Sulivan. than in some the narrow channels between the «smaller» islands. the stream was so rapid. with 4 or 5 ft deep. that the whole boat could scarcely pull against it. & he was afraid to wind the boat! —

It is clear from Mr Sullivans sections that. no law with respect to Windward & Leeward. — Shape of lower Mountain


[19 verso]

matter — found out from Lagoon have determined. it — Compare with Beechey. —

NB I see there are some notes about the Coral Isle — in the scraps on the Portfolio

NB I see ... on the Portfolio] pencil.


«L Sulivans. outside deep soundings»

(85 or 55) (doubtful from boat drifting.) course Calc sand. fragments of dead Coral 203/10 sqaure bit of dead Gorgonia, living Corallina:

(275) Impression of large: Bivalve, little Calc: sand

(320) little Calc sand

(363) little Calc sand

[4]30 Coarse calc sand, bits of dead shells & Corallina ¾ «inch» square. stony branched Corals like those of T. del. Fuego

«Capt. Fitz Roy. near to Breakers»

(1) 8F Beautifully marked with Astrea. quite clean; axis smooth probably alive

(14) [sparing] Calc sand. arming broken

([illeg]7) arming quite clean «irregularly marked»

(30 & 28) abundant Calc sand

[illeg]15) very little Calc sand; arming much broken

(85 or 55) ... like those of T. del. Fuego] pencil.

(1) 8F ... arming much broken] pencil.

«irregularly marked»] ink.

20 verso

(30) Calc sand abundant

(21) Calc sand — do — arming broken

(16) Calc sand, tolerably abundant, mark of something hard

14) little Calc sand; arming broken

35) Much Calc sand,

«Mr. Sulivan thinks the latter time. depth 5-600 fathoms»

Mem the lines cut at great depths

«? what depths : — »

Mr. Stuart1 carried away his anchor in 13 F fathoms & lead in 16. F: The Capt when sounding in 10 & 12 fathoms. frequently had the lead jammed. so as not to be without much difficulty to extricate it. — How then rough the bottom must be. —

At Anchorage most impalpable sand or mud

generally sand in the hollows. excepting Capt Fitz Roy soundings & his in the deeper water chiefly only the foliaceous Madrepore. —

1 Peter Benson Stewart (d. 1864), mate on the Beagle. Darwin also spelled his name as Stuart in the Beagle diary. JvW

(30) Calc sand abundant ... 35) Much Calc sand,] pencil.


Monday [4 April 1836]
1— 270 — 0.6 «fraction of a mile:»
2 70 — 0.3-04
3 190 0.6
4 190 0.3
5 26 0.2
No Fathoms estimated distance

(1) Minute bits generally about 1/50 1/100 of inch (excepting one thin fragment of 1/20") partly rounded of either shells or Coral; a trace fragment of one. cell of a Flustra, which looked fresh, & «other» colored red. —

(2) a bit of living sponge; arming marked with rounded pebbles. & ∠r fragments nearly ½. inch in size.

Monday ... 1 - 270 - 0.6] pencil.

2 70 - 0.3-04 ... fragments nearly ½. inch in size.] pencil.

21 verso

(3). scales thin white «with serpulae. dead» ½ inch in size of dead inarticulate Corallina; sandy particles. – joint of white Halimeda: minute fragments of stony «2 or 3/10 long» Cellariae dead, (like those of T del Fuego) & encrusting Flustrae. — a small shell allied to Terebratulae = dead =

(4) very fine Calcareous sand & as before fragments. all owing to shells & Corals. minute bits «1/50" of inch» of stony Flustraceae, all dead. (excepting a very minute kind. a stony cellaria of a scarlet color)

(5) As above, joint of Holimeda «&c &c &c», courser. – perhaps. « .3 long» marks of pebbles: —

bits «one or» two or three tenth of inch —

one piece. very old & dead. 1 ½ . x 2 x 3 ½ , tenth of inch of a Lamelliform Coral. — with this exception I


saw no trace of that order nor of any living thing excepting the minute scarlet stony Cellariae

Tuesday morning — [5 April 1836]

numerous bits of dead, stony Milliporae « .1 to .2 long» minute. Corallina «Serpulae» & a scaly piece of Caryophillia (?) .3 square all white, brittle bleached 146 fathoms

237 fathom, arming almost clean, marked as from sand with two or three particles of do

184 do – with minute bits of same substances as before

22 verso

«31»). Courser, than any I have seen all dead. excepting perhaps. a scarlet Corallina. — fragments. scaly. some of them ½ inch square, ¾th of them. Corallina; one of them a bit of oyster shell, spine of Echini

The corallines partially retained their red color, the rest much bleached. —

NB. I feel little doubt about the impression of pebbles noticed yesterday

Friday 7th

(38) Marked with round pebbles, fragments 2-3/10 square of dead Coral. with bits of living Corallina. attached to it.

(42). «long» Piece of «lately» dead Gorgonia. fragments of dead Coral & Corallina (latter some living) ½ inch square.



(12). a little calc sand. bits of dead Corallina, & fragments one inch long. of very fresh. living (or exceedingly fresh. Millepora (of the branching species)

(17). Calc. S. coarse: bits of dead Coral.

(18). One or two specks of Calc sand: the impression of some large Coral

([27]). Well rounded piece & bleached 1 inch long, ½ broad, of the plate Millepora. much Calc: Sand:

(14). Coarse Calc sand. dead joints of Halimeda. «in quantity» bits of Corals

(39). Coursest fragments, I have yet seen
dead old well rounded encrusted with living Corallina & «.Flustra» a [Sigillina] & sea weed. One 1 & ½ inch in diameter: others 1 inch:

(Friday) ... others 1 inch:] pencil.

23 verso

branched stony Millepora. – The 2d kind (& possibly Astrea); Calc sand.

(14) Calc sand small dead univalve, dead Halimeda joints; in tolerable quantity

(34). coarse Calc sand. Minute dead fragments of shells.

(14). a very little Calc sand, perhaps the lead struck on something solid

(10) arming came up perfectly clean I could not trace any impressions

(11) arming. fresh broken pieces of a very fresh (or live) Millepora, color very pale brown: (Mem color of the branching kinds?) « It is this color»

branched stony Millepora ... the branching kinds?)] pencil.


(28). Sparing Calc sand

(12) Arming almost clean, a mark «one or» two grains —

NB. Very many of the fragments appear of family of Corallina. — the rarity of shells. remarkable. in any but extreme comminuted state. the fineness of fragments even in shoal water: the probably entire absence of Astrea because not easily broken. —

indeed all large fragments must be thrown up & small carried by recoil of the sea. —

We may conclude «that generally» deeper than 12 fathom all dead Coral sand; a fathom or two of fragments. —

(28). Sparing Calc sand ... We may conclude] pencil.

deeper than 12 fathom ... two of fragments. —] pencil.

24 verso

Soundings in [the] Lagoon:

The branched stony kinds grow in. 3 fathoms. — & in seven a purplish brown thin plate kind seems common «allied in Polypus. to Madrepore. V. NoteB.»

Saturday [9 April 1836]

(770) Fathoms little Calc sand, with a body cellular body encrusted with dead Serpulae «depth doubtful» «:(360 really)»

(240) Coarse sand, chiefly joints of Halimeda. bits of [mealted] old Coral. one ¾ x ½ inch & a case of a. [Pteropodous] animal

Soundings in [the] Lagoon: ... kind seems common] pencil.

Saturday ... with dead Serpulae] pencil.

«:(360 really)» ... [Pteropodous] animal] pencil.


to 12. Fathoms arming clean: Millepore. Astrea

(Above 20. — no sign of any thing hard. — in soundings)

(42 soundings. —)

even at great depth ½ [illeg] square. –

Few, but some dead fragments of shells «— comminuted. non living. —»; variously bleached Halimeda. & Millipora. living 92

«few» Bits of various minute corals; «Flustra» Gorgonia, The bit of a Lamelliform. — a scarlet cellaria alive 190

a case of [pteropodous] animal

Sponge alive 70: Echinus.

Sigillina. Flustra. Millipora. sea weed. 39

some doubt about living on account of drift. —

(4: inch diameter)

NB. Action of sea. cause of slope: cliffs 300 «500-600» cut

25 verso

inclination of [Teg de] 12°. 29'. Vesuvius 12° 41' Etna 10° 13' — Cones of Volcano. have a medium slope «from» 39° to 40° «Even the steepest parts but little exceeding. these numbers» — Humboldt

NB. Beechey mean slope. from the 20 fathom line appears nearly 45. — but much less as over towards the end. —

Prepidices of unfathomable wall

«4/10 1/4 = 4/40 1/4

8/80 1/4 = 8/80 1/10 »

18. soundings with bottom, above 30.

several between 200 & 300 fathoms

& two «two at» 360.

«1.2 12.00 6 7200»

inclination of ... unfathomable wall] pencil.

18. soundings with ... «two at» 360.] pencil.

«1.2 12.00 6 7200»] ink.


section on the East side between the settlement & old House.

[sketch] 15 yds / 50 yds

near the last

[sketch] 5 yds / 20 yds / 50 yds

at this point the rocks at the foot of the beach (Breccia) was more rugged with loose blocks on the top from one to 2 feet square. the beach covered with blocks of the same size the reef also very rugged —

NB. In this latter part the Breccia in one section extended to near the Breakers. about 50 yd. wide.

There is no law about the extension of discoloured water it is at least not less on the leeward. than windward side. & the broardest part occured there

no average width can be given: in one spot nearly 200 yds. –

On Eastern side whole reef (— from breakers to Beach) had average width of rather more than 100 yd: whilst on West side nearly 300 yd. — Breccia commonly not more than 10. or fewer hands. with some exceptions. —

Mr Sulivan seems to consider that very commonly the Breccia has been torn up & cast on the beach. in loose fragments. —

If the lower part of a beach was consolidated. & the land sunk a trifle. the appearances. here presented all round the Island woud probably be presented.

26 verso

Mem: Width of outer coral, as shown by width of Isld. —

Cause of fertility. —

Mem: Width of outer ... of fertility. —] pencil.

This document has been accessed 3787 times

Return to homepage

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

File last updated 14 October, 2013