Skip to main content

Full text of "The structure and distribution of coral reefs"

See other formats













Authorized Edition, 




For all that distinguishes the present from the second 
edition the reader has to thank Professor Bonney. 
He has added occasional footnotes (distinguished by 
square brackets), and he has given, in the form of an 
appendix, a careful summary of the more important 
memoirs published since 1874. 

My own contribution is merely the fulfilment of a 
pleasant duty — the expression of my sincere gratitude 
to Professor Bonney for the ready kindness with which 
he undertook a difficult task, and for the care and 
Bkill with which he has completed it. 

I must also be allowed the satisfaction of expressing 
my obligations to Captain Wharton, E.N., Hydro- 
grapher to the Admiralty, for an interesting series of 
notes, which are embodied by Professor Bonney in the 
present edition. 

Francis Darwin. 

CAMBRIDGE : February 28, 1889. 

^^ ^-'^^^oSS 




The fiest edition of this book appeared in 1842, and 
since then only one important work on the same 
subject has appeared, namely, in 1872, by Professor 
Dana, on Corals and Coral-Eeefs. In this work he 
justly says that I have not laid sufficient weight on 
the mean temperature of the sea, in determining the 
distribution of coral-reefs ; but neither a low tempera- 
ture nor the presence of mud-banks accounts, as it 
appears to me, for the absence of coral-reefs through- 
out certain areas ; and we must look to some more 
recondite cause. Professor Dana, also, insists that 
volcanic action prevents the growth of coral-reefs 
much more effectually than I had supposed; but 
how the heat or poisonous exhalations from a volcano 


can affect the whole circumference of a large island 
is not clear. Nor does this fact, if fully established, 
falsify my generalisation that volcanos in a state of 
action are not found within the areas of subsidence, 
whilst they are often present within those of elevation ; 
for I have not been influenced in my judgment by the 
absence or presence of coral-reefs round active volcanos ; 
I have judged only by finding upraised marme remains 
within the areas of elevation, and by the vicinity of 
atolls and barrier-reefs with reference to the areas 
of subsidence. Professor Dana apparently supposes 
(p. 320) that I look at fringing-reefs as a proof of 
the recent elevation of the land ; but I have ex- 
pressly stated that such reefs, as a general rule, 
indicate that the land has either long remained at the 
same level or has been recently elevated. Neverthe- 
less, from upraised recent remains having been found 
in a large number of cases on coasts which are fringed 
by coral-reefs, it appears that of these two alternatives 
recent elevation has been much more frequent than a 
stationary condition. Professor Dana further believes 
that many of the lagoon-islands in the Paumotu 
or Low Archipelago and elsewhere have recently 
been elevated to a height of a few feet, although 
originally formed during a period of subsidence ; but I 
shall endeavour to show in the sixth chapter of the 
present edition that lagoon-islands which have long 


remained at a stationary level often present the false 
appearance of having been slightly elevated. 

Although I thus demur to some of the remarks and 
criticisms made by this eminent naturalist, who has 
examined more coral formations than almost any other 
man, yet I do not the less admire his work.^ It has 
also afforded me the highest satisfaction to find that 
he accepts the fundamental proposition that lagoon- 
islands or atolls, and barrier-reefs, have been formed 
during periods of subsidence. 

The late Professor Jukes, in his account of the 
voj^age of H.M.S. Fly, published in 1847, devoted a 
chapter to the Barrier-Eeefs of Australia, and thus 
concludes : * After seeing much of the Great Barrier- 
reefs, and reflecting much upon them, and trying if it 
were possible by any means to evade the conclusions to 
which Mr. Darwin has come, I cannot help adding that 
his hypothesis is perfectly satisfactory to my mind, and 
rises beyond a mere hypothesis into the true theory of 

On the other hand, a distinguished naturalist. 
Professor Semper, differs much from me, although he 
seems willing to admit that some atolls and barrier- 
reefs have been formed in the manner in which I 
suppose. I will give in the Appendix, under the head 

' A friendly reply from Professor Dana, contesting some of the 
points mentioned above, will be found in Nature, Sept. 1874, p. 408. 


of the Pelew Islands, which were carefully examined 
by him, some account of his objections, and I will here 
only state that his view does not differ essentially from 
that of Chamisso, which will hereafter be discussed. 
It will be seen that the evidence in favour of atolls 
and barrier-reefs having been formed during sub- 
sidence is of a cumulative nature ; and that it is 
very difficult to judge with safety respecting any 
single lagoon-island or barrier-reef, or small group of 
them, even if the depth outside the reef and the slope 
of the encircled land are both known. 

In the present edition I have added some new facts 
and have revised the whole book ; the latter chapters 
having been almost re- written. The appended map of 
the Pacific and Indian Oceans remains in nearly the 
same state as before, for I have added only two red 
and two blue circles. I have removed an active vol- 
cano, which was formerly suj)posed to exist in Torres 
Straits. An account of a remarkable bar of sandstone 
off Pernambuco on the Brazilian coast has been added 
to the Appendix, as this bar is protected from the 
wear and tear of the waves by a coating of organic 
bodies, in the same manner as are most coral-reefs. 
It also resembles a coral-reef in shape or outline to a 
curiously deceptive degree. If I had been better 
situated during the last thirty years, for hearing of 
recent discoveries in the Pacific, and for consulting 


charts published in various countries, my map might 
have been greatly improved. But I hope that before 
long some one may be induced to colour a map on a 
large scale, on nearly the same principles as I have 
done, and in accordance with our advanced state of 
geographical knowledge ; for I believe that he would 
thus arrive at some new and striking generalisations. 

Down, Eeckenham, Kent : 
Febrtiary 1871. 



I SHALL HAVE OCCASION, in many parts of the follow^ 
ing volume, to acknowledge the valuable information 
I have received from several persons; but I must 
more particularly express my obligations to Captain 
E. Moresby, I.N., who conducted the survey of the 
Eed Sea, and of the archipelagoes of low coral-islands 
in the Indian Ocean. I beg, also, to be permitted to 
return my best thanks to Captain Beaufort, E.N., for 
having given me free access to the charts in the Ad- 
miralty, as weU as to Captain Beecher, E.N., for most 
kindly aiding me in consulting them. My thanks are 
likewise especially due to Captain Washington, E.N., 
for his invariable desire to assist me in every possible 
manner. Having in former publications had the 
pleasure of acknowledging how much I owe to Captain 


FitzEoy, for having permitted me to volunteer my 
services on board H.M.S. Beagle, and for his uniform 
kindness in giving me assistance in my researches, I 
can here only repeat my obligations to him. The 
materials for this volume were nearly ready two years 
ago ; but owing to ill health, its publication has been 
delayed. The two succeeding Parts — one on the vol- 
canic islands visited during the voyage of the Beagle, 
and the other on South America — will appear as soon 
as they can be prepared. 

May 2, 1842. 


Introduction page 1 



Corals on the outer margin — Zone of Nulliporaa — Exterior reef —Islets 
— Coral-conglomerate— Lagoon — Calcareous sediment— Scari and 
Holutlniria) subsisting on corals — Changes in the condition of tho 
reefs and islets — Probable subsidence of the atoU— Future state 
of the lagoons 7 to 27 


General form and size of atolls, their reefs and islets — External slope 
— Zone of Nulliporie — Conglomerate — Depth of lagoons — Sedi- 
ment — Eeefs submerged wholly or in part — Breaches in the reef 
— Ledge -formed shores round certain lagoons — Conversion of 
lagoons into land . . . . . . . . 27 to 43 



Maldiva Archipelago — Eing-formed reefs, marginal and central — 
Great depth in the lagoons of the southern atolls — Eeefs in the 
lagoons all rising to the surface — Position of islets and breaches 
in the reefs, with respect to the prevalent winds and action of the 
waves — Destruction of islets — Connection in the position and 
submarine foundation of distinct atolls — The apparent dissever- 
ment of large atolls — The Great Chagos Bank — Its submerged 
condition and extraordinary structure . . . . 43 to 55 




Closely resemble in general form and structure atoll-reefs — Width 
and depth of the lagoon-channels — Breaches through the reef in 
front of valleys, and generally on the leeward side — Checks to 
the filling up of the lagoon-channels — Size and constitution of 
the encircled islands — Number of islands within the same reef — 
Larrier-reefs of New Caledonia and Australia— Position of the 
reef relative to the slope of the adjoining land — Probable great 
thickness of barrier-reefs paok 56 to 08 



Eeefs of Mauritius— Shallow channel within the reef — Its slow 
tilling up — Currents of water formed within it — Upraised reefs 
— Narrow fringlng-reefs in deep seas — Eeefs on the coast of E. 
Africa and of Brazil — Fringing-reefs in very shallow seas, round 
banks of sediment and on worn-down islands — Fringing-reefs 
aO"ected by currents of the sea — Coral coating the bottom of the 
sea, but not forming reefs . . . . . . 69 to 79 





LIVE 108 to 118 



The atolls of the larger archipelagoes are not formed on submerged 
craters, or on banks of sediment— Immense areas interspersed 
with atolls— Their subsidence — The eilects of storms and earth- 
ijuakes on atolls — Recent changes in their state — The origin of 
barrier-reefs and of atolls — Their relative forms — The step-formed 


ledges and walls round the shores of some lagoons — The ring- 
formed reefs of the Maldiva atolls — The submerged condition of 
parts or of the whole of some annular reefs — The disseverment 
of large atolls — The union of atolls by linear reefs — The Great 
Chagos Bauk^-Objections, from the area and amount of subsi- 
dence required by the theory, considered — The probable composi- 
tion of the lower parts of atolls . . . page 119 to 157 



Description of the coloured map — Proximity of atolls and barrier- 
reefs— Relation in form and position of atolls with ordinary 
islands — Direct evidence of subsidence difficult to be detected — • 
Proofs of recent elevation where fringing-reefs occur — Oscilla- 
tions of level — Absence of active volcanos in the areas of subsi- 
dence — Immensity of the areas which have been elevated and 
have subsided — Their relation to the present distribution of the 
land — Areas of subsidence elongated, their intersection and alter- 
nation with those of elevation — Amount, and slow rate of sub- 
sidence — Eecapitulation 158 to 196 


Containing a detailed description of the Reefs and Islands in the 
coloured Map, Plate III 199 to 260 


[Summary of the principal contributions to the History of Coral- 
lieels since the year 1874] 281 to 8^2 

Genekal iNiiKs 333 


PLATE I. at end of Volume. 

In the several original surveys, from which the small plans on this 
plate have been reduced, the coral-reefs are engraved in very dif- 
ferent styles. For the sake of uniformity, I have adopted the 
style used in the charts of the Chagos Archipelago, published by 
the East India Company, from the survey by Capt. Moresby and 
Lieut. Powell. The surface of the reef, which dries at low water, 
is represented by a stippled surface with small crosses : the coral- 
islets on the reef are marked by small linear unstippled spaces, 
on which a few cocoa-nut trees, out of all proportion too large, 
have been introduced for the sake of clearness. The entire 
annular reef, which when surrounding an open expanse of water, 
forms an ' atoll,' and when surrounding one or more high islands, 
forms an encircling ' barrier-reef,' has a nearly uniform structure, 
and has been tinted, in order to catch the eye, of a pale blue 
colour. The reefs in some of the original surveys are represented 
merely by a single line with crosses, so that their breadth is not 
given ; I have had such reefs engraved of the width usually at- 
tained by coral-reefs. I have not thought it worth while to 
introduce all those small and very numerous reefs, which occur 
within the lagoons of most atolls and within the lagoon-channels 
of most barrier-reefs, and which stand either isolated, or are 
attached to the shores of the reef or land. At Peros Banhos 
none of the lagoon-reefs rise to the surface of the water ; a few 
of them have been introduced, and are marked by jjlain dotted 
circles. A few of the deepest soundings are laid down within 
each reef ; they are in fathoms, of six English feet. 

Fig. 1. — Vakikoko, situated in the western part of the S. Pacific; 

taken from the survey by Cajjt. D'Urville in the Astrolabe ; the 

scale is \ of an inch to a geographical mile ; the soundings on 

the southern side of the island, namely from 30 to 40 fathoms, 



are given from the Voyage of the Chev. Dillon ; the other sounl- 
ings are laid down from the survey by D'Urville ; hciglit of the 
summit of the island is 3,032 feet. The principal small detached 
reefs within the lagoon-channel have in this instance been repre- 
sented. The southern shore of the island is narrowly fringed by 
a reef ; if the engraver had carried this reef entirely round both 
islands, this figure would have served (by leaving out in imagina- 
tion the barrier-reef) as a good specimen of an abruptly-sided 
island, surrounded by a reef of the fringing class. 

Fig. 2. — HoGOLEU, or Rouo, in the Caroline Archipelago ; taken from 
the atlas of the Voyage of the Astrolabe, compiled from the 
surveys of Captains Duperrey and D'Urville ; scale ^ of an inch 
to a mile ; the depth of the immense lagoon-like space within the 
reef is not known. 

Fig. 3.— Baiatea, in the Society Archipelago; from the map given in 
the quarto edition of Cook's First Voyage; it is probably not 
accurate ; scale ^ of an inch to a mile. 

Fig. 4. — Bow, or Heyou atoll (or lagoon-island), in the Low Arch 
pelago ; from the survey by Capt. Beechey, R.N. ; scale ^j-, of an 
inch to a mile; the lagoon is choked up with reefs, but the 
average greatest depth of about 20 fathoms, is given from the 
published account of the voyage. 

Fig. 5.— BoLABOLA, in the Society Archipelago ; from the survey of 
Capt. Duperrey, in the Goquille ; scale ^ of an inch to a mile ; 
the soundings in this and the following figure have been altered 
from French feet to Enghsh fathoms ; height of highest point of 
the island 4,026 feet. 

Fig. G.— Maurua, in the Society Archipelago; from the survey by 
Capt. Duperrey in the Goquille; scale I- of an inch to a mile; 
height of land about 800 feet. 

Fig. 7.— PouYNiPKTE, or Seniavine, in the Caroline Archipelago; 
from the survey by Admiral Lutk6 ; scale \ of an inch to a mile. 

Fig. 8. — Gambier Islands, in the southern part of the Low Archi- 
pelago ; from the survey by Capt. Beechey ; scale ^ of an inch to 
a mile; height of highest island, 1,246 feet; the islands are sur- 
rounded by extensive and irregular reefs ; the reef on the southern 
aide is submerged. 

Fi^r. 9.— Peros Banhos atoll (or lagoon-island), in the ChagoH group 
in the Indian Ocean; from the survey by Capt. Moropby ami 


Lieut. Powell ; scale \ of an inch to a mile ; not nearly all tlio 
Bmall submerged reefs in the lagoon are represented ; the annular 
reef on the southern side is submerged. 

Fig. 10. — Keeling, or Cocos atoll (or lagoon-island), in the Indian 
Ocean ; from the survey by Capt. FitzPioy ; scale -^ of an inch to 
a mile ; the lagoon south of the dotted line is very shallow, and 
is left almost bare at low water; the part north of the line is 
choked up with irregular reefs. The annular reef on the N.W. 
side is broken, and blends into a shoal sand-bank, on which the 
Bea breaks. 

PLATE II. at end of Volume. 

Fig. 1. — Great Chagos Bank, in the Indian Ocean ; taken from the 
survey by Capt. Moresby and Lieut. Powell ; scale ^ of an inch 
to a mile (same scale as Hogoleu, in Plate I.) ; the parts which 
are shaded, with the exception of two or three islets on the 
western and northern sides, do not rise to the surface, but are 
submerged from 4 to 10 fathoms ; the banks bounded by the 
dotted lines lie from 15 to 20 fathoms beneath the surface, and 
are formed of sand ; the central space is of mud, and from 30 to 
50 fathoms deep. 

Fig. 2. — A vertical section, on the same scale, in an E. and W. line 
- across the Great Chagos Bank, given for the sake of exnibiting 
more clearly its structure. 

Fig. 3. — Menchicoff atoll (or lagoon-island), in the Marshall Archi- 
pelago, northern Pacific Ocean ; from Krusenstern's atlas of the 
Pacific ; originally surveyed by Capt. Hagemeister ; scale ^^ of an 
inch to a mile ; the depth within the lagoons is unknown. 

Fig. 4. — Mahlos Mahdoo atoll, together with Horsburgh atoll, in 
the Maldiva Archipelago ; from the survey by Capt. Moresby and 
Lieut. Powell ; scale 55 of an inch to a mile ; the white spaces in 
the middle of the separate small reefs, both on the margin and 
in the middle part, are meant to represent little lagoons ; but it 
vras found not possible to distinguish them clearly from the small 
iislets, which have been formed on these same small reefs ; many 
of the smaller reefs could not be introduced; the nautical mark 
(-1.) over the figures 250 and 200 between Mahlos Mahdoo and 
Horsburgh atoll and Powell's Island, signifies that soundings were 
not obtained at these depths. 


Fig. 5. — New Calkdonia, in the western part of the Pacific ; from 
Kiusenaterrib atlas, compiled from several surveys ; I have 
slightly altered the northern point of the reef, in accordance 
with the atlas of the Voyage of the Astrolabe. In Krusenstern's 
atlas, the reef is represented by a single line with crosses ; I have 
for the sake of uniformity added an interior line ; scale Jj of an 
inch to a mile. 

Fig. G. — Maidiva Archipelago, in the Indian Ocean ; from the survey 
by Capt. Moresby and Lieut. Powell; scale ^^ of an inch to a 

PLATE III. at beginning of Volume. 

The principles on which this map is coloured are explained in the 
beginning of Chapter VI. ; and the authorities for colouring each 
particular spot are detailed in the Appendix. The names printed 
in italics in the Index refer to the Appendix. 





The object of this volume is to describe from my own 
observation and the works of others, the principal 
kinds of coral-reefs, and to explain the origin of their 
peculiar forms. I shall not here treat of the poly- 
pifers, which construct these vast works, except as 
to their distribution, and the conditions favourable 
to their vigorous growth. 

Without any distinct intention to classify coral- 
reefs, most voyagers have spoken of them under the 
following heads : ' lagoon-islands ' or ' atolls,' ' barrier ' 
or 'encircling reefs,' and 'fringing' or 'shore reefs.' 
The lagoon-islands have received much the most atten- 
tion ; and it is not surprising, for everyone must be 
struck with astonishment, when he first beholds one of 


these vast rings of coral-rock, often many leagues in 
diameter, here and there surmounted by a low verdant 
island with dazzling white shores, bathed on the out- 
side by the foaming breakers of the ocean, and on the 
inside surrounding a calm expanse of water, which, 
from reflection, is generally of a bright but pale green 
colour. The naturalist will feel this astonishment 
more deeply after having examined the soft and almost 
gelatinous bodies of these apparently msignificant 
coral-polypifers, and when he knows that the solid reef 
increases only on the outer edge, which day and night 

No. 1. 

is lashed by the breakers of an ocean never at 
rest. Well did Francois Pyrard de Laval, in the 
year 1605, exclaim, * C'est une merueille de voir 
chacun de ces atollons, enuironne d'un grand banc de 
pierre tout autour, n'y ayant point d'artifice humain.' 
The above sketch of Whitsunday Island, in the 
S. Pacific, taken from Capt. Beechey's admirable 
Voyage, although excellent of its kind, gives but a 


faint idea of the singular aspect of one of these 
lagoon-islands. Whitsunday Island is of small size, 
and the whole circle has been converted into land, 
which is a comparatively rare circumstance. As the 
reef of a lagoon-island generally supports many sepa- 
rate small islands, the word ' island,' applied to the 
whole, is often the cause of confusion ; hence I have 
invariably used in this volume the term ' atoU,' which 
is the name given to these circular coral formations by 
their inhabitants in the Indian Ocean, and is syn- 
onymous with 'lagoon-island.' 

Barrier-reefs, when encircling small islands, have 
been comparatively little noticed by voyagers ; but 

No. 2. 

they well deserve attention. In their structure they 
are little less marvellous than atolls, and they give a 
singular and most picturesque character to the scenery 
of the islands they surround. In the accompanying 
sketch, taken from the Voyage of the Coquille, the reef 
is seen from within, from one of the high peaks of Bo- 


labola/ one of the Society Islands. Here, as in Whit- 
sunday Island, the whole of that part of the reef which 
is visible is converted into land. This is a circum- 
stance of rare occurrence ; more usually a snow-white 
line of great breakers, with here and there an islet 
crowned by cocoa-nut trees, separates the smooth 
waters of the lagoon-like channel from the waves of 
the open sea. The barrier reefs of Australia and of 
New Caledonia, owmg to their enormous dimensions, 
have excited much attention : in structure and form 
they resemble those encircling many of the smaller 
islands in the Pacific Ocean. 

With respect to fringing, or shore reefs, there is 
little in their structure which needs explanation ; and 
their name expresses their comparatively small ex- 
tension. They differ from barrier reefs in not lying 
far from the shore, and in not having within them a 
broad channel of deep water. Eeefs also occur around 
submerged banks of sediment and of worn-down rock ; 
and others are scattered quite irregularly where the 
sea is very shallow ; these are allied in most respects 
to fringing reefs, but are of comparatively httle 

I have given a separate chapter to each of the 
above classes, and have described some one reef or 
island, on which I possessed most information, as 
typical ; and have afterwards compared it with others 
of a like kind. Although this classification is useful 

' I have taken the liberty of simplifying the foreground, and 
leaving out a mountainous island in the far distance. 


from being obvious, and from including most of the 
coral-reefs existing in the open sea, it admits of a more 
fundamental division into barrier and atoll-formed 
reefs on the one hand, where there is a great apparent 
difficulty with respect to the foundation on which tbey 
must first have grown ; and into fringing reefs on the 
other, where, owing to the nature of the slope of the 
adjoining land, there is no such difficulty. The two 
blue tints and the red colour on the map (Plate III.) 
represent this main division, as explained in the be- 
ginning of the last chapter. In the Appendix, every 
existing coral-reef, except some on the coast of Brazil 
not included in the map, is briefly described in geo- 
graphical order, as far as I possessed information ; and 
any particular spot may be found by consulting the 

Several theories have been advanced to explain the 
origin of atolls or lagoon-islands, but scarcely one to 
account for barrier-reefs. From the limited depths at 
which reef-building polypifers can flourish, taken into 
consideration with certain other circumstances, we are 
compelled to conclude, as it will be seen, that both in 
atolls and barrier-reefs, the foundation to which the 
coral was primarily attached, has subsided ; and that 
during this downward movement, the reefs have grown 
upwards. This conclusion, it will be further seen, 
explains most satisfactorily, tbe outline and general 
form of atolls and barrier-reefs, and likewise certain 
peculiarities in their structure. The distribution, also, 
of the different kinds of coral-reefs, and their position 


with relation to the areas of recent elevation, and to 
the points subject to volcanic eruptions, fully accord 
with this theory of their origin.' 

' A brief account of my views on coral formations, now published 
ill my Journal of Kesearclics, was read May 31, 18:i7, before the Geo- 
luyical Society, and an abstract has appeared in the Proceedinga. 




Corals on the outer margin — Zone of Nulliporce— Exterior reef-' 
Islets — Coral-conglomerate — Lagoon — Calcareous sediment — ■ 
Scari and HoluthuricB subsisting on corals — Changes in the con- 
dition of the reefs and islets — Probable subsidence of the atoll — 
Future state of the lagoon. 

Keeling or Cocos atoll is situated in the Indian Ocean, 
in 12° 5' S. and long. 90° 55' E. : a reduced chart of it, 
from the survey of Capt. EitzRoy and the officers of 
H.M.S. Beagle, is given in Plate I. fig. 10. The 
greatest width of this atoll is nine miles and a half. 
Its structure is in most respects characteristic of the 
class to which it belongs, with the exception of the 
shallowness of the lagoon. The accompanying wood- 
cut (No. 3) represents a vertical section, supposed to be 
drawn at low water from the outer coast across one of 
the low islets (one being taken of average dimensions) 
to within the lagoon. The section is true to the scale 
in a horizontal line, but it could not be made so in a 
vertical one, as the average greatest height of the land 
is only between six and twelve feet above high- water 

8 ATOLLS. Cii. I. 

mark. I will describe the section, commencing with, 
the outer margin. But I must first observe that the 
reef-building polypifers, not being tidal animals, require 
to be constantly submerged or washed by the breakers. 
I was assured by Mr. Liesk, an intelligent resident on 
these islands, as well as by some chiefs at Tahiti (Ota- 
No. 3. 

150 Yttrcls 

A — Level of the sea at low water : where the letter A is placed, 
the depth is 25 fathoms, and the distance rather more than 150 yards 
from the edge of the reef. 

J5 — Outer edge of that flat part of the reef, which dries at low 
water: the edge either consists of a convex mound, as represented, 
or of rugged ^joints, like those a little farther seaward, beneath the 

C — A flat of coral -rock, covered at high water. 

D — A low projecting ledge of brecciated coral-rock, washed by the 
waves at high water. 

E — A slope of loose fragments, reached by the sea only during 
gales : the upper part, which is from six to twelve feet high, is 
clothed with vegetation. The surface of the islet gently slopes to 
the lagoon. 

F — Level of the lagoon at low water. 

neite) , that an exposure to the rays of the sun for a very 
short time invariably causes their destruction.* Hence 

' [This v.-ould be true of certain genera or species, but according 
to the observations made during the voyage of the Challenger (Re- 
ports xvi. pp. 23, 3f)), and by Mr. Guppy (Proc. E. S. Edin. xiii. p. 8(33), 
there are some which can bear exposure for a limited time, perhaps 
two or three hours, especially if occasionally wetted by the spray.] 


it is possible only under the most favourable circum- 
stances, afforded by an unusually low tide and smooth 
water, to reach the outer margin, where the coral is 
alive. I succeeded only twice in gaining this part, and 
found it almost entirely composed of a living Porites, 
which forms great irregularly rounded masses (like 
those of an Astrfea, but larger) from four to eight feet 
broad, and little less in thickness. These mounds are 
separated from each other by narrow crooked channels, 
about six feet deep, most of which intersect the line of 
reef at right angles. On the furthest mound, which I 
was able to reach by the aid of a leaping-pole, and over 
which the sea broke with some violence, although the 
day was quite calm and the tide low, the polypifers in 
the uppermost cells were all dead, but between three 
and four inches lower down on its side they were living, 
and formed a projecting border round the upper and 
dead surface. The coral being thus checked in its up- 
ward growth, extends laterally, and hence most of the 
masses, especially those a little further inwards, had 
broad flat dead summits. On the other hand I could 
see, during the recoil of the breakers, that a few yards 
further seaward, the whole convex surface of the 
Porites was alive : so that the point where we were 
standing was almost on the exact upward and shore- 
ward limit of existence of those corals which form the 
outer margin of the reef. We shall presently see 
that there are other organic productions, fitted to bear 
a somewhat longer exposure to the air and sun. 

Next, but much inferior in importance to the 

10 ATOLLS. Cn. T. 

Pontes, is the Millepora complanata} It gro^YS in 
thick vertical plates, intersectiiig each other at various 
angles, and forms an exceedingly strong honey-combed 
mass, which generally assumes a circular form, the 
marginal plates alone being alive. Between these plates 
and in the protected crevices on the reef, a multitude 
of branching zoophytes and other productions flourish, 
but the Porites and Millepora alone seem able to resist 
the fury of the breakers on its upper and outer edge ; 
at the depth of a few fathoms other kinds of stony 
corals live. Mr. Liesk, who was intimately acquainted 
with every part of this reef, and likewise with that of 
North Keelmg atoll, assured me that these corals in- 
variably compose the outer margin. The lagoon is 
inhabited by quite a distinct set of corals, generally 
brittle and thinly branched ; but a Porites, apparently 
of the same species with that on the outside, is found 
there, although it does not seem to thrive, and cer- 
tainly does not attain the thousandth part in bulk of 
the masses opposed to the breakers. 

The wood-cut (No. 3) shows the form of the bot- 
tom outside the reef : the water deepens very gradually 
for a space of between one and two hundred yards 
wide, to a depth of 25 fathoms {A in section), beyond 
which the sides plunge into the unfathomable ocean 
at an angle of 45° 2. To the depth of ten or twelve 

' This Millepora, (Palmipora of Blainvillo,) as -well as the IT. alci- 
cornis, possesses the singular property of stinging the skin where it 
is delicate, as on the face and arm. 

* The soundings from which this section is laid down were taken 
with great care by Captain FitzEoy himself: he used a bell-shaped 



fathoms, the bottom is exceedingly rugged and seems 
formed of great masses of Kving coral, similar to 
those on the margin. The arming of the lead here 
invariably came up quite clean, but deeply indented, 
and chains and anchors which were lowered, in the 
hopes of tearing up the coral, were broken. Many 
small fragments, however, of MiUcpora alcicornis 
were brought up ; and on the arming from an eight- 
fathom cast, there was a perfect impression of an 
Astrsea, apparently alive. I examined the rolled 
fragments cast on the beach during gales, in order 
further to ascertain what corals grew outside the reef. 
The fragments consisted of many kinds, of which the 
Porites already mentioned and a Madrepora, appa- 
rently the M. corymhosa, were the most abundant. As I 
searched in vain in the hollows on the reef and in the 
lagoon, for a living specimen of this Madrepore, I con- 
clude that it is confined to a zone outside, and beneath 
the surface, where it must be very abundant. Frag- 
ments of the Milleijora alcicornis and of an Astraea 
were also numerous ; the former is found, but not in 
proportionate numbers, in the hollows on the reef ; but 
the Astrsea I did not see living. Hence we may infer, 
that these are the hinds of coral which form the 
rugged sloping surface (represented in the wood-cut 

lead, having a diameter of four inches, and the armings each time 
were cut off and brought on board for me to examine. The arming 
is a preparation of tallow, placed in the concavity at the bottom of 
the lead. Sand, and even small fragments of rock will adhere to it ; 
and if the bottom be of rock, it brings up an exact impression of ita 

12 ATOLLS. Ch. T. 

by an uneven line) round and beneath the external 
margin. Between 12 and 20 fathoms the arming camo 
up an equal number of times smoothed with sand, and 
indented with coral : an anchor and lead were lost at 
the respective depths of 13 and 16 fathoms. Out of 
twenty-five soundings taken at a greater depth than 
20 fathoms, every one showed that the bottom was 
covered with sand ; whereas at a less depth than 
12 fathoms, every sounding showed that it was 
exceedingly rugged, and free from all extraneous 
particles. Two soundings were obtained at the depth 
of 360 fathoms, and several between 200 and 300 
fathoms. The sand brought up from these depths 
consisted of finely triturated fragments of stony 
zoophytes, but not, as far as I could distinguish, of a 
particle of any lamelliform genus : fragments of 
shells were rare. 

At a distance of 2,200 yards from the breakers, 
Captain FitzEoy found no bottom with a line 7,200 
feet in length ; hence the submarine slope of this coral 
formation is steeper than that of any volcanic cone. 
Off the mouth of the lagoon, and likewise off the 
northern point of the atoll, where the currents act 
violently, the inclination, owing to the accumulation of 
sediment, is less. As the arming of the lead from all 
the greater depths showed a smooth sandy bottom, I at 
first concluded that the whole consisted of a vast conical 
pile of calcareous sand, but the sudden increase of depth 
at some points, and the fact of the line having been 
cut, when between 500 and 600 fathoms were out, 


indicates the probable existence of submarine cliffs 
of rock. 

On the margin of the reef, close within the line 
■where the upper surface of the Porites and of the 
Mihepora is dead, three species of Nullipora flourish. 
One grows in thin sheets, like a Hchen on old trees ; the 
second in stony knobs, as thick as a man's finger, 
radiating from a common centre ; and the third, which 
is less common, in a moss-like reticulation of thin, but 
perfectly rigid branches.' The three species occur 
either separately or mingled together ; and they form 
by their successive growth a layer two or three feet in 
thickness, which in some cases is hard, but where formed 
of the lichen-like kind, readily yields an impression to 
the hammer : the surface is of a reddish colour. These 
NulliporEe, although able to exist above the limit of 
true corals, seem to require to be bathed during the 
greater part of each tide by breaking water, for they are 
not found in any abundance in the protected hollows on 
the back part of the reef, where they might be immersed 
during either the whole or an equal proportional time 
of each tide. It is remarkable that organic productions 
of such extreme simplicity, for the Nulliporae undoubt- 
edly belong to one of the lowest classes of the vegetable 
kingdom, should be limited to a zone so peculiarly cir- 

' This last species is of a beautiful bright peach-blossom colour. 
Its branches are about as thick as crow-quills ; they are shghtly 
flattened and knobbed at the extremities. The extremities only are 
alive and brightly coloured. The two other species are of a dirty 
purplish white. The second species is extremely hard; its short 
knob-like branches are cylindrical, and do not grow thicker at their 

14 ATOLLS. Crr. L 

cumstanced. Hence the layer composed by their growth, 
merely fringes the reef for a space of about 20 yards in 
"width, either under the form of separate mammillated 
projections, where the outer masses of coral are separate, 
or more commonly, where the corals are united into a 
solid margin, as a continuous smooth convex mound 
(B in wood-cut) like an artificial breakwater. Both 
the mound and mammillated projections stand about 
three feet higher than any other part of the reef, by 
which term I do not include the islets, formed by the 
accumulation of rolled fragments. We shall hereafter 
see that other coral reefs are protected by a similar 
thick growth of NulliporaB on the outer margin, the part 
most exposed to the breakers, and this must effectually 
aid in preserving it from being worn down. 

The wood-cut (at p. 8) represents a section across 
one of the islets on the reef, but if all that part which 
is above the levol of C were removed, the section 
would be that of the reef, as it occurs where islets 
have not been formed. It is this reef which essen- 
tially forms the atoll. In Keeling atoll the ring 
encloses the lagoon on aU sides except at the northern 
end, where there are two open spaces, through one 
of which ships can enter. The reef varies in width 
from 250 to 500 yards ; its surface is level, or very 
slightly inclined towards the lagoon, and at high-tide 
the sea breaks entirely over it : the water at low tide 
thrown by the breakers on the reef, is carried by the 
many narrow and shoal gullies or channels on its sur- 
face, into the la,goon : a return stream sets out of the 


lagoon tliroiTgh the main entrance. The most frequent 
coral in the hollows on the reef is Pocillopora ver- 
na'osa, which grows in short sinuous plates, or branches, 
and when alive is of a beautiful pale lake-red : a Mad- 
repora, closel}^ allied or identical with M. pocillifera, is 
also common. As soon as an islet is formed, and the 
waves are prevented from breaking entirely over the 
reef, the channels and hollows become filled up with 
fragments cemented together by calcareous matter; and 
the surface of the reef is converted into a hard smooth 
floor (C of wood-cut), like an artificial one of free- 
stone. This flat surface varies in width from 100 to 
200, or even 300 yards, and is strewed with a few large 
fragments of coral torn up during gales : it is uncovered 
only at low water. I could with difficulty, and only 
by the aid of a chisel procure chips of rock from 
its surface, and therefore could not ascertain how 
much of it is formed by the aggregation of detritus, 
and how much by the outward growth of mounds of 
corals, similar to those now living on the margin. No- 
thing can be more singular than the appearance at low 
tide of this ' flat ' of naked stone, especially where it is 
externally bounded by the smooth convex mound of 
Nulliporse, appearing like a breakwater built to resist 
the waves, which are constantly throwing over it sheets 
of foaming water. The characteristic appearance of 
this ' flat ' is shown in the foregoing wood-cut of Whit- 
sunday Atoll. 

The islets on the reef are first formed between 200 
and 300 yards from its outer edge, through the accu- 

16, ATOLLS. Cir. I. 

m Illation of a pile of fragments, thrown together by 
Bome unusually strong gale. Their ordinary width is 
under a quarter of a mile, and their length varies from 
a few yards to several miles. Those on the S.E. and 
wmdward side of the atoll, increase solely by the addi- 
tion of fragments on their outer side ; hence the loose 
blocks of coral, of which their surface is composed, as 
well as the shells mingled with them, almost exclu- 
sively consist of those kinds which live on the outer 
coast. The highest part of the islets (excepting hil- 
locks of blown sand, some of which are 30 feet high), is 
close to the outer beach {E of the wood-cut) and aver- 
ages from six to ten feet above ordinary high-water 
mark. From the outer beach the surface slopes gently 
to the shores of the lagoon ; and this slope no doubt 
is due to the breakers, the further they have rolled 
over the reef, having had less power to throw up 
fragments. The little waves of the lagoon heap up 
sand and fragments of thinly-branched corals on the 
inner side of the islets on the leeward side of the atoll ; 
and these islets are broader than those to wmdward, 
some being even 800 yards in width ; but the land thus 
added is very low. The fragments beneath the surface 
are cemented into a solid mass, which is exposed as a 
ledge (D of the wood-cut), projecting some yards in 
front of the outer shore, and from two to four feet high. 
This ledge is just reached by the waves at ordinary 
high- water : it extends in front of all the islets, and 
everywhere has a water- worn and scooped appearance. 
The fragments of coral which are occasionaUy cast on 


the * flat ' are during gales of unusual violence swept 
together on the beach, where the waves each day at 
high-water tend to remove and gradually wear them 
down ; but the lower fragments are firmly cemented 
together by percolated calcareous matter, and they resist 
the daily tides longer than the loose upper fi-agments ; 
and thus a projecting ledge is formed. The cemented 
mass is generally of a white colour, but in some few 
parts reddish from ferruginous matter : it is very hard 
and sonorous under the hammer : it is obscurely divided 
by seams, dipping at a small angle seaward : it consists 
of fragments of the corals which grow on the outer 
margin, some quite and others partially rounded, some 
small and others between two and three feet across ; and 
of masses of previously formed conglomerate, torn up, 
rounded, and recemented : or it consists of a calcareous 
sandstone, entirely composed of rounded particles of 
shells, corals, the spines of echini, and other organic 
bodies generally almost blended together ; — rocks, 
of this latter kind, occur on many shores, where there 
are no coral-reefs. The structure of the coral in the 
conglomerate has generally been much obscured by the 
infiltration of spathose calcareous matter ; and I col- 
lected an interesting series, beginning with fragments 
of unaltered coral, and ending with others, where it was 
impossible to discover with the naked eye any trace of 
organic structure. In some specimens I was unable, 
even with the aid of a lens, and by wotting them, to 
distinguish the boundaries of the altered coral and 
spathose limestone. Many even of the blocks of coral 

18 ATOLLS. Cii. T. 

lying loose on the beach, had their central parts altered 
and infiltrated.^ 

The lagoon alone remains to be described ; it is 
much shallower than that of most atolls of considerable 
size. The southern part is almost filled up with banks 
of mud and fields of coral, both dead and alive ; but 
there are considerable spaces, from three to four 
fathoms, and smaller basins from eight to ten fathoms 
deep. Probably about half its area consists of sedi- 
ment, and half of coral-reefs. The corals composing 
these reefs have a very different aspect from those on 
the outside : they are numerous in kind, and most 
of them are thinly branched. Meandrina, however, 
lives in the lagoon, and many great rounded masses 
of this coral lie loose or almost loose on the bottom. 
The other most common species are three closely 
allied species of true Madrepora with thin branches ; 
Seriatajjora suhulata; two species of Porites^ with 
cylindricalbranches, one of which forms circular clumps, 
with only the exterior branches alive; and lastly, a 
coral something like an Explanaria, but with stars on 
both surfaces, growing in thin, brittle, stony, foliaceous 

' [Dead coral still lying on the beach has been found to contain 
at least 5 per cent, of carbonate of magnesia, though only & very 
small quantity is present in fresh coral (usually less than 1 per cent.). 
In old coral-rock as much as 3807 per cent, has been found (Dana's 
Corals and Coral Islands, ch. vi. § 9).] 

- This Porites has somewhat the habit of P. clavaria, but the 
branches are not knobbed at their ends. When alive it is of a 
yellow colour, but after having been washed in fresh water and 
placed to dry, a jet-black slimy substance exuded from the entire 
Furfacft, so that the specimen now appears as if it had boon dipped 


expansions, especially in the deeper basins of the 
lagoon. The reefs on which these corals grow are 
very irregular in form, are full of cavities, and have 
not a solid flat surface of dead rock, like that surround- 
ing the lagoon ; nor can they be nearly so hard, for 
the inhabitants by the aid of crowbars made a channel 
of considerable length through these reefs, in which a 
schooner, built on the S.E. islet, was floated out. It is 
a very interesting circumstance, pointed out to us by 
Mr. Liesk, that this channel, although made less than 
ten years before our visit, was then, as we saw, almost 
choked up with living coral, so that fresh excavations 
would be absolutely necessary to aUow another vessel 
to pass through it. 

The sediment from the deepest parts in the lagoon, 
when wet, appeared chalky, but when dry, like very fine 
sand. Large soft banks of similar, but even finer 
grained mud, occur on the S.E. shore of the lagoon, 
affording a thick growth of a Fucus, on which turtle 
feed; this mud, although discoloured by vegetable 
matter, appears from its entire solution in acids to be 
purely calcareous. I have seen in the Museum of the 
Geological Society, a similar but more remarkable sub- 
stance, brought by Lieut. Nelson from the reefs of 
Bermuda, which, when shewn to several experienced 
geologists, was mistaken by them for true chalk. On 
the outside of the reef much sediment must be formed 
by the action of the surf on the rolled fragments of 
coral ; but, in the calm waters of the lagoon, this can 
take place only in a small degree. There are, however, 

20 ATOLLS. Cn. t 

other and unexpected agents at work here : large shoala 

of two species of Scarus, one inhabiting the surf outside 
the reef and the other the lagoon, subsist entirely, as I 
was assured bj^ Mr. Liesk, the intelligent resident before 
referred to, by browsing on the living polypifers. I 
opened several of these fish, which are very numerous 
and of considerable size, and I found their intestines 
distended by small pieces of coral, and finely ground cal- 
careous matter. This must daily pass from them as the 
finest sediment ; much also must be produced by the 
infinitely numerous vermiform and molluscous animals 
which make cavities in almost every block of coral. Dr. 
J. Allan of Forres, who has enjoyed the best means of 
observation, informs me in a letter, that the Holuthuriae 
(a family of Eadiata), subsist on living coral ; ^ and the 
singular structure of bone within the anterior extremity 
of their bodies, certainly appears well adapted for this 
purpose. The number of the species of Holuthuria, 
and of the individuals which swarm on every part of 
these coral-reefs, is extraordinarily great; and many 
ship-loads are, as is well known, annually fi*eighted 
for China with trepan g, which is a species of this 
genus. The amount of coral yearly consumed, and 
ground down into the finest mud, by these several crea- 
tures, and probably by many other kinds, must bo 
immense. These facts are, however, of more importance 

• [Mr. Guppy, Proc. E. S. Edin. xiii. p. 894, expresses the opinion 
that the Holothurians do not subsist on the living coral, but obtain 
nutiimont from swallowing the sand and detrital material, of which 
broken coral forms a large constituent.] 


in another point of view, as showing us that there are 
Uving checks to the growth of coral-reefs, and that the 
ahnost universal law of ' consume and be consumed,' 
holds good even with the polypifers forming those 
massive bulwarks, which are able to withstand the force 
of the open ocean. 

Considering that Keeling atoll, like other coral for- 
mations, has been entirely formed by the growth of 
organic beings, and the accumulation of their detritus, 
one is naturally led to enquire, how long it has con- 
tinued, and how long it is likely to continue, in its 
present state. Mr. Liesk informed me that he had 
Been an old chart in which the present long island on 
the S.E. side was divided by several channels into as 
many islets ; and he assures me that the channels can 
still be distinguished by the smaller size of the trees 
on them. On several islets, also, I observed that only 
young cocoa-nut trees were growing on the extremities, 
and that older and taller trees rose in regular succession 
behind them : which shows that these islets have very 
lately increased in length. In the upper and south- 
eastern part of the lagoon, I was much surprised by 
finding an irregular field of at least a mile square of 
branching corals, still upright, but entirely dead. They 
consisted of the species already mentioned ; they were 
of a brown colour, and so rotten, that in trying to 
stand on them, I sank half way up the leg, as if 
through decayed brushwood. The tops of the branches 
were barely covered by water at the time of lowest tide. 
Several facts having led me to disbelieve in any elev?»- 

22 ATOLLS. Cii. I. 

tion of the whole atoll, I was at first unable to imagine 
what cause could have killed so large a licld of coral. 
Upon reflection, however, it appeared to me that the 
closing up of the above mentioned channels would be 
a sufficient cause ; for before this, a strong breeze 
by forcing water through them hito the head of the 
lagoon, would tend to raise its level. But now this 
cannot happen, and the inhabitants observe that tlie 
tide rises to a less height, during a high S.E. wind, at 
the head than at the mouth of the lagoon. The corals, 
which, under the former condition of things, had at- 
tained the utmost possible limit of upward growth, 
would thus occasionally be exposed for a short time to 
the sun, and be hilled. 

Besides the increase of drj'' land, indicated by the 
foregoing facts, the exterior solid reef appears to have 
grown outwards. On the western side of the atoll, the 
* flat ' lying between the margin of the reef and the 
beach, is very wide : and in front of the regular beach 
with its conglomerate basis, there is, in most parts, a 
bed of sand and loose fragments with trees growing 
out of it, which apparently is not reached even by the 
spray at high water. It is evident some change has 
taken piace since the waves formed the inner beach : 
that they formerly beat against it with violence was 
evident, from a remarkably thick and water- worn point 
of conglomerate at one spot, now protected by vegeta- 
tion and a bank of sand ; that they beat against it in 
the same peculiar manner in which the swell from 
windward now obliquely curls round the margin of the 


reef, was evident from the conglomerate having been 
worn into a point projecting from the beach in a simi- 
larly oblique manner. This retreat in the line of 
action of the breakers may have resulted, either from 
the surface of the reef in front of the islets having 
formerly been submerged, and afterwards having been 
raised by accumulated fragments, or from the mounds 
of coral on the margin having grown outwards. That 
an outward growth of this part is in process, can 
hardly be doubted from the existence of the mounds 
of Porites with their summits apparently lately killed, 
and their sides only three or four inches lower down 
thickened by a fresh layer of living coral. But there 
is a difficulty in this supposition which I must nob 
pass over. If the whole, or a large part of the ' fiat,* 
had been formed by the outward growth of the margin, 
each successive margin would naturally have been 
coated by the NuUipora), and so much of the surface 
would have been of equal height with the existing zone 
of living NuiliporoB : this is not the case, as may be 
seen in the wood-cut. It is, however, evident from 
the abraded state of the ' flat,' with its original ine- 
qualities filled up, that its surface has been much 
modified ; and it is possible that the inner portions of 
the zone of Nulliporse, perishing as the reef grows out- 
wards, might be worn down by the surf. If this has 
not taken place, the reef can in no part have increased 
outwards in breadth since its formation, or at least 
since the Nulliporas formed the convex mound on its 
margin : for the zone thus formed, which stands 

24 ATOLLS. Ch. L 

between two and three feet above the other parts of 
the reef, is nowhere much above twenty yards in 

Tims far we have considered facts, which indicate, 
with more or less probability, an increase in the 
diameter of the atoll ; but there are others having an 
opposite tendency. On the S.E. side, Lieut. Sulivan, 
to whose kindness I am indebted for many interesting 
observations, found the conglomerate (D, in wood-cut 
p. 8) projecting on the reef nearly fifty yards in front of 
the islets : we may infer from what we elsewhere see 
that the conglomerate was not originally so much 
exposed, but formed the base of an islet, the front and 
upper part of which has since been swept away. The 
degree to which the conglomerate, round nearly the 
whole atoll, has been scooped, broken up, and the frag- 
ments cast on the beach, is certainly very surprising, 
even on the view that it is the office of occasional galea 
to pile up fragments, and of the daily tides to wear 
them away. On the western side, also, of the atoll, 
where I have described a bed of sand and fragments 
with trees growing out of it, in front of an old beach, 
it struck both Lieut. Sulivan and myself, from the 
manner in which the trees were being washed down, 
that the surf had lately recommenced an attack on this 
Ime of coast. Appearances indicating a slight en- 
croachment of the water on the land, are plainer within 
the lagoon: I noticed in several places, both on its 
windward and leeward shores, old cocoa-nut trees 
falling with their roots undermined, and the rotten 


stumps of others on the beach, where the inhabitants 
assured us the cocoa-nut could not now grow. Capt. 
FitzPioy pointed out to me, near the settlement, the 
foundation posts of a shed, now washed by every tide, 
but which the inhabitants stated, had seven years 
before stood above high water-mark. In the calm 
waters of the lagoon, directly connected with a great, 
and therefore stable ocean, it seems very improbable 
that a change in the currents, sufficiently great to 
cause the water to eat into the land on all sides, should 
have taken place within a limited period. From these 
considerations I inferred, that probably the atoll had 
lately subsided to a small amount ; and this inference 
was strengthened by the circumstance, that in 1834, 
two years before our visit, the island had been shaken 
by a severe earthquake, and by two slighter ones during 
the ten previous years. If, during these subterranean 
disturbances, the atoll did subside, the downward 
movement must have been very small, as we must con- 
clude from the fields of dead coral still lipping the 
surface of the lagoon, and from the breakers on the 
western shore not having yet regained the line of their 
former action. The subsidence most, also, have been 
preceded by a long period of rest, during which the 
islets extended to their present size, and the living 
margin of the reef grew either upwards, or as I believe 
outwards, to its present distance from the beach. 

Whether this view be correct or not, the above 
facts are worthy of attention, as showing how severe a 
struggle is in progress on these low coral-formations 

26 ATOLLS. Cn. L 

between the two nicely balanced powers of land and 
^Yater. With respect to the future state of Keeling 
atoll, if left undisturbed, we can see that the islets 
may still extend in length ; but as they cannot resist 
the surf until it is broken by rolhng over a wide space, 
their increase in breadth must depend on the in- 
creasing breadth of the reef; and this must be limited 
by the steepness of the submarine flanks, which can be 
added to only by sediment derived from the wear and 
tear of the coral. From tlie rapid growth of the coral 
in the channel cut for the schooner, and from the 
several agents at work in producing fine sediment, it 
might be thought that the lagoon would necessarily 
become quickly filled up. Some of this sediment, 
however, is transported mto the open sea, as appears 
from the soundings off the mouth of the lagoon, in- 
stead of being deposited within it. The deposition, 
moreover, of sediment, checks the growth of coral reefs, 
BO that these two agencies cannot act together with full 
effect in filling up the lagoon. "VVe know so little of 
the habits of the many different species of corals which 
form the lagoon-reefs, that we have no more reason for 
supposing that their whole surface would grow up as 
quickly as the coral did in the schooner-channel, than 
for supposing that the whole surface of a peat-moss 
would increase as quickly as parts are known to do in 
boles, where the peat has been cut away. These 
agencies, nevertheless, tend to fill up the lagoon ; but 
in proportion as it becomes shallower, so must the 
polypifers be subject to many injurious agencies, such 

Sfct. ti. atolls. 27 

as impure water and loss of food. For instance, Mr. 
Liesk informed me, that some years before our visit 
unusually heavy rain killed nearly all the fish in the 
lago(m, and probably the same cause would likewise 
injure the corals. The reefs also, it must be remem- 
b3red, cannot possibly rise above the level of the 
lowest spring-tide, so that the final conversion of the 
lagoon into land must be due to the accumulation of 
sediment : and in the midst of the clear water of the 
ocean, and with no surrounding high land, this process 
must be exceedmgly slow. 


General form and size of atolls, their reefs and islets — External 
slope — Zone of Nulliporce — Conglomerate — Depth of lagoons — ■ 
Sediment — Beefs submerged wholly or in part — Breaches in the 
reef — Ledge-formed shores round certain lagoons — Conversion of 
lagoons into lai d. 

I WILL here give a sketch of the general form and 
structure of the many atolls and atoll-formed reefs 
which occur in the Pacific and Indian oceans, compar- 
ing them with Keeling atoll. The Maldiva atolls and 
the Great Chagos Bank differ in so many respects, that 
I shall devote to them, besides occasional references, a 
third section of this chapter. Keeling atoll may bo 
considered as of moderate dimensions and of regular 
form. Of the thirty-two islands surveyed by Capt. 
Beechey in the Low Archipelago, the longest was found 
to be thirty miles, and the shortest less than a mile ; 
but Vliegen atoll, situated in another part of the same 

28 ATOLLS. Cn. I. 

group, appears to be sixty miles long and twenty broad. 
Most of the atolls in this group are of an elongated 
form ; thus Bow Island is thirty miles in length, and 
on an average only six in width (See Fig. 4, Plate I.), 
and Clermont Tonnere has nearly the same proportions. 
In the Marshall Archipelago (the Ealick and Radack 
group of Kotzchue) several of the atolls are more than 
thirty miles in length, and Rimsky Korsacoff is fifty- 
four long, and twenty wide at the broadest part of its 
^regular outline. Most of the atolls in the Maldiva 
Archipelago are of great size, one of them (which, how- 
ever, bears a double name), measured in a medial and 
slightly curved line, is no less than eighty-eight geo- 
graphical miles long, its greatest width being under 
twenty, and its least only nine and a half miles. Some 
atolls have spurs projecting from them ; and in the 
Marshall group there are atolls united together by 
linear reefs, for instance Menchioff Island (See Fig. 3, 
Plate II.), which is sixty miles in length, and consists 
of three loops tied together. In far the greater num- 
ber of cases an atoll consists of a simple elongated ring, 
with its outline moderately regular. 

The average width of the annular reef may be taken 
at about a quarter of a mile. Capt. Beechey ' says 
tliat in the atolls of the Low Archipelago it exceeded 
in no instance half a mile. The description given of 
the structure and proportional dimensions of the reef 
and islands of Keeling atoll, appears to apply perfectly 
to nearly all the atolls in the Pacific and Indian 

' Beechey'a Voyage to the Pacific and Behring's Straits, chap. viii. 

Sect. II. ATOLLS. 29 

oceans. The islets are first formed some way back 
either on the projecting points of the reef, especially if 
its form be angular, or on the sides of the main en- 
trances into the lagoon — that is in both cases, on points 
where the breakers can act during gales of wind in 
somewhat different directions, so that the matter 
thrown up from one side may accumulate against that 
before thrown up from another. In Lutke's chart of 
the Caroline atolls, w^e see many instances of the former 
case ; and the occurrence of islets, as if placed for 
beacons, on the points where there is a gateway or 
breach through the reef, has been noticed by several 
authors. There are some atoll-formed reefs, rising to 
the surface of the sea and partly dry a{ low water, on 
which from some cause islets have never been formed; 
and there are others, on which they have been formed, 
but have subsequently been worn away. In atolls of 
small dimensions the islets frequently become united 
into a single horse-shoe or ring-formed strip ; but 
Diego Garcia, although an atoll of considerable size, 
being thirteen miles and a half in length, has its 
lagoon entirely surrounded, except at the northern end, 
by a belt of land, on an average a third of a mile in 
width. To show how small the total area of the annu- 
lar reef and the land is in islands of this class, I may 
quote a remark from the voyage of Lutke, namely, that 
if the forty-three rings, or atolls, in the Caroline Archi- 
pelago were put one within another, and over a steeple 
in the centre of St. Petersburg, the whole would not 
cover that city and its suburbs. 

30 ATOLLS. Cii. I. 

The form of the bottom, as given by Captain 
Beechey in his sections of the atolls in the Low 
Archii)elago, exactly coincides with that already de- 
scribed in Keeling atoll : it gradually slopes to about 
twenty fathoms, at the distance of between one and 
two hundred yards from the edge of the reef, and 
then plunges at an angle of 45° into unfathomabhi 
depths.' The nature, however, of the bottom seems 
to differ, for this officer'^ informs me that all the 
soundings, even the deepest, were on coral, but he 
does not know whether dead or alive. The slope 
round Christmas atoll (Lat. 1°4' N., 157° 45' W.), de- 
scribed by Cook,^ is considerably less ; at about half a 
mile from the edge of the reef, the average depth was 
about fourteen fathoms on a fine sandy bottom, and at 
a mile, only between twenty and forty fathoms. It has 
no doubt been owing to this gentle slope, that the strip 
of land surrounding its lagoon, has increased in one 
part to the extraordinary width of three miles ; it is 
formed of successive ridges of broken shells and corals, 
like those on the beach. I know of no other instance 
of such width in the reef of an atoll ; but Mr. F. D. 

• The elope of the bottom round the Marshall atolls in the 
Northern Pacific is probably similar: Kotzebue (First Voyage, vol, ii. 
p. 1(5) says, ' We had at a small distance from the reef, forty fatlioms 
depth, which increased a little further so much that we could find no 

■■' I must be permitted to express my obligation to Captain Beechey, 
for the very kind manner in which he has given me information on 
several points, and to own the great assistance I have derived from 
his excellent published work. 

* Cook's Third Voyage, vol. ii. chap. 10. 

Sect. II. ATOLLS. oi 

Bennett informs me that the inclination of the hottom 
round Carohne atoll in the Pacific, is like that off 
Christmas island, very gentle. Off the Maldiva and 
Chagos atoUs, the inclination is much more abrupt ; 
thus at Heawandoo Pholo, Lieut. Powell ^ found 50 and 
60 fathoms close to the edge of the reef, and at 300 
yards distance there was no bottom with a 300 yard 
Ime. Capt. Moresby informs me, that at 100 fathoms 
from the mouth of the lagoon of Diego Garcia he found 
no bottom with 150 fathoms : this is the more remark- 
able, as the slope is generally less abrupt in front of 
channels through a reef, owing to the accumulation 
of sediment. At Egmont Island, also, at 150 fathoms 
from the reef, soundings were struck with 150 fathoms. 
Lastly, at Cardoo atoll, only sixty yards from the reef, 
no bottom was obtained, as I am informed by Captain 
Moresby, with a line of two hundred fathoms ! The 
currents run with great force round these atolls, and 
where they are strongest, the inclination appears to be 
most abrupt. I am informed by the same authority, 
that wherever soundings were obtained off these is- 
lands, the bottom was invariably sandy : nor was there 
any reason to suspect the existence of submarine cliffs, 
as there was at Keeling Island.^ Here, then, occurs a 

' This fact is taken from a MS. account of these groups lent me 
by Capt. Moresby. See also Capt. Moresby's paper on the Maldiva 
atolls in the Geographical Journal, vol, v. p. 401. 

^ Off some of the atolls in the Low Archipelago the bottom 
appears to descend by ledges. Off Elizabeth Island, which consists 
of raised coral-rock, Capt. Beechey (p. 45, quarto ed.) describes three 
ledges ; the first slopes gently from the beach to a distance of about 
fifty yards ; the second extends two hundred yards with a depth of 

32 ATOLLS. Ch. L 

difficulty; — can sand accumulate on a slope, which, 
in some cases, appears to exceed fifty-five degrees ? It 
must be observed, that I speak of slopes where sound- 
ings were obtained, and not of such cases, as that of 
Cardoo, where the nature of the bottom is unknown, 
and where its inclination must be nearly vertical. M. 
Elio de Beaumont ' has argued, and there is no higher 
authority on this subject, from the inclination at which 
snow slides down in avalanches, that a bed of sand or 
mud cannot be formed at a greater angle than thirty 
degrees. Considering the number of soundings on 
sand, obtained round the Maldiva and Chagos atolls, 
which appear to indicate a greater angle, and the ex- 
treme abruptness of the sand-banks m the West Indies 
as will be mentioned in the Appendix, I must conclude 
that the adhesive property of wet sand counteracts its 
gravity, in a much greater ratio than has been allowed 
for by M. Elie de Beaumont. From the faciUty with 
which calcareous sand becomes agglutinated, it is not 
necessary to suppose that the bed of loose sand is thick. 
Capt. Beechey has observed, that the submarine 
slope is much less at the extremities of the more 
elongated atolls in the Low Archipelago, than at their 
sides; in speaking of Ducie's Island he says- the 
buttress, as it may be called, which ' has the most 

twenty-five fathoms, and then ends abruptly, like the first; an.l 
immeditttely beyond this there is no bottom with two hmidrcd 

' Memoires pour servir a une description G6olog. de France, toma 
iv. p. 216. 

•* Beechey's Voyage, 4to. ed. p. 44. 

Sdci. II. ATOLLS. 33 

powerful enemy (the S.W. swell) to oppose, is carried 
out much further, and with less abruptness, than the 
other.' In some cases, the less inclination of a certain 
part of the external slope, for instance of the northern 
extremities of the two Keeling atolls, is caused by a 
prevailing current which there accumulates a bed of 
sand. Where the water is perfectly tranquil, as within 
a lagoon, the reefs generally grow wp perpendicularly, 
and sometimes even overhang their bases : on the 
other hand, on the leeward side of Mauritius, where 
the water is generally tranquil although not invariably 
so, the reef is very gently inclined. Hence it appears 
that the exterior angle is much varied. We can, 
however, discern the effects of uniform laws in the 
close similarity in form between the sections of Keel- 
ing atoll and of the atolls in the Low Archipelago — 
in the general steepness of the reefs of the Makliva 
and Chagos atolls — and in the perpendicularity of 
those rising out of water always tranquil ; but from 
the complex action of the surf and currents on the 
growing powers of the coral and on the deposition of 
sediment, we can by no means follow out all the 

Where islets have been formed on the reef, that 
part which I have called the * flat,' and which is partly 
dry at low water, apjpears similar in every atoll. In 
the Marshall group in the N. Pacific, it may be 
inferred from Chamisso's description, that the reef, 
where islets have not been formed on it, slopes gently 
from the external marpjin to the shores of the lacjoou : 

34 ATOLLS. Cn. I. 

Flinders states tliat the Australian barrier lias a 
similar inclination inwards, and I have no doubt it is 
of general occurrence, although, according to Ehren- 
berg, the reefs of the Eed Sea offer an exception. 
Cliamisso observes that * the red colour of the reef (at 
the Marshall atolls) under the breakers is caused by a 
Nullipora, which covers the stone wherever the waves 
heat ; and, under favourable circumstances, assumes a 
stalactitical form,' — a description perfectly applicable 
to the margin of Keeling atoll. ^ Although Chamisso 
does not state that the masses of Nulliporse form 
l-)oints or a mound, higher than the flat, yet I believe 
that this is the case ; for Kotzebue,^ in another part, 
speaks of the rocks on the edge of the reef * as visible 
for about two feet at low-water,' and these rocks we 
may feel certain are not formed of true coral.' 

* Kotzebue's First Voyage, vol. iii. p. 142. Near Porto Praya, in 
the Cape de Verde Islands, some basaltic rocks, lashed by no incon- 
siderable surf, ■nere completely enveloped with a layer of Nullipora. 
The entire surface over many square inches, was coloured of a peach- 
blossom red ; the layer, however, was of no greater thickness than 
paper. Another kind, in the form of projecting knobs, grew in the 
same situation. These Nulliporse are closely related to those de- 
scribed on the coral-reefs, but I believe are of different species. 

* Kotzebue's First Voyage, vol. ii. p. 16. Lieut. Nelson, in his 
excellent memoir in the Geological Transactions (vol. ii. p. 105), 
alludes to the rocky points mentioned by Kotzebue, and infers thai 
they consist of Serpula), which compose incrusting masses on the 
reefs of Bermudas, as they likewise do on a sandstone-bar off the 
coast of Brazil, as described by me in the London Phil. Journal, 
Oct. 1841. I have added my description as a short supplement to 
the present volume. These masses of Serpulae hold the same posi- 
tion, relatively to the action of the sea, with the Nullipora3 on the 
coral-reefs in the Indian and Pacific oceans. 

* Capt. Moresby, in his valuable paper ' On the Northern AtoUa 

Sect. TI. ATOLLS. 35 

Whether a smooth convex mound of Nullipor^e, like 
that which appears as if artificially constructed to 
protect the margin of Keeling Island, is of frequent 
occurrence round atolls, I know not ; but we shall 
presently meet with it under precisely the same form, 
on the outer edge of the * barrier reefs ' which en- 
circle the Society Islands. 

There appears to be scarcely a feature in the 
structure of Keeling reef, which is not of common, if 
not of universal occurrence, in other atolls. Thus 
Chamisso describes' a layer of coarse conglomerate, 
outside the islets round the Marshall atolls, which 
* appears on its upper surface uneven and eaten 
away.' From drawings with appended remarks, of 
Diego Garcia in the Chagos group and of several of 
the Maldiva atolls, shown me by Captain Moresby,^ it 
is evident that then' outer coasts are subject to the 
same round of decay and renovation as those of 
Keeling atoll. From the description of the atolls in 
the Low Archipelago, given in Captain Beechcy's 
Voj^age, it is not apparent that any conglomerate 
coral-rock was there observed. 

The lagoon in Keeling atoll is shallow : in the 
atolls of the Low Archipelago the depth varies from 20 
to 38 fathoms, and in the Marshall Group, according 
to Chamisso, from 30 to 35 : in the Caroline atolls it 

of Maldivas' (Geographical Journal, vol. v.), says that the edges of 
the reefs there stand above water at low spring tides. 

' Kotzebue's First Voyage, vol. iii. p. 144. 

^ See also Moresby on the Northern Atolls of the Maldivas, Geo- 
graphical Journal, vol. v. p. 400. 

36 ATOLLS. CiT. L 

is only a little less. Within the Maldiva atolls there 
are large spaces with 45 fathoms, and some soundings 
are laid down at 49 fathoms. The greater part of the 
bottom in most lagoons, is formed of sediment ; large 
spaces have exactly the same depth, or the depth 
varies so insensibly, that it is evident that no other 
means excepting aqueous deposition, could have 
levelled the surface so equally. In the Maldiva atolls 
this is very consj)icuous, and likewise in some of the 
Caroline and Marshall Islands. In the former, large 
spaces consist of sand and soft clay ; and Kotzebue 
speaks of clay having been found within one of the 
Marshall atolls. No doubt this clay is calcareous mud, 
similar to that at Keeling Island, and to that at Ber- 
muda already referred to, as undistinguishable from 
disintegrated chalk, and which Lieut. Nelson says is 
called there pipe-clay.' 

Where the waves act with unequal force on the 
two sides of an atoll, the islets apj^ear to be first 
formed, and are generally of greater length on the more 
exposed shore. The islets, also, which are placed 

' I may here observe that on the coast of Brazil, where there is 
nrnch coral, the soundings near the land ai-e described by Admiral 
lioussin, in the Pilote du Brdsil, as siliceous sand, mingled with 
much finely comminuted particles of sliells and coral. Further in 
the offnig, for a space of 1,300 miles along the coast, from the 
Abrolhos islands to Maranham, the bottom in many places is com- 
posed of ' tuf blanc, mel6 ou form6 de madrepores broy^s.' This 
white substance, probably is analogous to that which occurs within 
the above-mentioned lagoons ; it is sometimes, according to Roussin, 
firm, and he compares it to mortar. [Probably the clay is commonly 
similar to that mentioned by Mr. Guppy (Proc. E. S. Edin. vol. xiii. 
p. 879 n.) and others. See the abstract of his paper in Appendix IL] 

Shot. H. ATOLLS. 37 

to leeward as regards the trade-wind, are in most 
l^arts of the Pacific hable to be occasionally swept 
entirely away by gales, equalling hurricanes in vio- 
lence, which blow in the opposite direction. The 
absence of islets on the leeward side of atolls, or, 
when present, their lesser dimensions compared with 
those to windward, is a comparatively unimportant 
fact ; but it is remarkable that in several instances 
the reef itself, although retaining its usual defined 
outline, does not rise to the surface by several fathoms 
on the leeward side. This is the case with the 
southern side of Peros Banhos (Plate I. fig. 9) in the 
Chagos group, with Mourileu atoll ^ in the Caroline 
Archipelago, and with the barrier reef (Plate I. fig. 8) 
of the Gambler Islands, where Captain Beechey was 
first led to observe the peculiarity in question. At 
Peros Banhos the submerged part is nine miles in 
length, and lies at an average depth of about 
five fathoms ; its surface is nearly level, and consists 
of hard stone with a thin covering of loose sand. 
There is scarcely any living coral on it, even on the 
outer margin, as I have been particularly assured by 
Captain Moresby : it is, in fact, a wall of dead coral-rock, 
having the same width and transverse section with the 
reef in its ordinary state, of which it is a continuous 
portion. The living and perfect parts termmate 
abruptly, and abut on the submerged portions, in 

' Frederic Lutk6's Voyage autour du Monde, vol. ii. p. 291. See 
also his account of Namonouito, at pp. 97 and 105, and the chart of 
OuUeay in the Atlas. 

38 ATOLLS. Ch, 1. 

the same manner as occurs where there is a passage 
through the reef. The reef to leeward in other cases is 
nearly or quite ohhterated, and one side of the lagoon 
is left open ; for instance, at Oulleay (Caroline Archi- 
pelago), where a crescent-formed reef is fronted by an 
ii-regular bank, on which the other half of the annular 
reef probably once stood. At Namonouito in the same 
Archipelago, both these modifications of the reef concur ; 
it consists of a great flat bank, with from 20 to 25 
fathoms of water on it ; for a length of more than 40 
miles on its southern side it is open and without any 
reef, whilst on the other sides it is bounded by a reef, in 
parts rising to the surface and perfectly characterised, 
in parts lying some fathoms submerged. In the Chagos 
group there are annular reefs entirely submerged, which 
have the same structure as the submerged and defined 
portions just described. The Speaker's Bank offers an 
excellent example of this structure ; its central expanse, 
which is about 22 fathoms deep, is 24 miles across ; the 
external rim is of the usual width of annular reefs, and 
is well-defined ; it lies between six and eight fathoms 
beneath the sm*face, and at the same depth there are 
scattered knolls ui the lagoon. Captain Moresby believes 
that the rim consists of dead rock thinly covered with 
sand, and he is certain that this is the case with the 
external rim of the Great Chagos Bank, which is also 
essentially a submerged atoll. In both these cases, as in 
the submerged portion of the reefatPerosBanhos,Capt. 
Moresby feels sure that the quantity of living coral, 
even on the outer edge overhanging the deep-sea water, 

Sect. II. ATOLLS. S9 

is quite insignificant. Lastly, in several parts of the 
Pacific and Indian Oceans there are banks, lying at 
greater depths than in the cases just mentioned, of the 
same form and size v/ith the neighbouring atolls, but 
"with their atoll-like structure wholly obliterated. It 
appears from the survey of Freycinet, that there are 
banks of this kind in the Caroline Archipelago, and, as 
is reported, in the Low Archipelago. When we discuss 
the origin of the different classes of coral formations, 
we shall see that the submerged state of the whole of 
some atoll-formed reefs, and of portions of others 
generally but not invariably on the leeward side, and 
the existence of more deeply submerged banks now 
possessing little or no signs of their original atoll-like 
structure, are probably the effects of a uniform cause, — 
namely, the death of the coral, during the subsidence of 
the area, in which the atolls or banks are situated. 

There are seldom (with the exception of the Maldiva 
atolls), more than two or three channels, and generally 
only one leading into the lagoon, of sufficient depth for 
u ship to enter. In small atolls, there is usually not 
(jven one. Where there is deep water, for instance 
above 20 fathoms, in the middle of the lagoon, the 
channels through the reef are seldom as deep as the 
centre, — it may be said that the rim only of the saucer- 
shaped hollow forming the lagoon is notched. Sir C. 
Lyell' has observed that the growth of the coral would 
tend to obstruct all the channels through a reef, except 
those kept open by discharging the water, which during 
' Principles of Geology, vol. iii. p. 289. [Vol. ii. p. G09, ed. 1872.] 

40 ATOLLS Ch. I. 

high tide and the greater part of each ebb is thrown 
over a large portion of its circumference. Several facts 
indicate that a considerable quantity of sediment is 
likewise discharged through these channels ; and 
Captain ]\Ioresby has observed, during the change of 
the monsoon, that the sea is discoloured to some dis- 
tance off the entrances into the Maldiva and Chagos 
atolls. This would probably check the growth of the 
coral in the channels, far more effectually than if they 
merely discharged a current of water. "Where there 
is not any channel, as in the case of many small atolls, 
these causes have not prevented the entu-e ring attain- 
ing the surface. The channels, like" the submerged and 
effaced parts of the reef, occur very generally, though 
not invariably on the leeward side of the atoll, or on 
that side, according to Beechey,' which, from extending 
m the same direction with the prevalent wind, is not 
fully exposed to it. Passages between the islets on 
the reef through which boats can pass at high- water, 
must not be confounded with ship-channels by which 
the annular reef itself is breached. The passages 
between the islets occur, of course, on the windward 
as well as on the leeward side ; but they are more 
frequent and broader to leeward, owmg to the lesser 
dimensions of the islets on that side. 

At Keeling atoll the shores of the lagoon shelve 
gradually where the bottom is of sediment, and irregu- 
larly or abruptly where there are coral reefs ; but this 
is by no means the universal structure in other atolls. 

' Cccchcy's Vovage, 4to ed. vol. i. p. 189. 

Sect. II. ATOLLS. 41 

Chamisso,' speaking in general terms of the lagoons in 
the Marshall atolls, says the lead generally sinks 'from 
a depth of two or three fathoms to twenty or twenty- 
four, and you may pursue a line in which on one side 
of the boat you may see the bottom, and on the other 
the azure-blue deep water.' The shores of the lagoon- 
like channel within the barrier-reef at Vanikoro have 
a similar structure. Captain Beechey has described a 
modification of this structure (and he believes it is not 
uncommon) in two atolls in the Low Archipelago, in 
which the shores of the lagoon descend by a few broad, 
slightly inclined ledges or steps : thus at Matilda 
atoll,^ the great exterior reef, the surface of which is 
gently inclined inwards, ends abruptly in a little 
submarine cliff three fathoms deep ; at its foot, a 
ledge 40 j^ards in width also shelves gently in- 
wards, like the surface-reef, and terminates in a 
second little cliff five fathoms deep ; beyond this, the 
bottom of the lagoon slopes to 20 fathoms, which 
is the average depth of its centre. These ledges 
seem to be formed of coral rock ; and Captain Beechey 
says that the lead often descended several fathoms 
through holes in them. In some atolls, all the coral 
reefs or knolls in the lagoon come to the surface at 
low-water ; in other cases of rarer occurrence, all 
lie at nearly the same depth beneath it, but most 

' Kotzebue's First Voyage, vol. iii. p. 142, 

* Beechey's Voyage, 4to ed. vol. i. p. IGO. At Whitsunday Islaiid 
the bottom of the lagoon slopes gradually towards the centre, and 
then deepens suddenly, the edge of the bank being nearly parpen- 
dicular. This bank is formed of coral and dead shells. 

42 ATOLLS. Cn. I. 

frequently tliey are quite irregular — some with per- 
pendicular, some with sloping sides— some rising 
to the surface, and others lying at all intermediate 
depths from the hottom upwards. I cannot, tliere- 
fore, suppose that the union of such reefs could pro- 
duce even one uniformly sloping ledge, and much 
less two or three one heneath the other, and each ter- 
minated by an abrupt wall. At Matilda Island, which 
offers the best example of the step-like structurBj 
Captain Beechey observes that the coral knolls withiii 
the lagoon are quite irregular in their height. We 
shall hereafter see that the theory which accounts for 
the ordinary form of atolls, apparently includes this 
occasional peculiarity m their structure. 

In the midst of a group of atolls, there sometimes 
occur small, flat, very low islands of coral formation, 
which probably once included a lagoon, since filled 
up with sediment and coral-reefs. Captain Beeyhey 
entertains no doubt that this has been the case with 
the two small islands, which alone of thirty-onf< sur- 
veyed by him in the Low Archipelago, did not con- 
tain lagoons. Eomanzoff Island (in lat. 15° S.) is 
described by Chamisso' as formed b}' a dam of madre- 
poritic rock inclosing a flat space, thinly covered with 
trees, into which the sea on the leeward side occasion- 
ally brealvs. North Keeling atoll appears to be m a 
rather less forward stage of conversion mto land : it 
consists of a horse-shoe shaped strip of land surround- 
ing a muddy flat, one mile in its longest axis, which ia 

■ Kotzebue'B First Voyage, vol. iii. p. 221. 

Sect. II. ATOLLS. 43 

covered by tlie sea only at high-water. When de- 
scribmg South Keehng atoll, I endeavoured to show 
how slow the final process of fillmg up a lagoon must 
be ; nevertheless, as all causes do tend to produce this 
effect, it is very remarkable that not one insta,nce, as I 
believe, is known of a moderately sized lagoon being 
filled up even to the low-water line at sprmg-tides, 
much less of such a one being converted mto land. It 
is, likewise, in some degree remarkable, how few atolls, 
except small ones, are surrounded by a single linear 
strip of land formed by the union of separate islets. 
We cannot suppose that the many atolls in the Pacific 
and Indian oceans all have had a late origin, and yet 
should they remain at their present level, subjected 
only to the action of the sea and to the growing powers 
of the coral, during as many centuries as must have 
elapsed since any of the earlier tertiary epochs, it can- 
not, I think, be doubted that their lagoons and the 
islets on their reef, would present a totally different 
appearance from what they now do. This considera- 
tion leads to the suspicion that some agency (namely, 
subsidence) comes into play at intervals, and renovates 
their original structure. 

44 ATOLLS. Ch. I. 


Maldiva Archipelago — Pdng-foTincd reefs, marginal and central — 
Li real depth in the lagoons of the southern atolls — lieefs in the 
lagoons all reach the surface — Position of islets, and breaches in 
the reefs with respect to the prevalent winds and action of the 
waves — Destriiction of islets — Relation in position betiveen dis- 
trict atolls — The apparent dissevervient of large atolls — The 
Great Chagos Ba)ik — Its submerged condition and extraordinai-y 

Although occasional references have been made to tlie 
Maldiva atolls and to the banks in the Chagos group, 
Bome points of their structure deserve further consi- 
deration. My description is derived from an exami- 
nation of the admirable charts lately published from 
the survey of Captain Moresby and Lieut. Powell, 
and more especially from information which Captain 
Moresby has communicated to me in the kmdest 

The Maldiva Archipelago is 470 miles in length, 
with an average breadth of about 50 miles. The form 
and dimensions of the atolls, and their singular posi- 
tion in a double line, may be seen, though imperfectly, 
m the greatly reduced chart (fig. 6) in Plate II. The 
dimensions of the longest atoll in the group (called l)y 
the double name of Milla-dou-Madou and Tilla-dou- 
Matte) have already been given ; it is 88 miles in a 
medial and slightly curved line, and is less than 20 
miles in its broadest part. Suadiva, also, is a noble 
atoll, being 44 miles across in one direction, and 34 in 


anotlier, and the great included expanse of water has a 
depth of between 250 and 300 feet. The smaller atolls 
in this group differ in no respect from ordinary ones; 
h lit the larger ones are remarkable from being breached 
by numerous deep-water channels leading into the 
lagoon ; for instance, there are 42 channels through 
which a ship could enter the lagoon of Suadiva. In 
the three southern large atolls, the separate portions of 
reef between these channels have the ordinary structure 
and are linear ; but m the other atolls, especially the 
northern ones, these portions are ring-formed like 
miniature atolls. Other ring- formed reefs rise out of 
the lagoons, in the place of those irregular ones which 
ordinarily occur there. In the reduction of the chart 
of Mahlos Mahdoo (Plate II. fig. 4), it was not found 
easy to define the islets and the little lagoons within 
each reef, so that the ring- formed structure is very im- 
perfectly shown : in the large published charts of Tilla- 
dou-Matte, the appearance of these rings, from stand- 
ing further apart from each other, is very remarkable. 
The rings on the margin are generally elongated ; 
many of them are three, and some even five miles in 
diameter ; those within the lagoon are usupJly smaller, 
few being more than two miles across, and the greater 
number rather less than one. The depth of the little 
lagoon within these small annular reefs is generally frum 
five to seven fathoms, but occasionally more ; and in Ai'i 
atoll many of the central ones are twelve, and some 
even more than twelve fathoms deep. These rings rise 
abruptly from the platform or bank on which they 

4G ATOLLS. Cn. I. 

stand ; their outer margins are invariably bordered by 
living coral/ within which there is a flat surface of 
coral rock ; on this flat, sand and fragments have in 
many cases accumulated and been converted into islets 
clothed with vegetation. They are indeed larger, and 
contain deeper lagoons than many true atolls standing 
in the open sea ; and I can point out no essential 
difference between these little ring-formed reefs and 
the most perfectly characterised atolls, excepting that 
they are based on a shallow foundation, instead of on 
the floor of the ocean, and that instead of being 
scattered irregularly, they are grouped closely together 
with the marginal rings arranged in a rudely-formed 

The perfect series which can be traced from a linear 
reef like that surrounding an ordinary atoll, to others 
which are ring-formed and much elongated but con- 
taining only a very narrow lagoon, and to others which 
are oval or almost circular, renders it probable that the 
latter are merely modifications of a linear and normal 
reef. The fact that the marginal annular reefs 
generally have their longest axes directed in the line 
which the exterior Imear reef would have held, agrees 
with this view. We may also infer that the central 
annular reefs are modifications of those irregular ones, 
which are found in the lagoons of all common atolls. 
It appears from the charts on a large scale, that the 

' Captain Moresby informs me that Millcpora complanata is one 
of the commonest kinds on the outer margin, as it is at Keeling 

Sect. m. MALDIVA ATOLLS. 47 

ring-like structure in these central reefs is con- 
tingent on the marginal channels or breaches being 
wide ; and, consequently^ on the whole interior of 
the atoll being freely exposed to the waters of the 
open sea. When the channels are narrow or few in 
number, although the lagoon be of great size and 
depth (as in Suadiva), there are no ring-formed reefs ; 
where the channels are somewhat broader, the mar- 
ginal portions of reef, and especially those close to the 
larger channels, are ring-formed, but the central ones 
are not so ; where they are broadest, almost every 
reef throughout the atoll is more or less perfectly ring- 
formed. Although their presence is thus contingent 
on the openness of the marginal channels, the theory of 
their formation, as we shall hereafter see, is included 
in that of the parent atolls of which they form the 
separate portions. 

The lagoons of all the atolls in the southern part of 
the Archipelago are from 10 to 20 fathoms deeper than 
those in the northern part. This is well exemplified in 
the case of Addoo, the southernmost atoll in the group, 
for although only 9 miles in its longest diameter, it has 
a depth of 39 fathoms, whereas all the other small atolls 
have comparatively shallow lagoons ; I can assign no 
adequate cause for this difference in depth, excepting 
that the southern part of the Archipelago has subsided 
to a greater degree or at a quicker rate than the 
northern part ; and this conclusion agrees well with the 
fact that, in the Chagos group, lying 280 miles still 
further southwards, most of the atolls are sunken and 

48 ATOLLS. Ch I. 

half destroyed with the dead corals. In the central and 
deepest part of the Maldiva lagoons, the bottom consists, 
as I am informed by Captain Moresby, of stiff clay 
(probably a calcareous mud) ; nearer the border it con- 
sists of sand, and in the channels through the reef, of 
hard sand-banks, sandstone, conglomerate rubble, and a 
little live coral. Close outside the reef the bottom is 
sandy, and slopes abruptly into unfathomable depths. 
In most lagoons the dex^th is considerably greater 
in the centre than in the channels ; but in Tilla- 
dou-Matte, where the marginal ring-formed reefs 
stand far apart, the same depth is carried across the 
entire atoll, from the deep-water line on one side to 
that on the other. I cannot refrain from once again 
remarking on the singular structure of these atolls, 
— a great sandy and generally concave disk rises 
abruptly from the unfathomable ocean, with the central 
expanse studded and the margins symmetrically fringed 
with oval basins of coral-rock, just lipping the surface 
of the sea, sometimes clothed with vegetation, and each 
containing a little lake of clear salt water. 

In the southern Maldiva atolls, of which there are 
nine large ones, all the small reefs within the lagoons 
come to the surface, and are dry at low-water spring- 
tides ; hence in navigating them there is no danger 
from submarine banks. This circumstance is very 
remarkable, as within some atolls, for instance those of 
the neighbouring Chagos group, not a single reef comes 
to the surface, and in most other cases a few only do, 
and the rest lie at all intermediate depths from thy 


bottom upwards. When treating of the growth of coral 
I shall again refer to this subject. 

Although in the neighbourhood of the Maldiva 
Archipelago the winds, during the monsoons, blow 
during nearly an equal time from opposite quarters, 
and although, as I am informed by Captain Moresby, 
the westerly winds are the strongest, yet the islets are 
almost all placed on the eastern side of the northern 
atolls, and on the south-eastern side of the southern 
atolls. That the formation of the islets is due to 
detritus thrown up from the outside, as in the ordinary 
manner, and not from the interior of the lagoons, may, 
I think, be safely inferred from several considerations 
which it is hardly worth while to detail. As the easterly 
winds are not the strongest, their action probably is 
aided by some prevailing swell or current. 

In groups of atolls exposed to the trade wind, the 
ship-channels into the lagoons are almost always 
situated on the leeward or less exposed side of the reef, 
and the reef itself is sometimes either wanting there, or 
is submerged. A strictly analogous, but diiferent, fact 
may be observed at the Maldiva atolls — namely, that 
where two atolls stand near together, the breaches in 
the reef are most numerous on the sides which face each 
other, and are therefore less exposed to the waves. Thus 
on the sides of Ari and the two Nillandoo atolls which 
face S. Male, Phaleedoo, and Moloque atolls, there are 
seventy-three deep-water channels, and only twenty-five 
on the outer sides ; on the three latter-named atolls there 
are fifty-six openings on the near side, and only thirty- 

50 ATOLLS. Cn. L 

Beven on the outside. It is scarcely possible to attri- 
bute this difference to any other cause than the some- 
what different action of the sea on the two sides, which 
would ensue from the mutual protection afforded by 
the two rows of atolls. I may here remark that in 
most cases, the conditions favourable to the greater 
accumulation of fragments on the reef and to its 
more perfect continuity on one side of the atoll than 
on the other, have concurred, but this has not been 
the case with the Maldivas ; for we have seen that the 
islets are placed on the eastern or south-eastern sides, 
whilst the breaches in the reef occur indifferently on 
any side where protected by an opposite atoll. The reef 
being more continuous on the outer and more exposed 
sides of those atolls which stand near each other, 
accords with the fact, that the reefs of the southern 
atolls are more continuous than those of the northern 
ones, for the former, as I am informed by Captain 
Moresby, are more constantly exposed to a heavy surf 
than are the northern atolls. 

The date of the first formation of some of the islets 
in this Archipelago is known to the inhabitants ; on the 
other hand, several islets, and even some of those which 
are believed to be very old, are now fast wearing away. 
The work of destruction has, in some instances, been 
completed in ten years. Captain Moresby found on one 
water- washed reef the marks of wells and graves, which 
were excavated when it supported an islet. In South 
Nillandoo atoll, the natives say that three of the islets 
were formerly larger : in North Nillandoo there is one 


now being washed away ; and in this latter atoll Lieut. 
Prentice found a reef, about six hundred yards in 
diameter, which the natives positively affirmed was 
lately an island covered with cocoa-nut trees. It is now 
only partially dry at low-water spring tides, and is (in 
Lieut. Prentice's words) ' entirely covered with live 
coral and madrepore.' In the northern part, also, of 
the Maldiva Archipelago and in the Chagos group, it 
is known that some of the islets are disappearing. 
The natives attribute these effects to variations in the 
currents of the sea. For my own part I cannot avoid 
suspecting, that there must be some further cause, 
which gives rise to such a cycle of change in the action 
of the currents of the great and open ocean. 

Several of the atolls in this Archipelago are so 
related to each other in form and position, that at the 
first glance one is led to suspect that they have 
originated in the disseverment of a single one. Male 
consists of three perfectly characterised atolls, of which 
the shape and relative position are such, that a line 
drawn closely round all three gives a symmetrical 
figure ; but to see this, a larger chart is required than 
that of the Archipelago in Plate II. The channel 
separating the two northern Male atolls is only little 
more than a mile wide, and no bottom was found in it 
with 100 fathoms. Powell's Island is situated at the 
distance of two miles and a-half off the northern end 
of another atoll, namely Mahlos Mahdoo (fig. 4), at 
the exact point where the two sides of the latter, 
if prolonged, would meet : no bottom, however, 

52 ATOLLS. Cu. I. 

was found in the channel with 200 fathoms : in the 
wider channel between Horsburgh atoll and the south- 
ern end of Mahlos Mahdoo, no bottom was found with 
250 fathoms. In these cases, the relation consists 
only in the form and position of the atolls. But 
in the channel between the two Nillandoo atolls, 
although three miles and a-quarter wide, soundmgs 
were struck at the depth of 200 fathoms : the channel 
between Eoss and Ari atoUs is four miles wide, and only 
150 fathoms deep. Here then we have a submarine 
connection, besides a relation in position and form. The 
fact of soundings having been obtained between two 
separate and perfectly characterised atolls is in itself 
interesting, as it has never, I believe, been effected in 
any of the many other groups of atolls in the Pacific 
and Indian seas. In continuing to trace the con- 
nection of adjoining atolls, if a hasty glance be taken 
at the chart (fig. 4, Plate II.) of Mahlos Mahdoo and 
the line of unfathomable water be followed, no one 
will hesitate to consider it as one atoll. But a second 
look will show that it is divided by a bifurcating 
channel, of which the northern arm is about one mile 
and three-quarters in width, with an average depth 
of 125 fathoms, and the southern one three-quarters 
of a mile wide, and rather less deep. These channels 
resemble in the slope of their sides and general form, 
those which separate atolls in every respect distinct ; 
and the northern arm is wider than that dividing two 
of the Mule atolls. The ring-formed reefs on the 
northern and southern sides of this bifurcatmg channel 

Sect. in. GREAT CHAGOS BANK. 53 

are elongated, and so continuous that the northern and 
southern portions of Mahlos Mahdoo may claim to 
be considered as distinct atolls. But the reefs of 
the intermediate portion are less perfect, so that this 
portion hardly yet resembles a distinct atoll. Mahlos 
Mahdoo, therefore, is in every respect in an inter- 
mediate condition, so that it may be considered either 
as a single atoll nearly dissevered into three portions, 
or as three atolls almost perfect and intimately con- 
nected. This is an instance of a very early stage of 
the apparent disseverment of an atoll, and another 
is exhibited at Tilla-dou-Matte. In one part of 
this atoll, the ring-formed reefs stand so far apart 
from each other, that the inhabitants have given 
different names to the northern and southern halves : 
nearly all the rings, moreover, are so perfect, and 
stand so separate, and the space from which they rise 
is so level and unlike a true lagoon, that we can easily 
imagine the conversion of this one great atoU, not into 
two or three portions, but into a whole group of 
miniature atolls. A series such as we have here 
traced, impresses the mind with the idea of actual 
change ; and it will hereafter be seen, that the theory 
of subsidence together with the upward growth of the 
coral-reefs, modified by accidents of probable occur- 
rence, accounts for the occasional disseverment of large 

The great Chagos Bank alone remains to be de- 
scribed.' In the Chagos grou^j there are some ordi- 

' [See Appendix 11.] 

54 ATOLLS. Crr. L 

nary atolls, sonie annular reefs rising to the surface 
but without any islets on them, and some atoll- formed 
banks either quite or nearly submerged. Of the 
latter, the Great Chagos Bank is much the largest, and 
differe in its structure from the others ; a plan of it 
is given in Plate II. fig. 1, in which, for the sake of 
clearness, I have had the parts under ten fathoms deep 
finely shaded : an east and west vertical section is given 
in fig. 2, in which the vertical scale has been neces- 
sarily exaggerated. Its longest axis is ninety nautical 
miles, and another line drawn across the broadest part, 
at right angles to the first, is seventy. The central 
part consists of a level muddy flat between forty and 
fifty fathoms deep, which is surrounded on all sides, 
with the exception of some breaches, by the steep 
edges of a set of banks rudely arranged in a circle. 
These banks consist of sand with a very little live 
coral ; they vary in breadth from five to twelve miles, 
and on an average lie about sixteen fathoms beneath 
the surface ; they are bordered by the steep edges of a 
third narrow and upper bank, w'hich forms the rim to 
the whole. This rim is about a mile in width, and, with 
the exception of two or three spots where islets have 
beenformed, is submerged between fiveand ten fathoms. 
It consists of smooth hard rock, covered with a thin 
layer of sand, but with scarcely any live coral ; it is 
steep on both sides, and slopes abruptly outwards into 
unfathomable depths. At the distance of less than half 
a mile from one part, no bottom was found with 190 
fathoms ; and off another point, at a somewhat greater 


distance, there was none with 210 fathoms. Small 
steep-sided banks or knolls, covered with luxuriantly- 
growing coral, rise from the interior expanse to tlio 
same level with the external rim, which, as we have seen, 
is formed only of dead rock. It is impossible to look at 
the plan (fig. 1, Plate II.), although reduced to so 
small a scale, without at once perceiving that the Great 
Chagos Bank is, in the words of Captain Moresby,^ 'no- 
thing more than a half-drowned atoll.' But of what 
great dimensions, and of how extraordinary an internal 
structure ! We shall hereafter have to consider both 
the cause of its submerged condition, a state common 
to other banks in the group, and the origin of the 
singular submarine terraces which bound the central 
expanse ; these, I think it can be shown, have resulted 
from a cause analogous to that which has produced the 
bifurcating channel across Mahlos Mahdoo. 

' This officer has had the kindness to lend me an excellent MS. 
account of the Chagos Islands ; from this paper, from the published 
charts, and from verbal information communicated to me by Captaiu 
Moresby, the above account of the Great Chagos Bank is taken. • 



Closely rcacvihle hi general form and striichire atoll-reefs — Width 
and deptJi of the lagoon-channels — Breaches through the reef in 
front of valleys, and generally on the leeward side — Checks to the 
filling up of the lagoon-channels — Size and constitxdion of the 
encircled islands — Number of islands within the same reef — 
Barrier-reefs of Neiu Caledonia and Australia — Position of the 
reef relative to the slope of the adjoining land — Probable great 
thickness of barrier-reefs. 

The term * barrier ' has been generally applied to that 
vast reef which fronts the N.E. shore of Australia, and 
by most voyagers likewise to that on the western coast 
of New Caledonia. At one time I thought it con- 
venient thus to restrict the term, but as these reefs are 
similar in structure and in position relatively to the 
land, to those, which, like a wall with a deep moat 
within, encircle many smaller islands, I have classed 
them together. The reef, also, on the west coast of 
New Caledonia, cu-cling round the extremities of tho 
island, is an intermediate form between a small en- 
circling reef and the Austrahan barrier, which stretches 
for a thousand miles in nearly a straight Hne. 

The geographer Balbi has m effect described those 
barrier-reefs which encircle moderately sized islands, 
by calling them atolls with high land rising from 
within their central expanse. The general resem- 


blance between the reefs of the barrier and atoll 
classes may be seen m the small, but accurately re- 
duced charts on Plate I.,' and this resemblance can be 
further shown to extend to every part of their struc- 
ture. Beginning with the outside of the reef ; many 
scattered soundings off Gambler, Ualan, and some 
other encircled islands, show that close to the breakers 
there exists a narrow shelving margin, beyond which 
in most cases, the ocean suddenly becomes unfathom- 
able. Off the west coast of New Caledonia, Captain 
Kent ^ found no bottom with 150 fathoms, at two ship's 
lengths from the reef ; so that the slope here must be 
nearly as precipitous as off the Maldiva atolls. 

I can give little information regarding the kinds 
of corals which live on the outer margin. When I 
visited the reef at Tahiti, although it was low-water, 
the surf was too violent for me to see the living 
masses ; but, according to what I heard from some in- 
telligent native chiefs, they resemble in their rounded 
and branchless forms, those on the margin of Keeling 
atoll. The extreme verge of the reef which was 
visible between the breaking waves at low-water, con- 
sisted of a rounded, convex, artificial-like breakwater, 
entirely coated with Nulliporse, and absolutely similar 
to that which I have described at Keeling atoll. 
From what I heard when at Tahiti, and from the 

' The authorities from which these charts have been reduced, 
together with some remarks on them, are given in a separately ap- 
pended page, descriptive of the Plates. 

- Dalrymple, Hydrog. Mem. vol. iii. 


writings of the Eevs. W. Ellis and J. Williams, I con- 
clude that this peculiar structure is common to most 
of the encircled islands of the Society Archipelago. 
The reef within this mound or breakwater, has an ex- 
tremely irregular surface, even more so than between 
the islets on the reef of Keeling atoll, with which 
alone (as there are no islets on the reef of Tahiti) it 
can properly be compared. At Tahiti the reef is very 
irregular in width ; but round many other encircled 
islands, for instance Vanikoro or Gambler Islands 
(figs. 1 and 8, Plate I.), it is quite as regular, and of 
the same average width, as in true atolls. Most 
barrier-reefs on the inner side slope irregularly into 
the lagoon-channel, (as the space of deep water sepa- 
rating the reef from the included land may be called,) 
but at Va,nikoro the reef slopes only for a short dis- 
tance, and then terminates abruptly in a submarine 
wall forty feet high, — a structure absolutely similar to 
that described by Chamisso in the Marshall atolls. 

In the Society Archipelago, Ellis' states that the 
reefs generally lie at the distance of from one to one 
and a-half miles, and, occasionally, even at more than 
three miles from the shore. The central mountains 
are generally bordered by a fringe of flat, and often 
marshy alluvial land, from one to four miles in width. 
This frmge consists of coral-sand and detritus thrown 
up from the lagoon-channel, and of soil washed down 
from the hills ; it is an encroachment on the channel, 

' Consult, on this and other points, the Polynesian Researches 
by the Rev. W. Ellis, an admirable work, full of curious information. 


analogous to that low and inner part of the islets in 
many atolls, which is formed by the accumulation of 
matter from the lagoon. At Hogoleu (fig. 2, Plate I.), 
in the Caroline Archipelago,^ the reef on the south 
side is no less than twenty miles ; on the east side, 
live ; and on the north side, fourteen miles from the 
encircled islands. 

The lagoon-channels may be compared in every 
respect with true lagoons. In some cases they are 
open, with a level bottom of fine sand ; in others they 
are choked up with reefs of delicately branched corals, 
which have the same general character as those within 
Keeling atoll. These internal reefs either stand sepa- 
rately, or more commonly skirt the shores of the in- 
cluded high islands. The depth of the lagoon-channel 
round the Society Islands varies from two or three, 
to thirty fathoms ; in Cook's ^ chart of Ulietea, how- 
ever, there is one sounding laid down of 48 fathoms : 
at Vanikoro there are several of 54 and one of 56i- 
fathoms (English), a depth which even exceeds by a 
little that of the interior of the great Maldiva atolls. 
Some barrier-reefs have very few islets on them ; whilst 
others are surmounted by numerous ones ; and those 
round part of Bolabola (Plate I., fig. 5), form a single 
linear strip. The islets first appear either on the 
angles of the reef, or on the sides of the breaches 

• See Hydrographical Mem. and the Atlas of the Voyage of the 
Astrolabe, by Capt. Dumont D'Urville, p. 428. 

* See the chart in vol. i. of Hawkesworth's 4to ed. of Cook's First 


through it, and are generally most numerous on the 
windward side. The reef to leeward retaining its usual 
width, sometimes lies submerged several fathoms be- 
neath the surface ; I have already mentioned Gambler 
Island as an instance of this structure. Submerged 
reefs, dead, covered with sand, and with a less defined 
outline, have been observed (see Appendix I.) off some 
parts of Huaheine and Tahiti. The reef is more fre- 
quently breached to leeward than to windward, although 
this is not so frequent as in the case of atolls. Thus I 
find in Krusenstcrn's Memoir on the Pacific, that there 
are passages through the encircling reef on the lee- 
ward side of the seven Society Islands, which possess 
ship-harbours ; but that there are openings to wind- 
ward through only three of them. The breaches 
in the reef are seldom as deep as the interior 
lagoon-like channel ; they generally occur in front of 
the main valleys, a circumstance which can be ac- 
counted for, as will be seen in the fourth chapter, 
without much difficulty. The breaches being generally 
situated in front of the valleys which descend on 
all sides, explains their more frequent occurrence 
through the windward side of barrier-reefs than 
through the windward side of atolls, — for in atolls 
there is no included land to influence the position of 
the breaches. 

It is remarkable that the lagoon-channels round 
mountainous islands have not in every instance been 
long ago filled up with coral and sediment ; but it is 
accounted for without much difficulty. In cases like 


that of Hogoleu and the Gamhier Islands, where a few 
small peaks rise out of a great lagoon, the conditions 
scarcely differ from those of an atoll ; and I have already 
shown at some length, that the filling up of a true 
lagoon must be an extremely slow process. Where 
the lagoon-channel is narrow, that agency, which on 
unprotected coasts is the most productive of sediment, 
namely the force of the breakers, is here entkely ex- 
cluded ; and owing to the reef being breached in the 
front of the main valleys, much of the finer mud from 
the rivers must be transported into the open sea. The 
water which is thrown over the edges of atoll-formed 
reefs causes a current which carries sediment from the 
lagoon through the breaches into the sea; and the 
same thing probably takes place in barrier-reefs. This 
would greatly aid in preventing the lagoon-channels 
from being filled up. The low alluvial border, how- 
ever, at the foot of the encircled mountains, shows 
that the work of filling up is in progress ; and at 
Maurua (Plate I., fig. 6), in the Society group, it has 
been almost effected, so that there remains only one 
harbour for small craft. 

If we look at a set of charts of barrier-reefs, and 
leave out in imagination the encircled land, we shall 
see that besides the many points already noticed of 
resemblance or rather of identity in structure with 
atolls, there is a close general agreement in form, aver- 
age dimensions, and grouping. Encircling reefs, like 
atolls, are generally elongated, and have an irregularly 
rounded, though sometimes angular outline. There are 


atolls of all sizes, from less than two miles in diameter 
to sixty miles (excluding Tilla-dou-Matte, which consists 
of a number of almost independent atoll-formed reefs) ; 
and there are encu-cling barrier-reefs from three miles 
and a-half to forty-six miles in diameter, — Turtle 
Island being an instance of the former, and Hogoleu of 
the latter. At Tahiti the encircled island is thirty-six 
miles in its longest axis, whilst at Maurua it is only a 
little more than two miles. It will also be shown in 
the last chapter, that there is the strictest resemblance 
between the grouping of atolls and of common islands, 
and there is the same resemblance between atolls and 
encircling barrier-reefs. 

The islands lying within reefs of this class, are of 
very various heights. Tahiti • is 7,000 feet ; Maurua 
about 800 ; Aitutaki 360, and Manouai only 50. The 
geological nature of the included land also varies ; in 
most cases it is of ancient volcanic origin, owing appa- 
rently to the fact that islands of this nature are the most 
frequent within all great seas ; some, however, are of 
madreporitic limestone, and others of primary forma- 
tion, of which latter kind New Caledonia offers the best 
example. The central land consists either of one island, 
or of several; thus in the Society group, Eimeo stands by 
itself; while Taha and Kaiatea (fig. 3, Plate I.), both 

' The height of Tahiti is given from Captain Beechey ; Maurua 
from Mr. F. D. Bennett (Geograph. Journ. vol. viii. p. 220) ; Aitutaki 
from measurements made on board the Beagle ; and Manouai, or 
Harvey Island, from an estimate by the Rev. J. Williams. The two 
latter islands, however, are not in some respects well characterised 
examples of the encircled class. 

Cii. n. BARRIER-EEEFS. 63 

moderately large islands, of nearly equal size, are in- 
cluded in one reef. Within the reef of the Gambler 
group there are four large and some smaller islands 
(fig. 8, Plate I.) ; within that of Hogoleu (fig. 2, 
Plate I.) nearly a dozen small islands are scattered over 
the expanse of one vast lagoon. 

After the details now given, it may be asserted that 
there is not one point of essential difference between 
encircling barrier-reefs and atolls ; — ^the latter enclose 
a simple sheet of water, the former encircle an expanse 
with one or more islands rising from it. I was much 
struck with this fact, when viewing, from the heights 
of Tahiti, the distant island of Eimeo standing within 
smooth water, and encircled by a ring of snow-white 
breakers. Eemove the central land, and an annular 
reef like that of an atoll in an early stage of its forma- 
tion is left ; remove Bolabola, and there remains a 
circle of linear coral-islets crowned with tall cocoa- 
nut trees, like one of the many atolls scattered over the 
Pacific and Indian oceans. 

The barrier-reefs of Australia and of New Caledonia 
deserve a separate notice from their great dimensions. 
The reef on the west coast of New Caledonia (fig. 5, 
Plate II.) is 400 miles in length ; and for a length of 
many leagues seldom approaches within eight miles of 
the shore. Near the southern end of the island, the 
space between the reef and the land is sixteen miles in 
width. The Australian barrier extends, with a few 
interruptions, for about eleven hundred miles; its 
average distance from the land is between twenty and 


thirty milos, but in parts from fifty to ninety. The 
great arm of the sea thus included, is from ten to 
twenty-five fathoms deep, with a sandy bottom ; but 
towards the southern end where the reef is further 
from the shore, the depth gradually increases to forty, 
and in some parts to more than sixty fathoms. Flinders 
has described the surface of the reef as consisting of a 
hard white agglomerate of different kinds of coral, 
with rough projecting points. A few low islets have 
been formed on it. The outer edge is the highest 
part ; it is traversed by narrow gullies, and at intervals 
by ship-channels. The sea close outside is in most 
parts profoundly deep ; but to the north, near New 
Guinea, and to the south, the depth is much less, and 
here the bottom slopes gradually from the reef, as it 
generally does in front of the ship-channels.^ 

There is one important point in the structure of 
barrier-reefs which must here be considered. The 
accompanying diagrams represent north and south ver- 
tical sections, taken through the highest points of Vani- 
koro. Gambler, and Maurua islands, as well as through 
their ench'cling reefs. The scale both in the horizontal 
and vertical direction is the same, namely, a quarter of 
an inch to a nautical mile. The height and width of 
these islands are known ; and I have attempted to repre- 
sent the form of the land from the shading of the hills 

' The foregoing details are taken chiefly from Flinders' Voyage 
to Terra Australis, vol. ii. p. 88 ; but these have been corrected by 
the account given by Prof. Jukes, Narrative of the Voyage of the Fly, 
vol. i. 1817, cliap. xiii. 

Ch. II. 



in the large published charts. It has long been 
remarked, even from the time of Dampier, that a 
considerable degree of relation subsists between the 
inclination of that part of the land which is beneath 
water and that above it : hence the dotted hne in the 
three sections is probably a moderately accurate repre- 
sentation of the actual submarine prolongation of the 
land. If we now look at the outer edge of the reef 

No. 4. 

South. J^orth. 

1 — Vanikoro, from the Atlas of the Voyage of the Astrolabe, by 
D. D'Urville. 

2 — Gambier Island, from Beechey. 

3 — Maurua, from the Atlas of the Voyage of the Coquille, by 

The horizontal line is the level of the sea, from which on the 
right hand a plummet descends, representing a depth of 200 fathoms, 
or 1,200 feet. The vertical shading shows the section of the land, 
and the horizontal shading that of the encircling barrier-reef ; from 
the smallness of the scale, the lagoon-channel could not be repre- 

A A — Outer edge of the coral-reefs, where the sea breaks. 

B B — The shore of the encircled islands. 


(A A), and bear in mind that the plummet on the right 
hand represents a depth of 1,200 feet, we must conclude 
that the vertical thickness of these barrier coral-reefs is 
very great. 

1 must observe,that if the sections had been taken in 
any other direction across those islands, or across other 
encircled islands,^ the result would have been the same. 
In the succeeding chapter it will be shown that reef- 
building polypifers cannot flourish at great depths, — for 
instance, it is highly improbable that they could exist 
at above one-eighth of the depth represented by the 
plummet on the right hand of the woodcut. Here then 
is a great apparent difficulty — how were the basal parts 
of these barrier-reefs formed. It will perhaps occur to 
some that the actual reefs formed of coral are not of 
great thickness, but that before their first growth the 
sea had deeply eaten into the coasts of these encu-cled 
islands, and had thus left a broad but shallow sub- 
marine ledge, on the edges of which the corals grew ; 
but if this had been the case, the shore would have 
been invariably bounded by lofty cliffs, and not have 
sloped down to the lagoon-channel, as it does in 
many instances. On this view,^ moreover, the cause 
of the reef springing up at such a great distance from 

- An East and West section across the Island of Bolabola and its 
barrier-reefs is given in the fifth chapter, for the sake of illustrating 
another point. The scale is -57 of an inch to a mile ; it is taken 
(rom the Atlas of the Voyage of the Coqidlle, by Duperrey. The 
depth of the lagoon-channel is exaggerated. 

2 The Eev. D. Tycrman and Mr. Bennett (Journal of Voyage and 
Travels, vol. i. p. 215) have briefly suggested this explanation of the 
origin of the encircling reefs of the Society Islands. 


the land, leaving a deep and broad moat -within, 
remains altogether unexplained. A supposition of the 
same nature and appearing at first more probable, is, 
that the reefs have risen from banks of sediment, 
which had accumulated round the shore previously to 
the growth of the coral ; but the extension of a bank 
to the saiae distance round an unbroken coast, and 
in front of deep arms of the sea (as in Eaiatea, see 
Plate II., fig. 3), which penetrate nearly to the heart of 
some encircled islands, is exceedingly improbable. And 
why, again, should the reef, in some cases steep on both 
sides like a wall, spring up at a distance of two, three, or 
more miles from the shore, leaving a channel often be- 
tween 200 and 300 feet deep — a depth which, we have 
good reason to believe, is too great for the growth of 
coral ? The existence, also, of this same channel pre- 
cludes the idea of the reef having grown outwards, on a 
foundation slowly formed by the accumulation of its own 
detritus and sediment. Nor, again, can it be asserted 
that the reef-building corals will not grow, excepting at 
a great distance from the land ; for, as we shall soon see, 
there is a whole class of reefs which take their name 
from growing (especially where the sea is deep) closely 
attached to the shore. At New Caledonia (see Plate 
XL, fig. 5), the reefs which run in front of the west coast 
are prolonged in the same line for 150 miles beyond the 
northern extremity of the island, and this shows that 
some explanation, quite different from any one of those 
just suggested is requisite. If the island had been 
originally prolonged to this distance, and if the northern 


end had been worn away until it was a little beneath 
the level of the sea, why sliould the coral-reefs have 
become attached, not on the central crest, but in tho 
Bame line with the reefs which still front the existing 
shores ? "We shall hereafter see, that there is one, and 
I believe only one solution of this difficulty. 

One other supposition to account for the position 
of encircling reefs remains, but it is almost too pre- 
posterous to be mentioned ; — namely, that they rest 
on enormous submarine craters surroundmg the in- 
cluded islands. When the size, height, and form of 
the islands in the Society group are considered, 
together with the fact that all are thus encircled, 
such a notion will be rejected by everyone. New 
Caledonia, moreover, besides its size, is composed of 
primitive formations, as are some of the Comoro 
Islands ; ^ and Aitutaki consists of calcareous rock. 
We must, therefore, reject the several explanations, 
and conclude that the vertical thickness of barrier- 
reefs, from then- outer edges to the foundation on 
which they rest (from A A in the sections No. 4 to 
the dotted lines), is really great: but this presents 
no real difficulty, as I hope to show hereafter when 
the upward growth of coral-reefs, during the slow 
subsidence of their foundation, is discussed. 

» I have been informed that this is the case by Dr. Allan 3I 
Forres, who has visited this grouj). 



Beefs of Mauritius — Shallow channel within the reef — Its sloiB 
filling up -Citrrents of water formed within it — Upraised reefs 
— Narrow fringing -reefs in deep seas — Reefs on the coast of E. 
Africa and of Brazil — Fringing-rcefs in very shallow seas, round 
hanks of sediment, and on worn-doiun islands — Fringing-reefs 
affected by currents of the sea — Coral coating bottom of the sea, 
hut not forming reefs. 

Fringing-eeefs, or, as they have been called by some 
voyagers, shore-reefs, whether sldrting an island or 
part of a continent, at first appear to differ little from 
barrier-reefs, except that they are generally of less 
breadth. As far as the superficies of the actual reef 
is concerned, this is the case ; but the absence of an 
interior deep-water channel, and the close relation in 
their horizontal extension with the probable slope of 
the adjoinmg land beneath the sea, present essential 
points of difi'erence. 

The reefs which fringe the island of Mauritius offer 
a good example of this class. They extend round its 
whole circumference, with the exception of two or 
tliree parts' where the coast is almost precipitous, and 

' This fact is stated on the authority of the Officier du Eoi, in 
his extremely interesting ' Voyage a I'lsle de France,' undertaken in 
1768. According to Captain Carniichael (Hooker's Bot. Misc., vol. ii. 
p. 316), on one part of the coast there is a space of sixteen miles 
without a reef. 


where, if as is probable the bottom of the sea has a 
similar inclination, the coral would have no foundation 
on which to become attached. A similar fact may 
sometimes be observed even in reefs of the barrier 
class, which follow much less closely the outline of 
the adjoining land ; as, for instance, on the S.E. and 
precipitous side of Tahiti, where the encircling reef is 
interrupted. On the western side of the Mauritius, 
which was the only part I visited, the reef generally 
lies at the distance of about half a mile from the 
shore ; but in some parts it is distant from one to two, 
and even three miles. Even in this last case, as the 
coast-land is gently inclined from the foot of the 
mountains to the sea-beach, and as the soundings 
outside the reef indicate an equally gentle slope 
beneath the water, there is no reason for supposing 
that the basis of the reef, formed by the prolongation 
of the strata of the island, lies at a greater depth than 
that at which the polypifers could begin constructing 
the reef. Some allowance, however, must be made for 
the outward extension of a foundation formed of sand 
and detritus, from the wear of the corals ; and this 
would give to the reef a somewhat greater vertical 
thickness than would otherwise be possible. 

The outer edge of the reef on the western or 
leeward side of the island, is tolerably well defined, 
and is a little higher than any other part. It chiefly 
consists of large strongly branched corals of the genus 
Madrepora, which also form a sloping bed some way 
out to sea: the hinds of coral growing in this part 


will be described in the ensuing chapter. Between 
the outer margin and the beach, there is a flat space 
with a sandy bottom and a few tufts of living coral ; in 
some parts it is so shallow, that people, by avoiding 
the deeper holes and gullies, can wade across it at low 
water ; in other parts it is deeper, seldom, however, 
exceeding ten or twelve feet, so that it offers a safe 
coasting channel for boats. On the eastern and 
windward side of the island which is exposed to a 
heavy surf, the reef was described to me as having 
a hard smooth surface, very slightly inclined inwards, 
just covered at low- water, and traversed by gullies ; 
it appears to be quite similar in structure to the reefs 
of the barrier and atoll classes. 

The reef of Mauritius, in front of every river and 
streamlet, is breached by a straight passage : at Grand 
Port, however, there is a channel like that within a 
barrier-reef : it extends parallel to the shore for four 
miles, and has an average depth of ten or twelve 
fathoms ; its presence may jDrobably be accounted for 
by two rivers which enter at each end of the channel, 
and bend towards each other. The fact of reefs of 
the fringing class being always breached in front of 
streams, even of those which are dry during the 
greater part of the year, will be explained, when the 
conditions unfavourable to the growth of coral are 
considered. Low coral-islets, like those on barrier- 
reefs and atolls, are seldom formed on reefs of this 
class, apparently owing in some cases to their narrow- 
ness, and m others to the gentle slope of the reef 


outside not yielding many fragments to the breakers. 
On the windward side, however, of the Mauritius, two 
or three small islets have been formed. 

It appears, as will be shown in the ensuing chapter, 
that the action of the surf is favourable to the vigorous 
growth of the stronger corals, and that sand or sedi- 
ment, if agitated by the waves, is hijurious to them. 
Hence it is probable that a reef on a shelving shore, 
like that of Mauritius, would at first grow up, not 
attached to the actual beach, but at some little distance 
from it ; and the corals on the outer margin would be 
the most vigorous. A shallow channel would thus be 
formed within the reef ; and this channel could be filled 
up only very slowly with sediment, for the breakers 
cannot act on the shores of the island, and they do 
not often tear up and cast inside fragments h-om the 
outer edge of the reef, whilst every streamlet carries 
away its mud in a straight line through breaches in 
the reef. But a beach of sand and of fragments of 
the smaller kinds of coral seems, in the case of Mauri- 
tius, to be slowly encroaching on the shallow channel. 
On many shelving and sandy coasts, the breakers tend 
to form a bar of sand a little way from the beach, with 
a slight increase of depth within it — for instance. Cap- 
tain Grey^ states that the west coast of Australia, hi lat. 
24°, is fronted by a sand bar about 200 yards in width, 
on which there is only two feet of water ; but within 
it the depth increases to two fathoms. Similar l)ars, 

• Captain Grey's Journal of Two Expeditions, vol. i. p. 309. 


more or less perfect, occur on other coasts. In these 
cases I suspect that the shallow channel, (which no 
doubt during storms is occasionally obliterated,) is 
scooped out by the flowing away of the water thrown 
beyond the line on which the waves break with the 
greatest force. At Pernambuco the bar of hard sand- 
stone, before alluded to, has the same external form 
and height as a coral reef, and extends nearly parallel to 
the coast ; within this bar currents, apparently caused 
by the water thrown over it during the greater part of 
each tide, run strongly, and are wearing away its inner 
wall. From these facts it can hardly be doubted that 
within most fringing reefs, especially within those 
lying some distance from the land, a return stream 
must carry away the water thrown over the outer edge ; 
and the current thus produced would tend to prevent 
the channel being filled up with sediment, and might 
even deepen it under certain circumstances. To 
this latter belief I am led, by finding that channels are 
almost universally jpresent within the fringing reefs of 
those islands which have undergone recent elevatory 
movements ; and this could hardly have been the case 
if the conversion of the very shallow channel into land 
had not been counteracted to a certain extent. 

A fringing-reef, if elevated in a perfect condition 
above the level of the sea, would present the singular 
appearance of a broad dry moat bounded by a low wall 
or mound. The author ^ of an interesting pedestrian 

' Voyage k I'lsle de France, par un Officier du Eoi, Part L 
pp. 192, 200. 


tour round the Mauritius seems to liave met with a 
structure of this kind : he says, * J'observai que la, ou 
la mer etale independamment des rescifs du large, 
il y a a terre une esjjcce d'effoncement, ou chemin 
couvcrt naturel. On y pourrait mettre du canon,' 
&c. In another place he adds, * Avant de passer le 
Cap, on remarque un gros banc de corail eleve 
de plus de quinze pieds : c'est une espece de res- 
cif, que la mer a abandonne : il regne au pied 
une longue flaque d'eau, dont on pourrait faire un 
bassin pour de petits vaisseaux.' But the margin of 
the reef, although the highest and most perfect part, 
from being most exposed to the surf, ■would generally 
during a slow rise of the land be either partially or 
entirely worn down to that level at which corals could 
renew then' growth on its upper edge. On some parts 
of the coast-land of Mauritius there are little hillocks 
of coral-rock, which are either the last remnants of a 
continuous reef, or of low islands formed on it. I 
observed two such hillocks between Tamarin Bay and 
the Great Black Pdver ; they were nearly 20 feet 
high, about 200 yards from the present beach, and 
about 30 feet above its level. They rose abruptly 
from a smooth surface, strewTi with worn fragments 
of coral. They consisted in their lower part of hard 
calcareous sandstone, and in then* upper of great 
blocks of several species of Astraa and Madrepora, 
loosely aggregated ; they were divided into irregular 
beds, dipping seaward, in one hillock at an angle of 8°, 
and in the other at 18°. The upraised reefs round 


this island have been much less worn and modified by 
the action of the sea than in most other cases. 

Many islands ' are fringed by reefs quite similar to 
those of Mauritius: but on coasts where the sea 
deepens very suddenly, the reefs are much narrower, 
and their limited extension seems evidently to depend 
on the high inclination of the submarine slope; — a 
relation which, as we have seen, does not exist in reefs 
of the barrier class. The fringing-reefs on steep coasts 
are frequently not more than from 50 to 100 yards in 
width: they have a nearly smooth, hard surface, 
scarcely uncovered at low-water, and without any 
interior shoal channel like that within those fringing- 
reefs which lie at a greater distance from the land. 
The fragments torn up during gales from the outer 
margin are thrown over the reef on the shores of the 
island. I may give as instances, Wateeo, where the 
reef is described by Cook as being 100 yards wide ; 
and Mauti and EHzabeth ^ Islands, where it is only 
50 yards in width : the sea round these islands is 
very deep. 

Fringing-reefs, like barrier-reefs, surround islands 

' I may give Cuba, as another instance ; Mr. Taylor (Loudon'a 
Mag. of Nat. Hist., vol. ix. p. 449) has described a reef several miles 
in length between Gibara and Vjaro, which extends parallel to the 
shore at the distance of between half and the third part of a mile, 
and encloses a space of shallow water, with a sandy bottom and 
tufts of coral. Outside the edge of the reef, which is formed of 
great branching corals, the depth is six and seven fathoms. This 
coast has been upheaved at no very distant geological period. 

2 Mauti is described by Lord Byron in the Voyage of H.M.S. 
Blonde, and Elizabeth Island by Captain Beechey. 


and front the shores of continents. In the charta 
of the eastern coast of Africa, by Captain Owen, 
many extensive IVinging-reefs are laid down ; — thus, 
for a space of nearly 40 miles, from lat. 1° 5' to 
1° 45' S., a reef frmges the sliore at an average 
distance of rather more than one mile, and therefore 
at a greater distance than is usual in reefs of this 
class ; but as the coast-land is not high, and as the 
bottom shoals very gradually, (the depth being only 
from 8 to 14 fathoms at a mile and a-half outside the 
reef), its extension thus far from the land offers no 
difficulty. The external margin of this reef is de- 
scribed as formed of projecting points ; and within 
it there is a channel from six to twelve feet deep, 
with patches of living coral. At Mukdeesha (lat. 
2° r N.) * the port is formed,' it is said,' * by a long 
reef extending eastward four or five miles, within 
which there is a narrow channel, with ten to twelve 
feet of water at low spring tides : ' it lies at the distance 
of a quarter of a mile from the shore. Again, in the 
plan of Mombas (lat. 4° S.) a reef extends for thirty- 
six miles, at the distance of from half a mile to one 
mile and a-quarter from the shore ; within it, there 
is a channel navigable ' for canoes and small craft,' 
between six and fifteen feet deep : outside the reef the 
depth is about 30 fathoms at the distance of nearly 
half a mile. Part of this reef is very symmetrical, and 
has a uniform breadth of 200 yards. 

• Owen's Africa, vol. i. p. 357 ; from which work the foregoing 
facts are likewise taken. 

Ch. in. PKINGING KEEFS. < i 

The coast of Brazil is in many parts fringed by reefs. 
Of these, some are not of coral formation ; for instance, 
those near Bahia and in front of Pernambuco ; but a 
few miles south of this latter city, the reef follows * every 
turn of the shore so closely, that I can hardly doubt it 
is of coral. It runs at the distance of three-quarters 
of a mile from the land, and within it the depth ia 
from ten to fifteen feet. I was assured by an intelli- 
gent pilot, that at Ports Frances and Maceio, the outer 
part of the reef consists of living coral, and the inner 
of a white stone full of large irregular cavities com- 
municating with the sea.^ The bottom of the sea oflf 
the coast of Brazil shoals gradually to between thirty 
and forty fathoms, at the distance of between nine and 
ten leagues from the land. 

From the description now given, we may conclude 
that the dimensions and structure of fringing-reefs 
depend entirely on the greater or less inclination of the 
submarine slope, conjoined with the fact, that reef- 
building polypifers can exist only at limited depths. 
It follows from this, that where the sea is very shallow, 
as in the Persian Gulf and in parts of the East Indian 
Archipelago, the reefs lose their fringing character, 

» Baron Eoussin's Pilote du Br6sil, and the accompanying hydro- 
grapbical memoir. See also the suiDplemeut to this volume on a Bar 
of Sandstone off Pernambuco. 

- [Kathbun (Amer. Nat., xiii. 539-551) describes a reef on the 
Brazilian coast. The lower part of the reef consists of true corals, 
the upper of nullipores and annelid tubes. The reef has a loose 
structure near the surface, compact below. The coral fragments cover- 
ing the channel within the reef ' form beds of considerable tliicknesa 
in places, often more or less consolidated.'] 


and appear as separate and irregularly scattered 
patches often of considerable area. As the conditions 
are less favourable in several respects on the inner 
side of these patches, the growth of the coral is more 
vigorous on the outside ; thus causing the reefs to be 
generally higher and more perfect in their marginal 
than in their central parts. Hence these reefs some- 
times assume (and this circumstance ought not to be 
overlooked) the ai)pearance of atolls ; but as they are 
based on a shallow foundation, and as their central 
expanse is much less deep and their form less defined, 
this resemblance is easily seen to be merely superficial. 
On the other hand, when, in a deep sea, banks of sedi- 
ment have accumulated round islands or submerged 
rocks, and they become fringed with reefs, they are dis- 
tinguished with difficulty from encircling barrier-reefs 
or atolls. In the West Indies there are reefs, which I 
should probably have arranged under these two classes, 
if the existence of large and level banks, lying a little 
beneath the surface and ready to serve as the basis 
for the attachment of coral, had not been present ; the 
formation of such banks through the accumulation of 
sediment being sufficiently evident. Fringing-reefs 
sometimes coat, and thus protect the foundations of 
islands, which have been worn down by the surf to the 
level of the sea. According to Ehrenberg, this has been 
extensively the case with the islands in the Red Sea, 
which formerly ranged parallel to the shores of the 
mainland, with deep water within them: hence the 
reefs now coating their bases, are situated relatively 


to the land like barrier-reefs, although not belonging 
to that class ; — but there are, as I believe, in the Eed 
Sea some true barrier-reefs. The reefs of this sea and 
of the West Indies will be described in the Appendix. 
In some cases, fringing-reefs appear to be considerably 
modified in outline by the course of the prevailing 
currents ; Dr. J. Allan informs me tha^t on the east 
coast of Madagascar, almost every headland and low 
point of sand has a coral-reef extending from it in 
a S.W. and N.E. line, parallel to the currents on that 
shore. I should think the influence of the currents 
chiefly consisted in causing an extension, in a certain 
direction, of a proper foundation for the attachment 
of the coral. Bound many intertropical islands, for 
instance the Abrolhos on the coast of Brazil sur- 
veyed by Captain FitzBoy, and, as I am informed by 
Mr. Cuming, round the Philippines, the bottom of the 
sea is entirely coated by irregular masses of coral, 
which although often of large size, do not reach the 
surface and form proper reefs. This must be owing 
either to insufficient growth, or to the absence of those 
kinds of corals which can withstand the breaking of 
the waves. 

The three classes, atoll-formed, barrier, and fringing 
reefs, together with the modifications just described of 
the latter, include all the most remarkable coral-forma- 
tions anywhere existing. At the commencement of 
the last chapter in the volume, where I detail the 
principles on which the map (Plate III.) is coloured, 
the exceptional cases will be enumerated. 



In this chapter I will give all the facts, which I have 
collected, relating to the distribution of coral-reefs, — • 
to the conditions favourable to their increase, — to the 
rate of their growth, — and to the depth at which they 
are formed. 

These subjects have an important bearing on tho 
theory of the origin of the different classes of coral- 

Section I. 

On the Distribution of Coral-Rccfs, and on the Conditions 
favourable to their Increase. 

With regard to the hmits of latitude over which coral- 
reefs extend, I have nothing new to add. The Ber- 
muda Islands in 32° 15' N., is the pomt furthest re- 
moved from the equator in which they appear to exist ; 
and their extension here so far northward no doubt is 
du(! to the warmth of the Gulf Stream. In the Pacific, 
the Loo Choo islands, in lat. 27° N., have reefs on theur 
Bhores, and there is an atoll in 28° 30', situated N.W. 


of the Sandwich Archipelago. In the Bed Sea there 
are coral-reefs in lat. 30°. In the Southern Hemisphere 
coral-reefs do not extend so far from the equatorial sea. 
In the Southern Pacific there are only a few reefs 
beyond the line of the tropic, but Houtmans Abrolhos, 
on the western shores of Australia, in lat. 29° S., are of 

The proximity of volcanic land, owing to tho lime 
generally evolved from it, has been thought to be 
favourable to the increase of coral-reefs. There is, 
however, no foundation for this view; for nowhere 
are coral-reefs more extensive than on the shores of 
New Caledonia and of north-eastern Australia, which 
consist of primary formations ; and the Maldiva, 
Chagos, Marshall, Gilbert, and Low Archipelagoes, 
the largest groups of atolls in the world, are formed 
exclusively of coral. 

The entire absence of coral-reefs in certain large 
areas within the tropical seas, is a remarkable fact. 
Thus no coral-reefs were observed during the survey- 
ing voyages of the Beagle on the west coast of South 
America south of the equator, or round the Galapagos 
Islands. It appears, also, that there are none * on this 
coast north of the equator ; Mr. Lloyd, who surveyed 
the isthmus of Panama, remarked to me, that although 
he had seen corals living in the Bay of Panama, yet he 
had never observed any reefs formed by them. I at first 
attributed this absence of reefs on the coasts of Peru and 

' I have been informed that this is the case, by Lieut. Ryder, E.N., 
and others who have had ample opportunities for observation. 


of the Galapagos Islands,' to the coldness of the cur- 
rents from the south, but the Gulf of Panama is one 
of the hottest pelagic districts in the world.^ In the 
central parts of the Pacilic there are islands entirely 
free from reefs ; and in some of these cases this appears 
to be due to recent volcanic action : but the existence 
of reefs, though scantily developed, and according to 
Dana, confined to one part of Hawaii (one of the Sand- 
wich Islands), shows that recent volcanic action docs 
not absolutely prevent their growth.^ 

' The moan temperature of the surface sea, from observations 
made by the direction of Captain FitzRoy on the shores of the 
Galapagos Islands, between the IGth of September and the 20th of 
October, 1835, was 68° Fahr. The lowest temperature observed was 
58°-o at the S.W. end of Albemarle Island ; and on the west coast 
of this island, it was several times 02° and 63^. The mean tem- 
perature of the sea in the Low Archipelago of atolls, and near Taliiti, 
from similar observations made on board the Beagle, was (although 
further from the equator) 77°'5, the lowest any day being 76°-5. 
Therefore we have here a difference of 9°*5 in mean temperature, 
and 18° in extremes ; a difference doubtless quite sufficient to affect 
the distribution of organic beings in the two areas. 

^ Humboldt's Personal Narrative, vol. vii. p. 43-4. 

3 [Mr. S. J. Whitmee (Nature, August 12, 1875, p. 291) states that 
in Savaii (Samoan grou2J), one of four exami^les of islands which 
Professor Dana brings forward as instances indicating that recent 
volcanic action has prevented the formation of extensive coral-reefs, 
the cause is more probably the depth of water on the coast. More- 
over, parts of Savaii differ in change of level from the rest of tho 
island, and it is in these (the upheaved regions) that coral-reefs are 
almost wanting. He also says that after examining ' a good many 
intertropical islands of the Pacific belonging to tho three orders — 
(1) Volcanic islands with fringing coral-reefs, such as Samoa, the New 
Hebrides, &c. ; (2) Atolls, such as the Low Archipelago, Ellice, Gilbert 
Islands, &c.\ (3) Upraised coral-islands, such as Niue or Savage 
Island, part of the Friendly, the Loyalty Islands, tfec' — he has been 
the more firmly convinced, the further he has gone, of the correctness 
of Mr. Darwin's theory. Mr. E. Webb (Nature, id. p. 475) disputes 


In tlie last chapter I stated that the bottom of the 
sea round some islands is thickly coated with living 
corals, -vrhich nevertheless do not form reefs, either 
from insufficient growth, or from the species not being 
adapted to contend with the breaking waves. 

I have been assured by several navigators that 
there are no coral-reefs on the west coast of Africa,^ or 
round the islands in the Gulf of Guinea. This perhaps 
may be attributed to the sediment brought down by 
the many rivers debouching on that coast, and to the 
extensive mud-banks which line great part of it. But 
the islands of St. Helena, Ascension, the Cape Verdes, 
St. Paul's, and Fernando Noronha, are, also, entirely 
destitute of reefs, although they lie far out at sea, are 
composed of the same ancient volcanic rocks, and have 
the same general form with those islands in the Pacific, 
the shores of which are surrounded by gigantic walls of 
coral-rock. With the exception of Bermuda, there is 
not a single coral-reef in the central expanse of the 
Atlantic ocean. It will, perhaps, be suggested that 
the quantity of carbonate of lime in diflerent parts of 
the sea may regulate the presence of reefs. But this 
cannot be the case, for at Ascension, the waves, charged 
to excess, precipitate a thick layer of calcareous matter 

Mr. Whitmee's statement as to the upheaval of the above-mentioned 
part of Savaii. (From materials collected by Mr. Darwin.)] 

' It might be conchaded, from a paper by Captain Owen (Geo- 
graph. Journ., vol. ii. p. 89), that the reefs off Cape St. Anne and the 
Sherboro' Islands were of coral, although the author states that they 
are not purely coralline. But I have been assured by Lieut. Hol- 
land, E.N., that these reefs are not of coral, or at least that tbey da 
not at all resemble those in the West Indies. 


on tlic tidal rocks ; and at St. Jago in the Cape Verdes, 
carbonate of lime not only is abundant on the shores, 
but it forms the chief part of some upraised post- 
tertiary strata. The apparently capricious distribution, 
therefore, of coral-reefs, cannot be fully explained by 
any of the above obvious causes ; but, as the study of 
the terrestrial and better-known half of the world, 
must convince everyone that no station capable of 
supporting life is lost, — nay more, that there is a 
struggle for each station between different organisms, 
— we may conclude that in those parts of the inter- 
tropical sea in which there are no coral-reefs, there 
are other organic beings, supplying the place of the 
reef-building polypifers. It has been shown in the 
chapter on Keeling atoll that there are some species of 
large j&sh, and the whole tribe of Holothuria?,^ which 
prey on the tenderer parts of the corals. On the other 
hand, the polypifers in their turn must prey on other 
organic beings ; and they would suffer by the diminu- 
tion of their j)rey through any cause. The relations, 
therefore, which determine the formation of reefs on 
any shore, by the vigorous growth of the efficient kinds 
of coral, must be very complex, and with our imperfect 
knowledge inexplicable. From these considerations, 
we may infer that changes in the condition of the sea, 
not obvious to our senses, might destroy all the coral- 
reefs in one area, and cause them to appear in another : 
thus, the Pacitic or Indian ocean might become as 
barren of coral-reefs as now is the Atlantic, without 

' [See Appendix ii.] 


our being able to assign any adequate cause for such a 

It has been a question with some naturalists, which 
part of a reef is most favourable to the growth of coral. 
The great mounds of living Porites and of Millepora 
round Keeling atoll occur exclusively on the extreme 
verge of the reef, which is washed by a constant suc- 
cession of breakers ; and living coral nowhere else 
forms solid masses. At the Marshall islands the larger 
kinds of corals (chiefly a species of Astrsea, a genus 
closely allied to Porites), * which form rocks measuring 
several fathoms in thickness,' prefer, according to 
Chamisso,^ the most violent surf. I have stated 
that the outer margin of the Maldiva atolls consists of 
living corals, (some of which, if not all, are of the same 
species with those at Keeling atoll), and here the surf 
is so tremendous, that even large ships have been 
thrown, by a single heave of the sea, high and dry on 
the reef, all on board thus escaping with their lives. 

Ehrenberg^ remarks, that in the Eed Sea the 

' I have left the foregoing paragraphs nearly as they stood m the 
first edition ; but, as stated in the Preface to the present work, Dana 
has shown that I have undervalued the importance of the mean 
temperature of the sea during the coldest season of the year, on the 
distribution of coral-reefs, as well as perhaps the injurious effects of 
recent volcanic action. But I cannot see that the absence of coral- 
reefs round certain islands in the Atlantic, for instance Ascension, 
St. Paul's Rock, and Fernando Noronha, or from the shores of the 
Gulf of Panama, is explicable through any known cause. 

2 Kotzebue's First Voyage (Eng. Transl.), vol. iii. pp. 142, 143, 

2 Ehrenberg, Ueber die Natur und Bildung der Corallen Biinke 
im rotheu Meere, p. 49. 


strongest corals live on the outer reefs, and appear to 
love the surf ; he adds, that the more branched kinds 
abound a little way within, but that these in still 
more protected places become smaller. Many other 
facts having a similar tendency might be adduced.' It 
has, however, been doubted by MM. Quoy and Gaimard, 
whether any kind of coral can even withstand, much 
less flourish in, the breakers of an open sea ; ^ they 
affirm that the saxigenous lithophytes flourish only 
where the water is tranquil, and the heat intense. 
This statement has passed from one geological work to 
another ; nevertheless, the protection of the whole reef 
is undoubtedly due to those kinds of coral, which 
cannot even exist in the situations thought by these 
naturalists to be most favourable to them. For should 
the outer and living margin perish, of any one of the 
many low coral-islands, round which a line of great 
breakers is incessantly foaming, the whole, it is 
scarcely possible to doubt, would be washed away and 
destroyed in less than half a century. But the vital 
energies of the corals conquer the mechanical power of 
the waves ; and the large fragments of reef torn up 
by every storm, are replaced by the slow but steady 
gi'owth of the innumerable polypifers which form the 
living zone on its outer edge. 

' In tbo West Indies, as I am informed by Captain Bird Allen, 
E.N., it is the common belief of those who are best acquainted with 
the reefs, that the coral flourishes most where freely exposed to the 
swell of the open sea. 

- Annates des Sciences Naturelles, tom. vi. pp. 27'!, 278. — ' La on 
IcB ondes sont agit(jes, lea Lytophytes ne pcuvent travailler, parce 
qu'elles diitruiraient leurs fragiles Edifices,' &c. 


From these facts, it is certain, that the strongest 
and most massive corals flourish where most exposed. 
The less perfect state of the reef of most atolls on the 
leeward and less exposed side, compared with its state 
to windward ; and the analogous case of the greater 
number of breaches on the near sides of those atolls 
in the Maldiva Archipelago which afford some pro- 
tection to each other, are obviously explained by this 
circumstance. If the question had been, under what 
conditions the greater number of species of coral, not 
regarding their bulk and strength, were developed, I 
should answer, — probably in the situations described 
by MM. Quoy and Gaimard, where the water is 
tranquil and the heat intense. The total number of 
species of coral in the circumtropical seas must be 
very great ; in the Eed Sea alone, 120 kinds, accord- 
ing to Ehrenberg,' have been observed. 

The same author has observed that the recoil of 
the sea from a steep shore is injurious to the growth 
of coral, although waves breaking over a bank are 
not so. Ehrenberg also states that where there is 
much sediment, placed so as to be liable to be moved 
by the waves, there is little or no coral ; and a col- 
lection of living specimens placed by him on a sandy 
shore died in the course of a few days.^ An experi- 
ment, however, will presently be related in which 
some large masses of living coral increased rax)idly in 
size, after having been secured by stakes on a sand- 
bank. That loose sediment should be injurious to 

' Ehrenberg Ueber die Natur, &c. &c., p. 46. '■' Ibid p. 49. 


the living polypifers, appears at first siglit probable ; 
and in sounding off Keeling atoll and Mauritius, the 
arming of the lead invariably came up clean, where 
the coral was growing vigorously. A strange belief, 
which, according to Captain Owen,^ is general amongst 
the inhabitants of the Maldiva atolls, namely, that 
corals have roots, and therefore grow up again if 
merely broken down to the surface, but if rooted 
out, are permanently destroyed, — I am inclined to 
believe arises from the fact that loose sand injures 
the polypifers. For it is probable that sand would 
accumulate in the hollows formed by tearing out the 
corals, but not on the broken and projecting stumps ; 
and therefore, in the former case, the fresh growth 
of the coral would be prevented. By this means 
the inhabitants keep their harbours clear ; and thus 
the French governor of St. Mary's, in Madagascar, 
* cleared out and made a beautiful little port at that 

In the last chapter I remarked, that fringing- 
reefs are almost universally breached where streams 
enter the sea.^ Most authors have attributed this fact 
to the injurious effects of the fresh water, even where 

' Captain Owen on the Geography of tlie Maldiva Islands, Geo- 
graph. Journal, vol. ii. p. 88. 

■■^ Lieut. Wellstead and others have remarked that this is the case 
in the Ecd Sea : Dr. Euiipell (Reise, in Abyss. Band. i. s. 142) says 
that there are pear-shaped harbours in the upraised coral-coast, into 
which periodical streams enter. From this circumstance, I presume, 
we must infer that, before the upheaval of the strata now forming 
the coast-land, fresh water and sediment entered the sea at these 
points ; and the coral being thus prevented growing, the pnar-sliaped 
harbours were produced. 


it enters the sea only in small quantity and during a 
IJart of the year. No doubt brackish water would pre- 
vent or retard the growth of coral ; but I believe that 
the mud and sand, which is deposited, even by small 
rivulets when flooded, is a much more efiicient check. 
The reef on each side of the channel leading into Port 
Louis at Mauritius, ends abruptly in a wall, at the 
foot of which I sounded, and found a bed of thick mud. 
This steepness of the sides appears to be a general 
character in such breaches : Cook,^ speaking of one 
at Eaiatea, says, * like all the rest, it is very steep on 
both sides.' Now, if it were the fresh water mingling 
with the salt, which prevented the growth of coral, 
the reef certainly would not terminate abruptly ; but as 
the polypifers nearest the impure stream would grow 
less vigorously than those farther off, so would the 
reef gradually thin away. On the other hand, the 
sediment brought down from the land would only 
prevent the growth of the coral in the line of its 
deposition, but would not check it on the side, so that 
the reefs might increase till they overhung the bed of 
the channel. The breaches are much fewer in number, 
and front only the larger valleys in reefs of the en- 
circling barrier class. They probably are kept open 
in the same manner as those into the lagoon of an 
atoll, namely, by the force of the currents and the 
drifting outwards of fine sediment. Their position m 
front of valle3^s, although often separated from the 
land by deep-water lagoon-channels, which it might 

' Cook's First Voyage, vol. ii. p. 271. (Hawkcsworth's edit.) 


be thought would entirely remove the injurious effects 
both of the fresh water and the sediment, will receive 
a simple explanation when we discuss the origin of 

In the vegetable kingdom every different station 
has its peculiar group of plants, and similar relations 
appear to prevail with corals. We have already de- 
scribed the great difference between the corals within 
the lagoon of an atoll and those on its outer margin. 
The corals, also, on the margin of Keeling Island oc- 
curred in zones : thus the Porites and Millepora compla- 
nata grow to a large size, only where they are washed by 
a heavy sea, and are killed by a short exposure to the 
air ; whereas, three species of Nullipora also hve amidst 
the breakers, but are able to survive uncovered for a 
part of each tide : at greater depths, a strong Madre- 
pora and Millepora alcicornis are the commonest kinds ; 
the former appearing to be confined to this part : be- 
neath the zone of massive corals, minute encrusting 
corallines and other organic bodies live. If we com- 
pare the external margin of the reef at Keeling atoll 
with that on the leeward side of Mauritius, which are 
very differently circumstanced, we shall find a corre- 
sponding difference in the appearance of the corals. At 
the latter place, the genus Madrepora is preponderant 
over every other kind ; and beneath the zone of massive 
corals there are large beds of Seriatopora. There is 
also amarked difference, according to Captain Moresby,* 

' Captain Moresby on the Northern Maldiva Atolls. Geograph. 
Journ., vol. v. p. 401. 


between the great branching corals of the Eed Sea and 
those on the reefs of the Maldiva atolls. 

These facts, which in themselves are deserving of 
notice, bear, perhaps, not very remotely on a remarkahlo 
circumstance which has been pointed out to me by 
Captain Moresby, namely, that with very few excep- 
tions, none of the coral-lmolls within the lagoons of Peros 
Banhos, Diego Garcia, and the Great Chagos Bank (all 
situated in the Chagos group), rise to the surface of the 
water ; whereas, with equally few exceptions, all those 
within Solomon and Egmont atolls in the same group, 
and likewise those within the large southern Maldiva 
atolls, reach the surface. I make these statements, after 
having examined the charts of each atoll. In the lagoon 
of Peros Banhos, which is nearly twenty miles across, 
there is only one single reef which rises to the surface: 
in Diego Garcia there are seven, but several of these lie 
close to the margin of the lagoon, and need scarcely 
have been reckoned : in the Great Chagos Bank there is 
not one. On the other hand, in the lagoons of some of 
the great southern Maldiva atolls, although thickly 
studded with reefs, every one without exception rises to 
the surface ; and on an average there are less than two 
submerged reefs in each atoll : in the northern atolls, 
however, the submerged lagoon-reefs are not quite so 
rare. The submerged reefs in the Chagos atolls gene- 
rally have from one to seven fathoms water on them, but 
Bome have from seven to ten. Most of them are small 
with very steep sides ; ^ at Peros Banhos they rise from 

' Some of these statements were not communicated to me verb- 


a depth of about thirty fathoms, and some of tliem in 
the Great Cliagos Bank from above forty fathoms: they 
are covered, Captain Moresby informs mo, with living 
and healthy coral two and three feet high, consisting 
of several species. Why then have not these lagoon- 
reefs reached the surface, like the innumerable ones in 
the atolls above named ? If we attempt to assign any 
difference in their external conditions, as the cause of 
this diversity, we are at once baflled : the lagoon of 
Diego Garcia is not deep, and is almost wholly sur- 
rounded by its reef ; Peros Banhos is very deep, much 
larger, with many wide passages communicating with 
the open sea. On the other hand, of those atolls in 
which all, or nearly all the lagoon-reefs have reached 
the surface, some are small, others large, some shallow, 
others deep, some well enclosed, and others open. 

Captain Moresby informs me that he has seen a 
French chart of Diego Garcia made eighty years before 
his survey, and apparently very accurate; and from it lio 
infers, that during this interval there has not been the 
smallest change in the depth on any of the knolls within 
the lagoon. It is, also, known that durmg the last fifty- 
one years, the eastern channel into the lagoon has 
neither become narrower, nor decreased in depth ; and 
as there are numerous small knolls of living coral within 
it, some change might have been anticipated. Moreover, 
as the whole reef round the lagoon of this atoll has been 
converted into land — an unparalleled case, I believe, in 

ally by Captain Moresby, but are taken from the MS. account, before 
alluded to, of the Cliagos Grcup. 


an atoll of such large size, — and as the strip of land is for 
considerable spaces more than half a mile wide — also a 
very unusual circumstance, — we have the best possible 
evidence that Diego Garcia has remained at its present 
level for a very long period. With this fact, and with 
the knowledge that no sensible change has taken place 
during eighty years in the coral knolls, and considering 
that every single reef has reached the surface in other 
atolls, which do not present the smallest appearance of 
being older than Diego Garcia and Peros Banhos, and 
which are placed under the same external conditions 
with them, one is led to conclude that these submerged 
reefs, although covered with luxuriant coral, have no 
tendency to grow upwards, and that they would remain 
at their present levels for an indefinite period. 

From the number of these knolls, from their posi- 
tion, size, and form, — many of them being only one or 
two hundred yards across, with a rounded outline and 
precipitous sides, — it is indisputable that they have been 
formed by the growth of coral; and this makes the case 
much more remarkable. In Peros Banhos and in the 
Great Chagos bank, some of these almost columnar 
masses are 200 feet high, and their summits lie only from 
two to eight fathoms beneath the surface ; therefore, 
a little greater proportional amount of growth would 
cause them to attain the surface, like those numerous 
knolls which rise from an equally great depth withhi 
the Maldiva atolls. We can hardly suppose that time 
has been wanting for the upward growth of the coral ; 
as in Diego Garcia, the broad annular strip of land, 


formed by the continued accumulation of detritus, 
bIiows how long this atoll has remained at its present 
level. We must look to some other cause than the 
rate of growth ; and I suspect it will be found in the 
reefs being formed of different species of corals, adapted 
to live at different depths. 

The Great Chagos bank is situated in the centre of 
the Cbagos group, and the Pitt and Speaker banks at 
its two extreme points. These banks resemble atolls, 
except in their external rim being about eight fathoms 
submerged, and in being formed of dead rock, with very 
little living coral on it : a portion nine miles long of 
the annular reef of Peros Banhos atoll is in the same 
condition. These facts, as will hereafter be shown, 
render it probable that the whole group at some 
former period subsided seven or eight fathoms ; and 
that the coral perished on the outer margins of those 
atolls which are now submerged, but that it contmued 
alive and grew up to the surface on the others now 
perfect. If all these atolls did formerly subside, and if 
from the suddenness of the movement or from any other 
cause, those species of corals which are best adapted 
to live at a certain depth, once got possession of the 
knolls, supplanting their former occupants, they would 
have little or no power to grow upwards. To illustrate 
this, I may observe that if the corals of the upper 
zone on the outer edge of Keeling atoll were to perish, 
it is improbable that those of the lower zone would grow 
to the surface, and thus become exposed to conditiona 
for which they do not appear to be adapted. The con- 


jecture that tlie corals on the submerged knolls with in 
the Chagos atolls have analogous habits with those of 
the lower zono outside Keeling atoll, receives some sup- 
port from a remark by Captain Moresby, namely, that 
they have a different appearance from those on the reefs 
in the Maldiva atolls, which, as we have seen, all rise to 
the surface : he compares the kind of difference to that 
of the vegetation under different climates. I have 
entered at considerable length into this case, although 
unable to throw much light on it, in order to show that 
coral-reefs situated in different places or at different 
depths, whether forming the rmg of an atoll or the 
knohs within a lagoon, need not all be supposed to 
have an equal tendency to upward growth. The infer- 
ence, therefore, that one reef could not grow to the 
surface within a given time, because another, not 
known to be covered with the same species of corals, 
and not known to be placed under exactly the same 
conditions, has not within the same time reached the 
surface, is unsound. 

Section II. 

On the Rate of Growth of Coral-reefs. 

The remark made at the close of the last section, 
naturally leads to this division of our subject, which 
has not, I think, hitherto been considered under a 
right point of view. Ehrenberg ^ has stated that in 
the Red Sea, the corals only coat other roclis in a 

' Ehrenberg, as before cited, pp. 39, 46, and 50. 


laj'er from one to two feet in thickness, or at most to 
a fathom and a-half ; and he disbelieves that, in any 
case, they form by their own proper growth great 
masses, stratmn over stratum. A nearly similar ob- 
servation has been made by MM. Quoy and Gaimard,' 
with respect to the thickness of some upraised beds of 
coral, which they examined at Timor and some other 
places. Ehrenberg^ saw certain large massive corals 
in the Eed Sea, which he imagines to be of such vast 
antiquity, that they might have been beheld by 
Pharaoh; and according to Sir C. LyelP there are 
certain corals at Bermuda, which are known by tra- 
dition to have been living for centuries.^ To show 
how slowly coral-reefs grow upwards. Captain Beechey* 
has adduced the case of the Dolphin Eeef off Tahiti, 
which has remained at the same depth beneath the 
surface, namely, about two fathoms and a-half, for a 
period of sixty-seven years. There are reefs in the 
Eed Sea, which certamly do not appear ^ to have in- 

' Annales des Sciences Nat., torn. vi. p. 28. 
' Ehrenberg, ut sup. p. 42. 

* Lyell's Principles of Geology, book iii. ch. xviii. 

* Since the preceding pages (of the first edit.) have been printed 
off, I have received from Sir C. Lyell an interesting pamphlet, en- 
titled Remarks upon Coral-Formations, &c., by J. Couthouy, Boston, 
United States, 1812. A statement (p. C) is here given on the 'lutho- 
rity of the Rev. J. Williams, corroborating the above remarks on the 
antiquity of certain individual corals, namely, that at Upolu, one of 
the Navigator islands, ' particular clumps of coral are known to ihe 
fishermen by name, derived from either some particular configuration 
or tradition attached to them, and handed down from time imme- 

* Beechey's Voyage to the Pacific, ch. viii. 
" Ehrenberg, ut sup. p, 43. 


creased in dimensions during the last half century, and 
from the comparison of old charts with recent surveys, 
probably not during the last two hundred years. 
These, and other similar facts, have so strongly im- 
pressed many with the belief of the extreme slowness 
of the growth of corals, that they have even doubted 
the possibility of islands in the great oceans having 
been formed by their agency. Others again, who have 
not been overwhelmed by this difficulty, have ad- 
mitted that it would require thousands, and tens of 
thousands of years, to form a mass even of incon- 
siderable thickness : but the subject has not, I believe, 
been viewed in the proper light. 

That masses of considerable thickness have been 
formed by the growth of coral, may be inferred with 
certainty from the following facts. In the deep 
lagoons of Peros Banhos and of the Great Chagos 
bank, there are, as already described, small steep- 
sided knolls covered with living coral. There are 
similar knolls in the southern Maldiva atolls, some of 
which, as Captain Moresby assures me, are less than 
a hundred yards in diameter, and rise to the surface 
from a depth of between 250 and 300 feet. Con- 
sidering their number, form, and position, it would be 
preposterous to suppose that they are based on pin- 
nacles of rock, or on isolated cones of sediment. As 
no kind of living coral grows above the height of a 
few feet, we are compelled to suppose that these knolls 
have been formed by the successive growth and death 


of many individuals, — first one being broken off or 
killed by some accident, and then another, and one set 
of species being replaced by another set with different 
habits, as the reef rose nearer the surface, or as other 
changes supervened. The spaces between the corals 
would become filled up with fragments and sand, and 
such matter would probably soon be consolidated, for 
we learn from Lieut. Nelson's ' observations at Bermuda 
that a process of this kind takes place beneath water, 
without the aid of evaporation. In reefs, also, of the 
barrier class, we may feel sure, as I have shown, that 
masses of great thickness have been formed by the 
growth of coral. In the case of Vanikoro, judging 
only from the depth of the moat between the land 
and the reef, the wall of coral-rock must be at least 
300 feet in vertical thickness. 

So again some of the upraised islands in the Pacific 
show what thick masses of coral-rock have been 
formed. Dana - states that Metia, in the Paumotu 
or Low Archipelago, consists of white solid limestone 
with some disseminated corals ; and this island once 
existed as an atoll, though now surrounded by chffs 
250 feet in height. The cliffs round Elizabeth Island 
in the same archipelago are 80 feet high, and are 
composed, according to Beechey, of homogeneous coral- 
rock. Mangaia in the Hervey Group, and Eurutu, 
appeal both to have once existed as encircled islands, 

' Geological Transactions, vol. v. p. 113. 

^ Corals and Coral Islands, 1872, p. 193. See also Mr. Couthouy's 
pamphlet above referred to. 


and their barrier-reefs are now in parts 300 feet above 
the level of the sea.' 

Some attempts have been made, with but little 
success, to ascertain by boring the thickness of coral 
formations. At Bow Island, in the Low Archipelago, 
Sir E. Belcher^ bored to a depth of 45 feet, and 
below the first 20 found only coral- sand. During 
Wilke's Expedition,^ in a boring of 21 feet in depth 
on one of the islands in the same archipelago, coral- 
sand was iDasscd through for the first 10 or 11 feet, 
and then solid reef rock. On one of the Maldiva 
atolls in the Indian Ocean, Captain Moresby bored to 
a depth of 26 feet, when his auger broke. He gave 
me the matter brought up, and it was perfectly white 
like finely triturated coral-rock. 

In my description of Keeling atoll, I have ad- 
vanced some facts showing that the reef has probably 
grown outwards ; and I found, just within the outer 
margin, the great mounds of Porites and of Millepora, 
with their summits lately killed, and their sides sub- 
sequently thickened by the growth of the coral : a 
layer, also, of Nullipora had already coated the dead 
surface. As the external slope of the reef is the same 
round the whole of this and many other atolls, the 
angle of inclination must result from an adaptation 

' Dana, Corals and Coral Islands, p. 336. Also Forster's Voyage 
round the World with Cook, vol. ii. pp. 163, 167. Williams's Narra- 
tive of Missionary Enterprise, pp. 30, 48, and 249. 

2 Voyage Round the World, vol. i. 1843, p. 369. 

^ Narrative U.S. Exploring Expedition, vol. iv. p. 268. Dana, 
Corals and Coral Islands, p. 184. 

100 RATE or GROWTH. Cn. IV 

between tlie growing powers of the coral and the 
force of the breakers, and their action on the loose 
sediment. The reef, therefore, could not increase out- 
wards without a nearly equal addition to every part 
of the slope, so that the original inclination might be 
preserved, and this would require a large amount of 
sediment, all derived from the wear of corals and 
shells, to be added to the lower part. Moreover, at 
Keeling atoll and probably in many other cases, the 
different kinds of coral would have to encroach on each 
other; thus the Nulliporse cannot increase outwards 
without encroaching on the Porites and Millepora 
zomplanata, as is now taking place ; nor these latter 
without encroaching on the strongly branched Madre- 
pora, the Millepora alcicornis, and some Astr^eas ; nor 
these again without a foundation being formed for 
fchem within the requisite depth, by the accumulation 
of sediment. How slow, then, must be the ordinary 
lateral or outward growth of such reefs ! But ofi 
Christmas atoll, where the sea is much more shallow 
than is usual, we have good reason to believe that, 
within a period not very remote, the reef has in- 
creased considerably in width. The land has the 
extraordinary breadth of three miles ; it consists of 
parallel ridges of shells and broken corals, which 
furnish * an incontestable proof,' as observed by Cook,* 
* that the island has been produced by accessions from 
the sea, and is in a state of increase.' The land is 
ironte d by a coral-reef, and from the manner in which 

' Cook's Third Vovage, book iii. ch. i. 

Sect. H. EATE OF GROWTH. 101 

islets are known to be formed, we may feel confident 
that the reef was not three miles wide when the first, 
or most inward ridge, was thrown up ; and, there- 
fore, we must conclude that the reef has grown out- 
wards during the accumulation of the successive ridges. 
Here then, a wall of coral-rock of very considerable 
breadth has been formed by the outward growth of 
the living margin, within a period, during which 
ridges of shells and corals, lying on the bare surface, 
have not decayed. There can be little doubt, from the 
account given by Captain Beechey, that Matilda atoll in 
the Low Archipelago has been converted in the space 
of thirty-four years, from being, as described by the 
crew of a wrecked whaling vessel, a * reef of rocks,' 
into a lagoon-island fourteen miles in length, with 
* one of its sides covered nearly the whole way with 
high trees.' ^ The islets, also, on Keeling atoll, it has 
been shown, have increased in length, and since the 
construction of an old chart, several of them have 
become united into one long islet : but in this case, 
and in that of Matilda atoll, we have no proof that 
the foundation of the islets, namely the reef, has 
increased in breadth, although it must be allowed that 
this is probable. 

I think, therefore, in regard to the possible rate of 
outivard growth of coral-reefs, but little importance 
need be attached to the fact that certain reefs in the 
Bed Sea have not increased during a long interval of 
time, or to other similar cases, such as that of Ouluthy 

' Beechey's Voyage to the Pacific, ch. vii and viii. 


atoll in the Caroline group, where every islet, described 
a hundred years before by Cantova, was found in the 
same state by Lutke.^ For it cannot be shown that, 
in these cases, the conditions were favourable to the 
vigorous and unopposed growth of the corals hving 
in the different zones of depth, and that a proper basis 
for the extension of the reef was present. These 
conditions must depend on many contingencies, and 
a basis within the requisite depth can rarely be pre- 
sent in the deej) oceans where coral formations most 

Nor do I think, when we consider the rate of the 
7ipward growth of reefs under favourable circumstances, 
that we should be influenced by the fact that certain 
submerged reefs, such as those off Tahiti or those within 
Diego Garcia, are not now nearer the surface than they 
were many years ago. For it has been shown that all 
the reefs have grown to the sm-face in some of the 
Chagos atolls, but that in neighbouring atolls which 
appear to be of equal antiquity and to be exposed to 
the same external conditions, every reef remains sub- 
merged ; we are, therefore, almost driven to attribute 
this to a difference, not in the rate of growth, but in 
the habits of the corals in the two cases. 

In an old-standing reef, the corals, which greatly 
differ in kind on different parts of it, are probably 

' F. Lml:6's Voyage autour du Monde. In the group Elato, 
however, it api^ears tbat what is now the islet Fahpi, is called in 
Cantova's Chart, the Banc de Falipi. It is not stated whether this 
has been caused by the growth of coral, or by the accumulation of 

Sect. H. KATE OF GROWTH. 103 

all adapted to the stations they occupy, and hold their 
places, like other organic beings, by a struggle one 
■with another and with external nature ; hence we may 
infer that their growth would be slow except under 
peculiarly favourable circumstances. Almost the only 
natural condition, allowing a quick upward growth of 
the whole surface of a reef, would be a slow subsidence 
of the area in which it stood ; — if, for instance, Keeling 
atoll were to subside two or three feet, can we doubt 
that the projecting margin of live coral, about half an 
mch in thickness, which surrounds the dead upper sur- 
faces of the mounds of Porites, would in this case form 
a concentric layer over them, and the reef thus increase 
upwards, instead of, as at present, outwards ? The 
NullipCrge are now encroaching on the Porites and 
Millepora, btit in this case might we not confidently 
expect that the latter would, in their turn, encroach 
on the NuUiporae ? After a subsidence of this kind, the 
sea would gain on the islets, and the great fields of dead 
but upright corals in the lagoon would be covered by a 
sheet of clear w'ater ; and might we not then expect 
that these reefs would rise to the surface, as they an- 
ciently did when the lagoon was less confined by islets, 
and as they did within a period of ten years m the 
schooner-channel cut by the inhabitants. In one of the 
Maldiva atolls, a reef, which withm a very few years 
existed as an islet bearing cocoa-nut trees, was found 
by Lieut. Prentice ' entirely covered tvitli lice coral and 
Madrepore.' The natives believe that the islet was 
washed away by a change in the currents, but if. 


instead of this, it had quietly subsided, Burely every 
part of the island which offered a BoUd foundation, 
would in a like manner have become coated with living 

Through steps such as these, any thickness of rock 
composed of a singular intermixture of various kinds 
of corals, shells, and calcareous sediment, might be 
formed ; but without subsidence, the thickness would 
necessarily be determined by the depth at which the 
reef-buildmg polypifers can exist. If it be asked, at 
what rate in years I suppose a reef of coral favourably 
circumstanced could grow up from a given depth ; I 
must answer that we have no precise evidence on 
this head. It will, however, be hereafter shown that 
in certain large areas where subsidence has probably 
been long in progress, the growth of the corals has 
been sufficient to keep the reefs up to the surface ; and 
this is a much more important standard of comparison 
than any cycle of years. 

It may, however, be inferred from the following 
facts, that the rate under favom-able circumstances 
would be far from slow. Dr. Allan of Forres has 
given, in his MS. Thesis deposited in the library of 
the Edinburgh University, the foUowmg account of 
some experiments, which he tried during his travels 
in the years 1830 to 1832 on the east coast of 
Madagascar. ' To ascertain the rise and progress 
of the coral family, and fix the number of species 
met with at Foul Point (lat. 17° 40'), twenty spe- 
cies of coral were taken off the reef and planted 

Sect. II. RATE OF GROWTH. 105 

apart on a sand-bank three feet deep at low water. 
Each portion weighed ten pounds, and was kept in 
its place by stakes. Similar quantities were placed 
in a clump and secured as the rest. This was done 
in December 1830. In July following, each detached 
mass was nearly level with the sea at low water, quite 
immovable, and several feet long, stretching, like the 
parent reef, in the line of the coast-current from north 
to south. The masses accumulated in a clump were 
found equally increased, but some of the species in 
such unequal ratios as to be growing over each other.' ^ 
The loss of Dr. Allan's magnificent collection by 
shipwreck, unfortunately prevents its being known 
to what genera these corals belonged ; but from the 
numbers experimented on, it is certain that all the 
more conspicuous kinds must have been included. 
Dr. Allan informs me, in a letter, that he believes 
it was a Madrepora which grew most vigorously. 
One may be permitted to suspect that the level of the 
sea might possibly have been somewhat different at the 
two stated periods ; nevertheless, it is quite evident 
that the growth of the ten-pound masses, during the 
six or seven months at the end of which they were 
found to be immovably fixed ^ and several feet in 
length, must have been very great. The fact of the 

' I owe the above extract to the kindness of Dr. Malcolmson. 

2 It is stated by Mr. De la Beche (Geological Manual, p. 143), on 
the authority of Mr. Lloyd, who surveyed the Isthmus of Panama, 
that some specimens of Polypifers, placed by him in a sheltered 
pool of water, were found in the course of a few days firmly fixed by 
the secretion of a stony matter, to the bottom. 


different kinds of coral, when placed in one clump, 
having increased in extremely unequal ratios, is very 
interesting, as it shows the manner in which a reef, 
supporting many species of coral, would probahly be 
ali\;cted by a change in the external conditions 
favouring one hind more than another. The growth 
of the masses of coral in N. and S. lines parallel to 
the prevailing currents, whether due to the drift- 
ing of sediment or to the simple movement of the 
water, is, also, an interesting circumstance. 

Lieut. Wellstead, I.N., informed me that in the 
Persian Gulf a ship had her copper bottom en- 
crusted in the course of twenty months with a layer 
of coral two feet in thickness, which it required great 
force to remove when the vessel was docked : it was 
not ascertained to what order this coral belonged.^ 

' Mr. Stutchbury (West of England Journal, No. I. p. 50) has 
described a specimen of Agaricia, ' weighing 2 lbs. 9 oz., which sur- 
rounds a species of oyster, whose age could not be more than two 
years, and yet is completely enveloped by this dense coral.' I pre- 
sume that the oyster was living when the specimen was procured ; 
otherwise the fact tells nothing. Mr. Stutchbury also mentions an 
anchor, which had become entirely encrusted with coral in fifty 
years; other cases, however, are recorded of anchors having long 
remained amidst coral-reefs without having become coated. The 
anchor of the Beagle, in 1832, after having been down exactly one 
month at Eio de Janeiro, was so thickly coated by two specie? of 
Tubularia, that large spaces of the iron were entirely concealed ; the 
tufts of this horny zoophyte were between two and three inches in 
length. Spallanzani states (Travels, Eng. Translat. vol. iv. p. 313) 
that in the Mediterranean, the red coral of commerce is usually 
dredged every ten years, during which time it grows to a height of 
one foot. It grows, however, at difierent rates in different places. 
It has been erroneously attempted to compute the rate of growth of 
a reef, from the fact mentioned by Captain Beechey of the Chama 

Sect. II. RATE OF GROWTH. 107 

This fact in some degree corroborates the result of 
Dr. Allan's experiments. The case of the schooner- 
channel, choked up with coral in an interval of less 
than ten years, in the lagoon of Keehng atoll, should 
be here borne in mind. We may also infer, from 
the trouble which the inhabitants of the Maldiva 
atohs take to root out, as they express it, the coral- 
knolls from their harbours, that their growth can 
hardly be very slow. I may add, that M. Duchassaing 
broke off all the Madrepores growing on a marked 
place in a bay at Guadaloupe ; and in the course of 
two months he found there a greater number of 
corals than before.^ 

From the facts given in this section, it may be 
concluded, first, that considerable thicknesses of rock 
have certainly been formed within the present geo- 
logical era by the growth of corals and the accumu- 
lation of their detritus ; and, secondly, that the 
increase of individual corals and of reefs, both out- 
wards or horizontally, and upwards or vertically, under 
conditions favourable to such increase, is not slow, 
when referred either to the standard of the average 

yigas being embedded in coral rock. But it should be remembered, 
that some species of this genus invariably live, both whilst young 
and old, in cavities, which the animal has the power of enlarging 
with its growth. I saw many of these shells thus embedded in the 
outer 'flat' of Keeling atoll, which is composed of dead rock; and 
therefore the cavities in this case had no relation whatever to the 
growth of coral, M. Lesson, also, speaking of this shell (Partie 
Zoolog., Voyage de la Coquille), has remarked, 'que constamment 
Bes valves etaient engagees completement dans la masse des Madr6- 

' L'Institut, 184G, p. 111. 


oscillations of level in the earth's crust, or to the 
more precise but less important one of a cycle of 

Section III. 

On the Depths at which Reef -building Corals live. 

I HAVE already described in detail the nature of the 
bottom of the sea immediately surrounding Keehng 
atoll ; and I will here describe with almost equal care, 
the soundings off the fringing-reefs of Mauritius. I 
sounded with the wide bell-shaped lead which Captain 
FitzEoy used at Keeling Island. My examination 
of the bottom was confined to a few miles of coast 
(between Port Louis and Tomb Bay) on the leeward 
side of the island. The edge of the reef is formed 
of great shapeless masses of branching Madrepores, 
which chiefly consist of two species, — apparently 
M. corpnhosa and pociWfera, — mingled with a few 
other kinds of coral. These masses are separated from 
each other by the most irregular gullies and cavities, 

' [See Dana, Corals and Coral Islands, ch. i. sec. iv. for additional 
facts relating to rate of growth of corals. Le Conte (Amer. Jour. Sci. 
Ser. 3, vol. x. pp. 34 -6) estimates that a Madrcpora (cervicornis ?) in 
shoal water at the Tortugas grew upwards at the rate of 3 .J inches 
per annum. Duncan (Proc. Eoy. Soc. xxvi. 133) estimates in the 
case of Lophohelia pwlifcra and Desmophyllum eriHtagalli growing in 
deep water to the north-west of Spain (522 to 550 fathoms) an in- 
ciease upwards at the rate of 0-29 inches per annum. The result of 
later researches indicates considerable variation in the rate of 
qrowth, depending probably on species, locality, &c., and coulirms the 
general conclusions of this paragraph.] 


into which the lead sinks many feet. Outside thia 
u'regular border of Madrepores, the water deepens 
gradually to 20 fathoms, which depth generally is 
found at the distance of from half to three-quarters 
of a mile from the reef. A little further out the depth 
is 30 fathoms, and thence the bank slopes rapidly 
into the depths of the ocean. This inclination is very 
gentle compared with that outside Keeling and other 
atolls, but compared with most coasts it is steep. The 
water was so clear outside the reef, that I could 
distinguish every object forming the rugged bottom. 
In this part, and to a depth of 8 fathoms, I 
sounded repeatedly, and at each cast pounded the 
bottom with the broad lead ; nevertheless the arming 
invariably came up perfectly clean, but deeply in- 
dented. From 8 to 15 fathoms a little calca- 
reous sand was occasionally brought up, but more 
frequently the arming was simply indented. In all 
this space the two Madrepores above mentioned, and 
two species of Astraea with rather large stars, seemed 
the commonest kinds ; and it must be noticed that 
twice at the depth of 15 fathoms, the arming was 
marked with a clean impression of an Astrgea. 
Besides these lithophytes, some fragments of the 
Millepora alcicornis which occurs in the same relative 
position at Keeling Island, were brought up ; and in 
the deeper parts there were large beds of a Seriato- 
pora, different from S. suhulata, but closely allied to 
it. On the beach within the reef, the rolled fragments 
consisted chiefly of the corals just mentioned, and of 


a massive Poritcs like that at Keeling atoll, of a 
Meandrina, Podllopora verrucosa, and of numerous 
fragments of Nullipora. From 15 to 20 fathoms the 
bottom was, with few exceptions, either formed of 
Band, 01 thickly covered with Seriatopora : this delicate 
coral seems to form at these depths extensive beds 
unmingled with any other kind. At 20 fathoms, one 
sounding brought up a fragment of Madrepora, ap- 
parently M. pocillifera, and I believe it to be the 
same species as that which mainly forms the upper 
margin of the reef ; if so, it grows in depths varying 
from to 20 fathoms. Between 20 and 23 fathoms I 
obtained several soundings, and they all showed a sandy 
bottom, with one exception at 30 fathoms, when the 
arming came up scoojDed out as if by the margin of a 
large Caryophjdlia. Beyond 33 fathoms I sounded 
only once ; and from 8G fathoms, at the distance of one 
mile and a third from the edge of the reef, the arming 
brought up calcareous sand with a pebble of volcanic 
rock. The chcumstance of the arming having in- 
variably come up quite clean when sounding within a 
certain number of fathoms off the reefs of Mauritius 
and Keeling atoll (8 fathoms in the former ease, 
and 12 in the latter), and of its having always 
come up (with one exception) smoothed and covered 
with sand when the depth exceeded 20 fathoms, 
probably indicates a criterion, by which the limits 
of the vigorous growth of coral might in all cases be 
readily ascertained. I do not, however, suppose that 
if a vast number of soundings were obtained round 


these islands, the limit above assigned would be found 
never to vary, but I conceive the facts are sufficient to 
show that the exceptions would be few. The circum- 
Btance of a gradual change, in the two cases, from a 
field of clean coral to a smooth sandy bottom, is far 
more important in indicating the depth at which the 
larger kinds of coral flourish, than almost any number 
of separate observations on the depth at which certain 
species have been dredged up. For we can understand 
the gradation only as a prolonged struggle against 
unfavourable conditions. If a person were to find the 
soil clothed with turf on the banks of a stream of 
water, but on going to some distance on one side of it 
he observed the blades of grass growmg thinner and 
thinner with intervening patches of sand, until he 
entered a desert of sand, he would safely conclude, 
especially if changes of the same kind were noticed in 
other places, that the presence of the water was abso- 
lutely necessary to the formation of a thick bed of 
turf: so may we conclude, with the same feeling of 
certainty, that thick beds of coral are formed only at 
small depths beneath the surface of the sea. 

I have endeavoured to collect every fact which 
might either invalidate or corroborate this conclusion. 
( 'aptain Moresby, whose opportunities for observation 
during his survey of the Maldiva and Chagos Archi- 
pelagoes were unrivalled, mforms me, that the upper 
part or zone of the steep-sided reefs on the inner and 
outer coasts of the atolls in both groups, invariably 
consisted of coral, and the lower parts of sand. At 


7 or 8 fathoms depth, the bottom is formed, as could 
be seen through the clear water, of great living 
masses of coral, which at about 10 fathoms generally 
stand some way apart from each other, with patches of 
white sand between them, and at a little greater depth 
these patches become united into a smooth steep slope 
without any coral. Captain Moresby, also, informs me 
in support of the above statement, that he only found 
decayed coral on the Padua Bank (northern part of the 
Laccadive group), which has an average depth of 25 
to 35 fathoms ; but that on some other banks in the 
same group, with a depth of only 10 or 12 fathoms 
(for instance, the Tillacapeni bank) the coral was 

Professor Dana likewise states that during the various 
and extensive surveys in the Pacific Ocean, made during 
the United States exploring expedition, no evidence 
was found of corals growing beyond the depth of 20 
fathoms.' I may here add that Sir E. Belcher, though 
he does not state to what depth living corals extended, 
says that many soundings were taken off Bow atoll, at 
depths ranging from 50 to 960 fathoms, and that the 
bottom always consisted of coral sand.^ 

"With regard to the coral-reefs in the Bed Sea, 
Ehrenberg has the following passage. ' The living 
corals do not descend there into great depths. On the 
edges of islets and near reefs, where the depth was 
small, very many lived ; but we found no more even at 

' Corals and Coral Islands, 1872, p. 116. 

■' Voyage Round the ^Yorld, 1843, p. 379, vol. i. 


Bix fathoms. The pearl-fishers at Yemen and Massaua 
asserted that there was no coral near the pearl-banks at 
nine fathoms depth, but only sand. We were not able 
to institute any more special researches.' ' I am, how- 
ever, assured both by Captain Moresby and Lieut. Well- 
stead, that in the more northern parts of the Eed Sea, 
there are extensive beds of living coral at a depth of 25 
fathoms, in which the anchors of their vessels were 
frequently entangled. Captain Moresby attributes the 
less depth at which the corals are able to live in the 
places mentioned by Ehrenberg, to the greater quantity 
of sediment there; the situations, where they were 
flourishing at the depth of 25 fathoms, were protected, 
and the water was extraordinarily limpid. On the 
leeward side of Mauritius, where I found the coral 
growing at a somewhat greater depth than at Keeling 
atoll, the sea, owing apparently to its tranquil state, 
was likewise very clear. Within the lagoons of some 
of the Marshall atolls, where the water can be but little 
agitated, there are, according to Kotzebue, living beds 
of coral in 25 fathoms. From these several facts, and 
considering the manner in which the beds of clean coral 
off Mauritius, Keeling Island, the Maldiva and Chagos 
atolls, graduated into a sandy slope, it appears very 
probable that the depth at which reef-building poly- 
pifers can exist, is partly determined by the extent of 
inclined surface which the currents of the sea and 
the recoiling waves have the power to keep free from 

' Ehrenberg, Ueber die Natur, &c p, 50. 

114 DErTir AT WHICH Cn. IV. 

MM. Quoy and Gaimard ' believe that the growth 
of coral is confined within very limited depths ; and 
they state that they never found any fragment of an 
Astra;a (the genus they consider most efficient in form- 
ing reefs) at a depth above 25 or 30 feet. But we 
have seen that in several places the bottom of the sea 
is paved with massive corals at more than twice this 
depth ; and at 15 fathoms (or thrice this depth) off the 
reefs of Mauritius the arming was marked with the 
distinct impression of a living Astrsea. Millepora 
alcicornis lives in from Oto 12 fathoms, and the genera 
Madrepora and Soriatopora from to 20 fathoms. 
Captain Moresby has given me a specimen of Sideropora 
scahm (Porites of Lamarck) brought up alive from 17 
fathoms. Mr. Couthouy ^ states that on the Bahama 
banks he dredged up considerable masses of Mean- 
drina from 16 fathoms, and has seen this coral growing 
in 20 fathoms. 

Captain Beechey mforms me that branches of pmk 
and yellow coral were frequently brought up from be- 
tween 20 and 25 fathoms off the Low atolls ; and Lieut. 
Stokes, writing to me from the N.W. coast of Australia, 
says that a strongly branched coral was procured there 
from 30 fathoms : unfortunately it is not known to what 
genera these corals belong. 

Although the limit of depth, at which each particular 
kind of coral ceases to exist, is thus far from being 
accurately known : yet when we bear in mind the 

' Annales des Sci. Nat. torn. vi. 

' Kemarks on Coral Formations, p. 12. 


manner in which the clumps of coral gradually became 
infrequent at about the same depth, and wholly dis- 
appeared at a greater depth than 20 fathoms on the 
slope round Keeling atoll, off the reefs in the Pacific 
(according to Dana), on the leeward side of the Mauri- 
tius, and at rather less depth both within and without 
the atolls of the Maldiva and Chagos Archipelagoes ; 
and when we know that the reefs round these islands do 
not differ from other coral formations in their form and 
structure, we ma.y, I think, conclude that in ordinary 
cases reef-building polypifers do not flourish at greater 
depths than between 20 and 30 fathoms, and rarely at 
above IS' fathoms.^ 

It has been argued ^ — that reefs may possibly rise 
from very great depths through the means of small 
corals first making a platform for the growth of the 
stronger kinds. This, however, is an arbitrary supposi- 
tion : it is not always remembered, that in such cases 
there is an antagonistic power at work, namely, the 
decay of organic bodies when not protected by a cover- 
ing of sediment or by their own rapid growth. We have, 
moreover, no right to calculate on unlimited time for 

' [The general conclusions of this paragraph do not appear to have 
been disturbed by recent researches, though Mr. Guppy (Proc. Eoy. 
Soc. Edin. xiii. p. 857, see Appendix II.) argues in favour of the 
possibility of reefs occasionally beginning to grow at depths of at 
least 50 fathoms ; and in the Eeports of the Challenger Expedition 
(Report on the Reef-building Corals, p. 35) cases of sjpecies which 
build reefs, living at a depth of 40 fathoms, and in two instances even 
at 70 fathoms, are recorded. Still even here it is admitted that ' the 
zone of most active growth is from 1 to 20 fathoms.'] 

- Journal of the Royal Geograph. Soc. 1831, p. 218. 



Cn. IV. 

the accumulation of small organic bodies into great 
nfasses.' Every fact in geology proclaims that neither 
tlie dry land nor the l)cd of the sea retains tlie same level 
for indctinite periods. As well might it be imagined 
that the British seas would in time become choked up 
with beds of oysters, or that the numerous small coral- 
lines off the inhospitable shores of Tierra del Fuego 
would in time form a solid and extensive coral- 

\ [This remark has, by anticipation, a direct bearing on an 
impoi'tant part of Mr. Murray's hypothesis. See Appendix II.] 

- I will here record the few facts which I have been able to collect 
as to the depths, both within and without the tropics, inhabited by 








Sertularia .... 


Cape Horn CO 

[Where none 




is given the 

,, A minute scarlet en- 

observation ig 

crusting species, found 

my own.] 

, living. 


Keeling At. 12- 

„ An allied, small stony 

sub-generic form 



A coral allied to Vincularia, 

with eight rows of cells 


Cape Horn 

Tubulipora, near to T. patina 



Do. do. 



Cellepora, several species and 

allied sub-generic forms 


Cape Horn 




1 ChonosArch. 
1 45° 



S. Cruz 50° 

Eschara. ^ , . . 


j Tierra del 
I Fuego 53° 



S. Cruz K. 50° 

Eetepora . - , . 


Cape Horn 




1 C. Good Hope 

mard, Ann. 

t 34° 

1 Kcicn. Nat., 
{ t. vi. p. 284. 

Sect. IIT. 







IkIillei:)ora, a strong coral with \ 
cylindrical branches, of a 
pink colour, abont two 
inches high, resembling in 
the form of its orihces M. 
aspera of Lamarck 

Coralium . . . . 
Antipathes . . . . 
Gorgonia (or an allied form) . 




E. Chiloe 43° 
Tierra del 
Fuego 53° 

Barbary 33° N. 

Chonos 45° 

(■Abrolhos, on 
•I the coast of 
[ Brazil, 18' 


Peyssonel, in 
paper read to 
Koyal Society 
May, 1752. 

Capt. Beechey 
informed me 
of this fact 
in a letter. 

those corals and corallines which we have no reason to suppose ever 
materially aid in the construction of a reef. Mr. Stokes also showed 
me a Caryophyllia which was dredged up alive by Captain King 
from a depth of 80 fathoms off Juan Fernandez, in lat. 63" S. EUis 
(Nat. Hist, of CoraUine, p. 96) states that Ombellularia was pro- 
cured in lat. 79° N. sticking to a Una from the depth of 236 fathoms ; 
hence this coral either must have been floating loose, or was en- 
tangled in a stray line at the bottom. Off Keeling atoll a compound 
Ascidia (Sigillina) was brought up from 39 fathoms, and a piece of 
sponge, apparently living, from 70, and a fragment of NuUipora, 
also apparently living, from 92 fathoms. At a greater depth than 
90 fathoms the bottom was thickly strewed with joints of a Halimeda 
and small fragments of other Nulliporre, but all dead. Captain B. 
Allen, R.N., informs me that in the survey of the Vv^est Indies it was 
noticed, that between the depth of 10 and 200 fathoms, the sounding- 
lead very generally came up coattd with the dead joints of a Hali- 
meda, of which he showed me specimens. Off Pernambuco, in 
Brazil, in about 12 fathoms, the bottom was covered with fragments, 
dead and alive, of a dull red NuUipora, and I infer from Koussin's 
chart, that a bottom of this kind extends over a wide area. On the 
beach, within the coral-reefs of Mauritius, vast quantities of frag- 
ments of Nulliporte were piled up. From these facts, it appears that 
these simply organised bodies, belonging to the vegetable kingdom, 
are amongst the most abundant productions of the sea. [Of late years 
corals, commonly solitary, have been found at much greater depths 

118 DErTH. Cn. TV. 

than those mentioned in this note ; for instance, Caryophyllia, 
down to at least 1,000 fathoms (The Depths of the Sea, pp. 28, 431), 
True corals referable to the Madroporaria ai'e not very abundant 
in deep water. According to Mr. Moseley's report, abuift ten genera 
reach a dc2)th of 1,000 fathoms ; four genera are found at 1,500 
fathoms ; and a single species extends practically through all depths, 
ranging from 30 to 2,900 fathoms. Challenger Ileports, vol. ii. 
(Zwlogy), ' On Corals,' pp. 132, 133. Pee also Keports, vol. xvi., 
•On Eeef-Corals,' p. 35. 



The atolls of the larger archipelagoes not formed on submerged craters, 
or on banks of sediment -Immense areas interspersed with atolls 
— Their subsidence— The effects of storms and earthquakes on 
atolls— Recent changes in their state— The origin of barrier-reefs 
and of atolls— Their relative forms— The step-formed ledges and 
walls round the shores of some lagoons— The ring-formed reefs of 
tlie Maldiva atolls — The submerged condition of parts or of the 
iohole of some annular reefs— The disscvcrmcnt of large atolls^ 
The union of atolls by linear reefs— The great Chagos Bank — 
Objections considered arising from the area and amount of sub- 
sidence required by the theory— The probable composition of the 
lower parts of atolls. 

The naturalists who have visited the Pacific, seem to 
have had their attention riveted by the lagoon-islands 
or atolls,— those singular rings oi" coral-land which 
rise abruptly out of the unfathomable ocean, — and 
have passed over, almost unnoticed, the scarcely less 
wonderful encircling barrier-reefs. The theory most 
generally received on the formation of atolls, is that 
they are based on submarine craters : but where can 
we find a crater of the shape of Bow atoll, which is five 
times as long as it is broad (Plate I., fig. 4) ; or like 
that of Menchicoff Island (Plate II., fig. 3), with its 


three loops, together sixty miles in length ; or like 
EimskyKorsacoll', narrow, crooked, and IH'ty-four miles 
long; or like the northern Maldiva atolls, made up of 
numerous ring-formed reefs, placed on the margin of 
a disk, — one of which disks is eighty-eight miles in 
length, and onl}' from ten to twenty in hreadth ? A 
further difficulty on this theory of the origin of atolls 
arises from the necessary assumption of so large a 
number of immense craters crowded together beneath 
the sea. But, as we shall presently see, a greater diffi- 
culty is involved, namely, that all these craters must 
lie within nearly the same level beneath the sea. 
Nevertheless, if the rim of a crater afforded a basis 
at the proper depth, I am far from denying that a 
reef like a perfectly characterized atoll might not be 
formed on it. Some such, perhaps, now exist ; but 
it is incredible that the greater number could have 
thus originated. 

An earlier and better theory was proposed by 
Chamisso : ^ he supposes that as the more massive kinds 
of corals prefer the sm-f , the outer portions of a reef will 
first reach the surface and consequently form a ring.* 
I remarked in the third chapter that a reef, growing on 
a detached bank, would tend to assume an atoll-hke 
structure ; if, therefore, corals were to grow up from a 
bank some fathoms submerged in a deep sea, having 
steep sides and a level surface, a reef not to be dis- 

' Kotzebue's First Voyage, vol. iii. p. 331.— 

* [By anticipation, some of the objections which have been raised 
of l^tc years (see Appendix II.) are considered in this section.] 

Cii. V. OF CORAL-EEEFS. 121 

tinguished from an atoll might be formed ; and I 
believe some such exist in the West Indies. But on 
this view it must be assumed, that in every case the 
basis consists of a flat bank ; for if it were conically 
formed like a mountainous mass, we can see no reason 
why the corals should spring up from the flanks instead 
of from the central and highest parts. As the lagoons 
of atolls are sometimes even more than 40 fathoms 
deep, it must, also, be assumed on this view, that at a 
depth at which the waves do not break, the coral grows 
more vigorously on the edges of a bank than on its 
central part : and this is an assumption without any 
evidence.' If we consider, moreover, the number of 
the atolls in the midst of the Pacific and Indian 
Oceans, this assumption of so many submerged banks 
is in itself very improbable. 

No theory worthy of notice has been advanced to 
account for those barrier-reefs which encircle islands 
of moderate dimensions. The great reef which fronts 
the coast of Australia has been supposed, but without 
any evidence, to rest on the edge of a submarme 
precipice parallel to the shore. The origin of the 
third class, or of fringing-reefs, presents, I believe, 
scarcely any difficulty, and arises simply from the poly- 
pifers growing in moderate depths, and not flourishing 
close to gently shelving beaches where the water is 
often turbid. 

' [The more vigorous growth of the coral on the outward part of a 
reef is, however, asserted and advocated by Mr. Murray as producing 
the atoll form. See Appendix II.] 


"What cause, then, has given to atolls and barrier- 
reefs their characteristic forms ? Let us see whether 
an important deduction does not follow from the 
following facts, — first, that reef-building corals only 
flourish at a very limited depth, — and secondly, that 
throughout areas of vast dimensions, none of the 
coral-reefs and coral-islets rise to a gi'eater height 
above the level of the sea than that attained by matter 
thrown uj) by the waves and winds. I do not make 
this latter statement vaguely ; I have carefully sought 
for descriptions of every island in the inter-tropical 
Beas ; and my task has been in some degree facili- 
tated by a map of the Pacific, corrected in 1834 by 
MM. D'Urville and Lottin, in which the low islands 
are distinguished from the high ones (even from those 
much less than a hundred feet in height) by being 
written without a capital letter.^ I have also ascer- 
tained, chiefly from the writings of Cook, Kotzebue, 
Bellingshausen, Duperrey, Beechey, and Lutke regard- 
ing the Pacific ; and from Moresby ^ with respect to the 
Indian Ocean, that in the following cases the term 
' low island ' strictly means land of the height com- 

' I have detected a few errors in this map, respecting the heights 
of some of the islands, which will be noticed in the Appendix, where 
I treat of coral-formations in geographical order. To the Appendix, 
also, I must refer for a more particular account of the data on which 
the following statements are grounded. 

^ See also Captain Owen's and Lieut. Wood's papers in the Geo- 
graphical Journal on the Maldiva and Laccadive Archipelagoes. 
These oliicers particularly refer to the lowness of the islets ; but I 
chiefly ground my assertion respecting these two groups, and the 
Chagos group, from information communicated to me by Captain 


monly attained by matter thrown up by the winds 
and the waves of an open sea. If we draw a Hne 
joining the external atolls of that part of the Low 
Archipelago in which the islands are numerous — the 
plan always adopted — the figure will be a pointed 
ellipse (reaching from Hood to Lazaref Island), of 
which the longer axis is 840 geographical miles, and 
the shorter 420 miles : in this space,^ none of the 
innumerable islets, united into great rings, rise above 
the stated level. The Gilbert group is very narrow, 
and 300 miles in length. In a prolonged line from 
this group, at the distance of 240 miles, is the 
Marshall Archipelago, the figure of which is an 
irregular square, one end being broader than the 
other ; its length is 520 miles with an average width 
of 240 : these two groups together are 1,040 miles in 
length, and all their islets are low. Between the 
southern end of the Gilbert and the northern end of 
Low Archipelago, the ocean is thinly strewed with 
islands, all of which, as far as I have been able to 
ascertain, are low : so that from nearly the southern 
end of the Low Archipelago, to the northern end of 
the Marshall Archipelago there is a narrow band of 
ocean more than 4,000 miles in length, containing a 

' Metia or Aurora Island has been upraised ; but it lies N.E. of 
Tahiti, and in the map appended to this volume is close without 
the line bounding the space here referred to. I shall have occasion 
hereafter to make some remarks on the supposed slight elevation (of 
about three feet) of the atolls of the Low Archipelago, subsequently 
to their original formation. [Other cases of upheaval have since been 
recorded. See Appendix II.] 


vast number of islands, all of which are low. In tlio 
western part of the Caroline Archipelago, there is a 
space of 480 miles in length, and about 100 in breadth, 
tliinlj'- interspersed with low islands. Lastly, in the 
Indian Ocean, the archipelago of the Maldivas is 470 
miles in length, and 60 in breadth ; that of the Lac- 
cadives is 150 by 100 miles : as there is a low island 
between these two groups, they may be considered as 
one group of a thousand miles in length. To this 
may be added the Chagos group of low islands, 
situated 280 miles distant, in a line prolonged from 
the southern extremity of the Maldivas. This group, 
mcluding the submerged banks, is 170 miles in length 
and 80 in breadth. So striking is the uniformity in 
direction of these three archij^elagoes, all the islands 
of which are low, that Captain Moresby, in one of his 
papers, speaks of them as parts of one great chain 
nearly 1,500 miles long. I am, then, fully justified 
in repeating that immense spaces, both in the 
Pacific and Indian Oceans, are interspersed with 
islands, of which none rise above the height to which 
the waves and winds in an open sea can heap up 

On what foundations, then, have these reefs and 
islets of coral been constructed ? A foundation must 
originally have been present beneath each atoll, at 
that limited depth which is indispensable for the 
first growth of the reef-building polypifers. A con- 
jecture will perhaps be hazarded, that the requisite 
bases may have been afforded by the accumulation of 


great banks of sediment, which did not quite reach 
the surface owing to the action of superficial currents, 
aided possibly by the undulatory movement of the 
sea. This appears actually to have been the case in 
some parts of the West Indian sea. But in the form 
and disposition of the groups of atolls, there is nothing 
to countenance this notion ; and the assumption that 
a number of immense piles of sediment have been 
heaped on the floor of the great Pacific and Indian 
Oceans in their central parts, far remote from land, 
where the dark blue colour of the limpid water 
bespeaks its purity, cannot for one moment be 

The many widely scattered atolls must, therefore, 
rest on rocky bases. But we cannot believe that a 
broad mountain summit lies buried at the depth of 
a few fathoms beneath every atoll, and nevertheless 
that throughout the immense areas above-named, not 
one point of rock projects above the level of the sea. 
For we may judge of mountams beneath the sea by 
those on the land; and where can we find a single 
chain, much less several such chains, many hundred 

' [This accumulation, it vdll be observed, is an integral part of Mr. 
Murray's hj^othesis. See Appendix II. for a sketch of the arguments 
by which it is supported. Perhaps I may be permitted to add that, 
in my opinion, the perusal of the observations of Mr. Murray, Mr. 
Guppy, and others would probably have led Mr. Darwin to modify 
slightly some of the clauses in these pages, and allow a more important 
rdle to the accumulation of organisms, other than corals, on submarine 
banks. I do not, however, anticipate that they would have seriously 
modified his general conclusions, or led him to regard modes of 
formation, which these authors consider to be normal, as other than 
exceptional.— T. G. B.] 



miles in length and of considerable breadth, with 
broad summits attaining the same height, from within 
120 to 180 feet ? Even if it be assumed without any 
evidence that the reef-building corals can flourish at a 
depth of 100 fathoms, yet the weight of the above argu- 
ment is but little diminished ; for it is almost equally im- 
probable, that as many submarine mountains, as there 
are low islands in the several great and widely-separated 
areas above-specified, should all rise within 600 feet of 
the surface of the sea and not one above it, as that they 
should be of the same height within the smaller hmit 
of one or two hundred feet. So highly improbable is 
this supposition, that we are compelled to believe, that 
the rocky foundations of the many atolls did never at 
any one period all lie submerged within the depth of a 
few fathoms beneath the surface, but that they were 
brought into the requisite position or level, some at one 
period and some at another, through movements in the 
earth's crust. But this could not have been effected 
by elevation ; for the belief that points so numerous and 
BO widely-separated were successively uplifted to a cer- 
tain level, but that not one point was raised above that 
level, is quite as improbable as the former supposition, 
and indeed differs little from it. It will probably occur 
to those who have read Ehrenberg's account of the reefs 
of the Red Sea, that many points in these great areas 
may have been elevated, but that as soon as raised, the 
protuberant parts were cut off by the destroying action 
of the waves : a moment's reflection, however, on the 
basin-like form of the atolls, will show that this ia 


impossible ; for the upheaval and subsequent abrasion 
of an island would leave a flat disk, which might become 
coated with coral, but not a deeply concave surface ; 
moreover, we should expect to see, at least in somfe 
parts, the rock of the foundation brought to the surface. 
If, then, the foundations of the many atolls were not 
uplifted into the requisite position, they must of neces- 
sity have subsided into it ; and this at once solves every 
difficulty,' for we may safely infer from the facts given 
in the last chapter, that during a gradual subsidence 
the corals would be favourably circumstanced for build- 
ing up their solid frameworks and reaching the surface, 
as island after island slowly disappeared. Thus areas 
of immense extent in the central and most profound 

' The additional difficulty on the crater hypothesis before alluded 
to, will now be evident ; for on this view the volcanic action must 
be supposed to have formed within the areas specified a vast number 
of craters, all rising within a few fathoms of the surface, and not one 
above it. The supposition that the craters were at different times 
upraised above the surface, and were there abraded by the surf and 
subsequently coated by corals, is subjected to nearly the same ob- 
jections with those given at the top of the page ; but I consider 
it superfluous to detail all the arguments opposed to such a notion. 
Chamisso's theory, from assuming the existence of so many banks, 
all lying at the proper depth beneath the water, is also vitally de- 
fective. The same observation applies to an hypothesis of Lieut. 
Nelson's (Geolog. Trans, vol. v. p. 122), who supposes that the ring- 
formed structure is caused by a greater number of germs of corals 
becoming attached to the declivity, than to the central plateau of a 
submarine bank ; it likewise applies to the notion formerly enter- 
tained (Forster's Observ. p. 151), that lagoon-islands owe their pecu- 
liar form to the instinctive tendencies of the polypifers. According 
to this latter view, the corals on the outer margin of the reef in- 
stinctively oppose themselves to the surf in order to afford protection 
to corals living in the lagoon which belong to other genera and to 
other families 1 


parts of the great oceans might become interspersed 
with coral-islets, none of which would rise to a greater 
height than that attained by detritus heaped up by the 
sea, and nevertheless they might all have been formed 
by corals, which absolutely require for their growth a 
solid foundation within a few fathoms of the surface. 

It would be out of place here to do more than 
allude to the many facts, showing that the supposition 
of a gradual subsidence over large areas is by no means 
improbable. We have the clearest proof that a move- 
ment of this kind is possible, in the upright trees 
buried under strata many thousand feet in thickness ; 
we have also every reason for believing that there are 
now large areas gradually sinking, in the same manner 
as others are rismg. And when we consider how many 
parts of the surface of the globe have been elevated 
within recent geological periods, we must admit that 
there have been subsidences on a corresponding scale, 
for otherwise the whole globe would have swollen. It 
is very remarkable that Sir C. Lyell,' even in the first 
edition of his Principles of Geology, mferred that the 
amount of subsidence in the Pacific must have exceeded 
tliat of elevation, from the area of land being very 
email relatively to the agents there tending to form it, 
namely, the growth of coral and volcanic action. But, 
although subsidence may explain a jphenomenon other- 
wise inexplicable, it may be asked, are there any direct 
proofs of a subsiding movement in these areas ? This, 

' Principles of Geology, sixth edition, vol. iii. p. 386. [Ch. xlix. 
ToJ. ii. p. C04, eleventh edition.] 


however, can hardly be expected, for it must ever be 
most difficult, excepting in countries long civilized, to 
detect a movement the tendency of which is to conceal 
the part affected. In barbarous and semi-civilized 
nations how long might not a slow movement, even ot 
elevation such as that now affecting Scandmavia, have 
escaped attention ! 

Mr. Williams^ insists strongly that the traditions of 
the natives, which he has taken much pains in collect- 
ing, do not indicate the appearance of any new islands : 
but on the theory of a gradual subsidence, all that would 
be apparent would be, the water sometimes encroaching 
slowly on the land, and the land again recovering by 
the accumulation of detritus its former extent, and 
perhaps sometimes the conversion of an atoll with coral 
islets on it, into a bare or into a sunken annular reef. 
Such changes would naturally take place at the periods 
when the sea rose above its usual limits during a gale 
of more than ordinary strength ; and the effects of the 
two causes would be hardly distinguishable. In Kotze- 
bue's Voyage there are accounts of islands, both in the 
Caroline and Marshall Archipelagoes, which have been 
partly washed away during hurricanes ; and Kadu, the 
native who was on board one of the Eussian vessels, 
said * he saw the sea at Eadack rise to the feet of the 
cocoa-nut trees ; but it was conjured in time.' ^ A storm 
lately entirely swept away two of the Carohne Islands 
and converted them into shoals ; it also partly destroyed 

' Williams's Narrative of Missionary Enterprise, p. 31. 
* Kotzebue's First Voyage, vol. iii. p. 168. 


two other islands.' According to a tradition which was 
communicated to Captain FitzPtoy, it is beHeved in the 
Low Archipelago that the arrival of the first ship caused 
a great inundation which destroyed many lives. Mr. 
Stutchbury relates that in 1825, the western side of 
Chain Atoll in the same group, was completely de- 
vastated by a hurricane, and not less than 300 lives 
lost : ' in this instance it was evident, even to the 
natives, that the hurricane alone was not sufficient to 
account for the violent agitation of the ocean.' "^ That 
considerable changes have taken place recently in some 
of the atolls in the Low Archipelago, appears certain 
from the case of Matilda Island given in the last chapter. 
With respect to Whitsunday and Gloucester Islands in 
this same grouj), we must either attribute great inac- 
curacy to their discoverer, the famous circumnavigator 
Wallis, or believe that they have undergone a consider- 
able change in the period of fifty-nine years between his 
voyage and that of Captain Beechey. Whitsunday 
Island is described by Wallis as ' about four miles long, 
and three wide,' now it is only one mile and a-half 
long. The appearance of Gloucester Island, in Captain 
Beechey's words,^ ' has been accurately described by its 
discoverer, but its present form and extent diff'er mate- 
rially.' Blenheim reef in the Chagos group, consists of a 
water- washed annular reef thirteen miles in circumfer- 
ence, surrounding a lagoon ten fathoms deep ; on its sur- 

' M. Desmoulins in Comptes Eendus, 1840, p. 837. 
* West of England Journal, No. 1, p 35. 

' Beechey's Voyage to the Pacific, chap, vii., and Wallis's Voyage 
In the Dolphin, chap, iv. 


face there are a few worn patches of conglomerate coral- 
rock of about the size of hovels ; and these Captain 
Moresby considers as being, without doubt, the last 
remnants of islets ; so that here an atoll has been con- 
verted into an atoll-formed reef. The inhabitants of 
the Maldiva Archipelago, as long ago as 1605, declared, 
* that the high tides and violent currents were always 
diminishing the number of the islands : ' • and I have 
already shown, on the authority of Captain Moresby, that 
the work of destruction is still in progress ; but that on 
the other hand the first formation of some islets is 
known to the present inhabitants. In such cases, it 
would be exceedingly difficult to detect a gradual sub- 
sidence of the foundation on which these mutable 
structures rest. 

Some of the archipelagoes of low coral-islands are 
subject to earthquakes : Captain Moresby informs me 
that they are frequent, though not very strong, in the 
Chagos group, which occupies a central position in the 
Indian Ocean, and is far from any land not of coral 
formation. One of the islands in this group was 
formerly covered by a bed of mould, which disap- 
peared after an earthquake, and was believed by the 
residents to have been washed by the rain into the 
underlying fractured rock : the island was thus ren- 
dered unproductive. Chamisso^ states that earth- 

' See an extract from Pyrard's Voyage in Captain Owen's paper 
on the Maldiva Archipelago, in the Geographical Journal, vol. ii, 
p. 84. 

^ See Chamisso, in Kotzebue's First Voyage, vol. iii. pp. 132 and 


quakes are felt in the Marshall atolls, which are far 
from any high land, and likewise in the islands of 
the Caroline Archipelago. On Oulleay atoll, one of 
the latter, Admiral Lutke informs me that he ob- 
served several straight fissures about a foot in width, 
running for some hundred yards obliquely across the 
whole width of the reef. Fissures indicate a stretching 
of the earth's crust, and, therefore, probably changes 
in its level ; but these coral-islands, which have been 
shaken and fissured, certainly have not been elevated, 
and, therefore, probably have subsided.' We shall 
hereafter see that the position of certain ancient build- 
ings in the Caroline Archipelago clearly indicates recent 
subsiden(*e. In the chapter on Keeling atoll. I have 
also attempted to show, by direct evidence, that the 
island subsided during the earthquakes lately felt there. 
The facts then stand as follows : — there are many 
large spaces of ocean, without any high land, inter- 
spersed with reefs and islets formed by the growth 
of those kinds of coral which cannot live at great 
depths; and the existence of these reefs and low 
islets in such numbers and at such distant points, 
is inexplicable, excepting on the theory that their 
rocky bases slowly and successively sank beneath the 
level of the sea, whilst the corals continued to gi-ow 
upwards. No positive facts are opposed to this view, 
and some direct evidence, as well as general considera- 
tions, render it probable. There is also evidence of 

• [It seems to me doubtful whether the argument from the ex« 
istence of fissures can be picsoed.— T. G. B.] 

Cn. V. 



change in form, whether or not from subsidence, on 
some of these coral-islands ; and there is evidence of 
subterranean disturbances beneath them. Will then 
the theory, to which we have thus been led, solve the 
curious problem — what has given to each class of reef 
its peculiar form ? 

Let us in imagination place within a subsiding area, 





Level of Sea- 

A A— Outer edge of the reef at the level of the sea. 

BB— Shores of the island. 

AA'— Outer edge of the reef, after its upward growth during a period 

of subsidence. 
C C— The lagoon -channel between the reef and the shores of the now 

encircled land. 
B' B' — The shores of the encircled island. 

N.B.— In this, and the following woodcut, the subsidence of the 
land could only be represented by an apparent rise in the level of 
the sea. 

an island surrounded by a ' fringing-reef '— that kind 
of which the origin alone offers no difficulty. Let the 
unbroken lines in the woodcut (No. 5) represent a 
vertical section through the land and water ; and the 
horizontal shading a section through the reef. Now, as 
the island sinks down, either a few feet at a time or quite 


insensibly, we may infer, from what we know of the 
conditions favourable to the growth of coral, that the 
living masses bathed by the surf on the margin of the 
reef, will soon regain the surface. The water, however, 
will encroach little by little on the shore, the island 
becoming lower and smaller, and the sj^ace between the 
edge of the reef and the beach proportionally broader. 
A section of the reef and island in this state, after a 
subsidence of several hundred feet, is given by the 
dotted lines : coral-islets are supposed to have been 
formed on the new reef, and a ship is anchored in 
the lagoon-channel. This section is in every respect 
that of an encircling barrier-reef, and is, in fact, 
taken E. and W. through the highest point of the 
encircled island of Bolabola,' of which a plan is given 
in Plate I., fig. 5. The same section is more clearly 
shown in the foUovving woodcut (No. 6) by the un- 
broken lines. The width of the reef and its slope 
both on the outer and inner side, will have been 
determined by the gi'owing powers of the coral, under 
different conditions, for instance, of the force of the 
breakers and currents to which it has been exposed ; 
and the lagoon-channel will be deeper or shallower, in 
proportion to the growth of the delicately branched 
corals within the reef, and to the accumulation of 
sediment ; relatively, also, to the rate of subsi- 

' The section has been made from the chart given in the Atlas 
of the Voyage of the Coquille. The scale is -57 of an inch to a mile. 
The height of the island, according to M. Lesson, is 4,020 feet. The 
deepest part of the lagoon-channel is 162 feet ; its depth is exag. 
gerated in the woodcut for the sake of clearness. 


dence and the length of the mtervening stationary 

It is evident in this section, that a Hne drawn per- 
pendicularly down from the outer edge of the new reef 
to the foundation of solid rock, exceeds, by as many 
feet as there have been feet of subsidence, that small 
limit of depth at which the effective polypifers can 
live,— the corals having grown up, as the whole sank 
down, from a basis formed of other corals and their con- 
solidated fragments. Thus the difficulty on this head, 
which before seemed so great, disappears. 

As the space between the reef and the subsiding 
shore continued to increase in breadth and depth, and 
as the injurious effects of the sediment and fresh water 
borne down from the land were consequently lessened, 
the greater number of the channels with which the reef 
in its fringmg state must have been breached, especially 
those which fronted the smaller streams, will have 
become choked up by the growth of coral : on the wind- 
ward side of the reef where the coral grows most 
vigorously, the breaches will probably have first been 
closed. In barrier-reefs, therefore, the breaches kept 
open by draining the tidal waters of the lagoon-channel, 
will generally be placed on the leeward side, and they 
will still face the mouths of the larger streams, although 
removed beyond the influence of their sediment and 
fresh water; — and this, it has been shown, is commonly 
the case. 

Eeferrmg to the following diagram (No. 6), in 
which the newly-formed barrier-reef is represented by 


unbroken lines, instead of by clots as in the former wood- 
cut, let the work of subsidence go on, and the doubly- 
pointed hill will form two small islands included within 
one annular reef. Let the island continue to subside, 
and the coral-reef will continue growing up on its own 
foundation, whilst the water gains inch by inch on the 
land, until the last and highest pinnacle is covered, and 
there remains a perfect atoll. A vertical section of this 

A'A' — Outer edges of the barrier-reef at the level of the sea. The 
cocoa-nut trees represent coral-islets formed on the reef. 

C C — The lagoon-channel. 

B'B' — The shores of the island, generally formed of low alluvial land 
and of coral detritus from the lagoon-channel. 

A"A" — The outer edges of the reef, now forming an atoll. 

C — The lagoon of the newly-formed atoll. According to the scale 
the depth of the lagoon and of the lagoon-channel is exaggerated. 

atoll is shown in the woodcut by the dotted lines ; — a 
ship is anchored in its lagoon, but islets are not supposed 
yet to have been formed on the reef. The depth of the 
lagoon and the width and slope of the reef, will depend 
on the different circumstances to which it has been 
exposed, as just stated with respect to barrier -reefs. 
Any further subsidence will produce no change in the 


atoll, except a diminution in its size, from the reef not 
growing vertically upwards. I may here observe, that 
a bank either of rock or of hardened sediment, level 
with the surface of the sea and fringed with living 
coral, would be immediately converted by subsidence 
into an atoll, without passing, as in the case of a reef 
fringing the shore of an island, through the inter- 
mediate form of a barrier-reef. As before remarked, if 
such a bank lay a few fathoms submerged, the simple 
growth of the coral, without the aid of subsidence, would 
produce a structure scarcely to be distinguished from a 
true atoll; for the corals on the outer margin, from being 
freely exposed to the open sea, would grow vigorously 
and tend to form a contmuous ring, whilst the growth 
of the less massive kinds on the central expanse, would 
be checked by the sediment formed there, and by that 
washed inwards by the breakers ; and as the space be- 
came shallower, their growth would also be checked by 
the impurities of the water, and probably by the small 
amount of food brought to them by the enfeebled cur- 
rents. The subsidence of a reef based on a bank of this 
kind, would give depth to the central expanse or lagoon, 
steepness to the flanks, and through the free growth of 
the coral, symmetry to the whole outline ; but, as we 
have seen, the larger groups of atolls in the Pacific and 
Indian Oceans cannot have been formed on banks of 
this nature. 

If, instead of an island, as in the diagram, the shore 
of a continent fringed by a reef were to subside, a great 
barrier-reef like that on the N.E. coast of Australia, 


would be the necessary result ; and it would be sepa- 
rated from the main land by a deep-water channel, 
broad in proportion to the amount of subsidence, and 
to the less or greater inclination of the bed of the sea. 
The efifcct of the continued subsidence of a barrier- 
reef, and its probable conversion into a chain of 
separate atolls, will be considered when we discuss the 
progressive disseverment of the larger Maldiva atolls. 
We now are able to perceive that the close similarity 
in form, dimensions, structure, and relative position 
between fringmg and encircling barrier-reefs, and be- 
tween these latter reefs and atoUs, is the necessary result 
of the transformation, during subsidence, of the one 
class into the other. On this view, the three classes of 
reefs ought to graduate into each other. Eeefs having 
an intermediate character between those of the fring- 
ing and barrier classes do exist ; for instance, on the 
S.W. coast of Madagascar, a reef extends for several 
miles, within which there is a broad channel from 
7 to 8 fathoms deep, but the sea does not deepen 
abruptly outside the reef. Such cases, however, are 
open to doubt, for an old fringing-reef which had 
extended itself on a basis of its own formation, would 
hardly be distinguishable from a barrier-reef produced 
by a small amount of subsidence, and with its lagoon- 
channel nearly filled up with sediment during a long 
stationary period. Between barrier-reefs, encircling 
either a single lofty island or several small low ones, 
and atolls including a mere expanse of water, a striking 
Beries can be shown : and in proof of this, I need only 


refer to Plate I., whicli speaks more plainly to the eye, 
than any description to the ear. The authorities from 
which the figures have been copied, together with some 
remarks on them, are given on a separate page descrip- 
tive of the plate. At New Caledonia (Plate II., fig. 5) 
the barrier-reefs extend for 150 miles on each side of 
the submarine prolongation of the island ; and at the 
northern extremity these reefs appear broken up and 
converted into a vast atoll supporting a few low coral- 
islets. We may imagine that we see in New Caledonia 
the effects of subsidence actually in progress, — the 
water always encroaching on the northern end of the 
island, towards which the mountains slope down, and 
the reefs steadily building up their massive fabrics in 
the line of their ancient growth. 

We have as yet only considered barrier-reefs and 
atolls in their simplest form ; but there remain some 
peculiarities in structure and some special cases, which 
were described in the two first chapters, to be accounted 
tor by our theory. These consist, firstly, in the presence 
of an inclined ledge terminated by a wall, and some- 
times succeeded by a second ledge with a wall, round 
the shores of certain lagoons and lagoon-channels ; for 
this structure cannot be explained by the mere groAvth 
of the corals; — secondly, in the ring or basin-like form 
of the central and circumferential reefs of the northern 
Maldiva atolls, — thirdly, in the disseverment of some 
of the Maldiva atolls,— fourthly, in the existence of 
irregularly formed atolls, some tied together by linear 
reefs, and others with spurs projecting from them, — 


fifthly, in the submerged condition of the whole, or of 
parts of certain barrier and atoll-formed reefs, and 
in the submerged parts being generally to leeward, — 
and, lastly, in the structure and origin of the Great 
Chagos bank. 

Step-formed ledges round certain lagoons. — If we 
suppose an atoll to subside at an extremely slow rate, 
the living corals would grow up on the outer margin 
and on the deeper parts of the bare and hard surface 
of the annular reef. Detritus would soon accumulate 
and become agglomerated on this surface, after a time 
forming islets. Consequently the whole atoll before long 
would recover its former structure and appearance. If, 
however, an atoll were to subside somewhat suddenly 
to the depth of a few fathoms, the whole annular reef, 
consisting of solid rock, would constitute an excellent 
basis for the attachment and subsequent upward growth 
of a great bed of living corals. But the corals would 
not be able to grow up from the sandy shores of the old 
lagoon ^ ; consequently the new annular reef would be 
separated from the new lagoon by an abrupt edge or 
wall. As the corals would grow upwards much more 
vigorously on the outer side, and more detritus would 
be accumulated there, the surface of the new annular 
reef would slope gently inwards. Hence the summit 
of the new annular reef on the inner side would pro- 
bably never rise above the level of the wew lagoon, 
and in this case would be covered with sand. If now 

' [This would seem possible in certain cases, though obviously the 
position would be an unfavourable one. See Appendix II.] 


a second and somewhat sudden subsidence were to 
occur, the same results would follow as during the 
previous and similar subsiding movement. Conse- 
quently the new lagoon would be surrounded by two 
inwardly sloping ledges, which once existed as tho 
summits of two successive annular reefs, both termi- 
nated on the inner side by vertical w^alls or cliffs.' 

The ring or hasin-formed reefs of the northern 
Maldiva atolls. — I must first observe that small reefa 
within large lagoons or within broad lagoon-channels, 
would grow up during subsidence; and therefore such 
reefs would sometimes be found rising abruptly from 
a greater depth than that at which the ef&cient poly- 
pifers can flourish. We see this well exemplified in 
the small abruptly-sided reefs with which the deep 
lagoons of the Chagos and Southern Maldiva atolls are 
studded. With respect to the ring or basin-formed reefs 
of the Northern Maldiva atolls (see Plate II., fig. 4), 
it is evident from the perfectly continuous series which 
exists, that the rings on the margin, although broader 
than the exterior or bounding reef of an ordinary 
atoll, are only modified portions of such a reef ; it is 

' According to Mr. Couthouy (p. 26) the external slope round 
many atolls descends by a succession of ledges or terraces. He 
attempts, but I doubt whether successfully, to explain this structure 
somewhat in the same manner as I have attempted, with respect to 
the internal ledges round the lagoons of certain atolls. More facta 
are wanted regarding the nature both of the interior and exterior 
stop-like ledges. Are all the ledges, or only the upper ones, covered 
with living coral ? If they are all so covered, do the species differ 
on the different ledges ? Do ledges occur on the inside and outside 
round the same atolls ? &c. [Some further information on this sub- 
ject has been obtained by recent investigations. See Appendix IL] 



also evident that the central rings, although broader 
than the knolls or reefs which commonly occur within 
lagoons, occupy the same relative position. The ring- 
like structure has been shown to be contingent on 
the breaches into the lagoon being wide and numerous, 
thus causing the inner side of the marginal reef and 
the central reefs to be placed under nearly the same 
conditions with the outside of an ordinary atoll which 
is exposed to the open sea. Hence the margins of these 
reefs have been favourably circumstanced for growing 
outwards and increasing beyond their usual breadth ; 
and the conditions have likewise been favourable for 
their growing vigorously upwards, during that subsi- 
ding movement to which by our theory the whole 
archipelago has been subjected ; and subsidence toge- 
ther with the upward growth of the margin would 
convert the central space of each little reef into a 
small lagoon. This, however, could only take place 
with reefs which had increased in breadth sufficiently 
to prevent their central spaces from being almost im- 
mediately filled up with the sand and detritus driven 
inwards by the waves from all sides. We can thus 
understand how it is that few reefs less than half 
a mile in diameter, even in the atolls where perfect 
ring-formed reefs are found, include lagoons. This 
remark, I may add, applies to aU coral- formations. 
The basin-formed reefs of the Maldiva Archipelago 
may, in fact, be briefly described as small atolls 
formed during subsidence over separate portions of a 
large and broken atoll, in the same manner as the 


latter was originally formed over a reef encircling one 
or more mountainous islands. 

The disseverment of the larger Maldiva atolls. — 
The ai)parent progressive disseverment of large atolls 
into smaller ones in the Maldiva Archipelago, demands 
an explanation. The graduated series which marks, 
as I believe, this process, can be observed only in the 
northern half of the group, where the atolls have im- 
perfect margins consisting of detached basin-formed 
reefs. The currents of the sea flow across these atolls 
with considerable force, as I am informed by Captain 
Moresby, and drift the sediment from side to side 
during the monsoons, transporting much of it seaward ; 
yet the currents sweep with greater force round their 
flanks. It is historically known that these atolls have 
long existed in their present state ; it is intelligible 
that they might thus remain, even during a slow sub- 
sidence, owing to the continued growth of the corals, 
and to the lagoon being kept at nearly its original depth 
by the accumulation of sediment. But during the 
action of such nicely balanced forces, it would be strange 
if the currents of the sea had never made a direct 
passage across some of these atolls, through the many 
wide breaches in their margins. As soon as this oc- 
curred the channels would be deepened by the removal 
of the finer sediment, and by the check to its further 
accumulation. The sides also of the channels would 
soon be worn into a slope like that on the outer coasts, 
from being exposed to the same force of the currents. 
In fact, a channel like that bifurcating one which 


divides Mahlos Mahdoo (Plate II., fig. 4) would almost 
necessarily be formed. The scattered reefs situated 
near the borders of the new channel, from being favour- 
ably placed for the growth of coral, would, by their 
extension, tend to produce fresh margins to the dis- 
severed portions : and a tendency of this kind is evident 
in the elongated reefs which border the two channels 
intersecting Mahlos Mahdoo. Such channels would 
become deeper with continued subsidence, and, from 
the reefs on both sides not growing up perpendicularly, 
somewhat broader. In this case, and more especially 
if the channels had been originally formed of consider- 
able breadth, the dissevered portions would soon be- 
come perfect and distinct atolls like Ari and Eoss atolls 
(Plate IL, fig. 6), or like the two Nillandoo atolls, which 
must be considered as distinct, although plainly related 
to each other in form and position, and separated only 
by moderately deep channels. Further subsidence 
would render such channels unfathomable, and the 
dissevered portions would then resemble Phaleedoo and 
Moluque atolls, or Mahlos Mahdoo and Horsburgh atolls 
(Plate II., fig. 4), which are related to each other only 
in proximity and position. Hence, on the theory of 
subsidence, the disseverment of large atolls which are 
exposed to strong currents and which have imperfect 
margins (for otherwise their disseverment would bo 
scarcely possible) is far from being an improbable 
event ; and the several stages, from a close connection 
to the entire isolation of some of the atolls in tlio 
Maldiva Archipelago, are readily explicable. 


It is even probable that the Maldiva Archipelago 
originally existed as a barrier-reef of nearly the same 
dimensions as that of New Caledonia (Plate II. fig. 5) : 
for if we complete in imagination the subsidence of 
this great island, we may infer from the broken condi- 
tion of the northern portion of the reef, and from the 
almost entire absence of reefs on the eastern coast, that 
the present barrier, after repeated subsidences, would 
become, during its subsequent upward growth, separated 
into distinct portions ; and these portions would tend 
to assume an atoll-like structure, owing to the corals 
growing with vigour where freely exposed to the open 
sea. As some large islands have subsided to a certain 
amount and are partly encircled by barrier-reefs, so our 
theory makes it probable that there should be other 
large islands wholly submerged ; and these, as we can 
now see, would be surmounted, not by one enormous 
atoll, but by several large ones like the atolls of the 
Maldiva group ; and these again, during long periods 
of subsidence, would sometimes become dissevered into 
smaller ones. In the Marshall and Caroline Archipela- 
goes, there are atolls standing close together which 
have an evident relationship in form; and we may 
; suppose that either two or more encircled islands ori- 
ginally stood close together and afforded bases for two 
or more atolls, or that one large atoll has been dis- 
Bevered. But from the position as well as the forms of 
three atolls in the Caroline Archipelago (the Namourrek 
and Elato groups), which are placed in an irregular 
circle, I am strongly inclined to believe that they owe 


their origin to the clisscverment of a single large 

Irregularly -formed Atolls. — In the Marshall group, 
Musquillo atoll consists of tf\'0 loops united by a single 
point ; and Menchicoff atoll is formed of three loops, 
two of which (as may be seen in fig. 3, Plate 11.) are 
connected by a mere ribbon-shaped reef; the three 
together being GO miles in length. In the Gilbert group 
some of the atolls have narrow reefs like spurs, pro- 
jecting from them. Linear and straight, or crescent- 
formed reel's with their extremities more or less curled 
inwards, may sometimes be found standing by them- 
selves in the open ocean. All these irregular forms 
would naturally follow from continued subsidence, 
combined with the upward growth of reefs fronting 
one side alone of a high island, the reefs on the op- 
posite side having perished or never having existed. 

Submerged and Dead Reefs. — In the second section 
of the first chapter, I have shown that there some- 
times exist in the neighbourhood of atolls, deeply 
submerged banks with level surfaces ; that there are 
others, less deeply but yet wholly submerged, having 
all the characters of a perfect atoll, but consisting 

' The same remark is, perhaps, applicable to the islands of Ollap, 
Fanadik, and Tamatam in the Caroline Archipelago, of which charts 
are given in the atlas of Duperrey's voyage ; a line drawn through 
the linear reefs and lagoons of these three islands form a semicircle. 
Consult also the atlas of Lulk6's voyage ; and for the Marshall group 
that of Kotzebue ; for the Gilbert group (which is refcrred to in the 
ensuing paragraph) consult the atlas of Duperrey's voyage. Most of 
the points here referred to may, however, be seen in Krusenstern'a 
general Atlas of the Pacific. 


merely of dead coral-rock ; that there are barrier-reefs 
and atolls with only a portion of the reef, generally on 
the leeward side, submerged; and that such portions 
eitlier retain their perfect outline, or appear to be 
more or less completely effaced, their former place 
being marked only by a bank, conforming in general 
outline with that part of the reef which remains 
j)erfect. These several cases are, I believe, intimately 
related, and can all be explained by the same agency 
of subsidence. 

We see that in those parts of the ocean where 
coral-reefs are most abundant, one island is fringed 
and another neighbouring one is not fringed, and that 
in the same archipelago, all the reefs are more perfect 
in one part than in another, — for instance, in the 
southern compared with the northern half of the 
Maldiva Archipelago, and likewise on the outer as 
compared with the inner coasts of the double row of 
atolls in this same archipelago. The existence of the 
innumerable polypifers forming a reef depends on 
their finding sustenance, and we know that they are 
preyed on by other organic beings, and that some 
inorganic causes are highly injurious to their growth. 
Can it, therefore, be expected that the reef-building 
polypifers should keep alive for perpetuity in any one 
place, during the round of change to which earth, 
air, and water are subjected ; and still less can this 
be expected during progressive subsidence, to which 
by our theory these reefs and islands have been liable ? 
Should such subsidence be at any time greater than 


the rate of upward growth of thepolypifers, the death 
of tlie reef must ensue, and it would have been strange 
had wo found no evidence of this. It is, then, not 
at all improbable that the corals should sometimes 
perish either on the whole or on part of a reef. 
If only on a part, the dead portion, after a small 
amount of subsidence, would still retain its proper 
outline and position beneath the water. After a more 
prolonged subsidence, it would form, owing to the 
accumulation of sediment, a more or less level bank 
marking the limits of the former lagoon. Such dead 
portions of a reef would generally lie on the leeward 
side,^ for the impure water and fine sediment are 
driven out from the lagoon over this side of the reef, 
where the force of the breakers is less than to wind- 
ward, and where the corals are, in consequence, less 
vigorous and less able to resist any destroying agency. 
It is owing to this same cause tha't reefs are fre- 
quently breached to leeward by channels which serve 

• Sir C. Lyell, in the first edition of his Principles of Geology, 
offered a somewhat different explanation of this structure. He snp- 
poses that there has been subsidence ; but he was not aware that the 
submerged portions of reef were in most cases, if not in all, dead ; 
and he attributes the difference in height in the two sides of most 
atolls chiefly to the greater accumulation of detritus to windward 
than to leeward. But as matter is accumulated only on the back- 
ward part of the reef, the front part would remain of the sama 
height on both sides. I may here observe that in most cases (for 
instance at Peros Banhos, the Gambler group and the Great Chag03 
bank), and I suspect in all cases, the dead and submerged portions 
do not blend or slope into the living and perfect parts, but are sepa- 
rated from thcra by an abrupt line. In some instances small patcbcs 
of living reef rise to the suiface from the middle of the submerged 
and dead parts. 


as ship-channels. If the corals perished entirely, or 
on the greater part of the circumference of an atoll, 
the result would be an atoll-shaped bank of dead 
rock more or less entirely submerged ; and further 
subsidence, together with the accumulation of sedi- 
ment, would obliterate its atoll-like structure, and 
leave only a bank with a nearly level surface. 

We meet with all these cases in the Chagos group 
of atolls. Here within an area of 160 miles by 60, 
there are two atoll-formed banks of dead rock (besides 
another very imperfect one) entirely submerged ; a 
third bank with merely two or three small pieces of 
living reef which rise to the surface ; and a fourth, 
namely, Peros Banhos (Plate I. fig. 9), with a por- 
tion nine miles in length dead and submerged. As 
by our theory this area has subsided, and as there is 
nothing improbable in the death of the corals on por- 
tions or over the whole surface of a reef, either from 
changes in the state of the surrounding sea or from the 
subsidence being great or sudden, these Chagos banks 
present no difficulty. So far, indeed, are any of the above- 
mentioned cases of dead submerged reefs from offering 
any difficulty, that their occurrence might have been 
anticipated on our theory ; and as fresh atolls are sup- 
loosed to be in progressive formation by the sub- 
sidence of encircling barrier-reefs, a weighty ob- 
jection might even have been raised, namely that 
atolls must increase indefinitely in number, unless 
proofs of their occasional destruction could have been 


The Great Charjos Bank.^—l have already shown 
that the submerged condition of the Great Chagos 
bank (Plate II. fig. 1, with its section, fig. 2), and of 
some other banks in the Chagos group, may in all pro- 
bability be attributed to the corals having perished 
during an unusually rapid or sudden subsidence. The 
external rim or upper ledge (shaded in the chart) con- 
sists of dead coral-rock thinly covered with sand ; it 
lies at an average depth of between 5 and 8 fathoms, 
and perfectly resembles in form the annular reef of an 
atoll. The banks of the second level, the boundaries 
of which are marked by dotted lines in the chart, lie 
from about 15 to 20 fathoms beneath the surface; 
they are several miles in breadth, and terminate m 
a very steep slope round the central expanse. This 
central expanse consists of a level muddy flat between 
30 and 40 fathoms deep. The banks of the second 
level appear at first sight to resemble the internal step- 
like ledges of dead coral-rock which border the lagoons 
of certain atolls, but their much greater width, and 
their being formed of sand, are points of essential dif- 
ference. On the eastern side of the atoll some of the 
banks are linear and parallel, like islets in a great river, 
and they point dh-ectly towards a great breach on the 
opposite side of the atoll : these are best seen in the 
large published chart. I inferred from this circum- 
stance, that strong currents sometimes set dh-ectly 
across this great bank ; and I hear from Captain Moresby 
that this is the case. I observed, also, that the channels, 

' [See Appendix II.] 


or breaches through the rim, were all of the same depth 
as the central expanse into which they lead ; whereas 
the channels into the other atolls of the Chagos group, 
and as I believe into most other large atolls, are not 
nearly as deep as the lagoons. For instance at Peros 
Banhos, the channels as well as the bottom of the lagoon 
for a space about a mile and a-half round its shores, are 
only between 10 and 20 fathoms in depth, whilst the 
central expanse is from 35 to 40 fathoms deep. Now, 
if an atoll during a gradual subsidence once became 
entirely submerged like the Great Chagos bank, and 
therefore no longer exposed to the surf, very little 
sediment could any longer be formed from it ; conse- 
quently the channels leading into the lagoon would be 
no longer filled up with drifted sand and coral detritus, 
and would continue increasing in depth, as the whole 
sank down. In this case we might expect that the 
currents of the open sea, instead of sweeping as at 
first round the submarine flanks, would, as the many 
breaches in the reef increased, flow directly across 
the lagoon, thus removuig the finer sediment from 
the channels, and preventing its further accumulation. 
The submerged reef would thus ultimately consist of an 
upper and narrow broken rim of rock, surrounded on 
the inner side by banks, the remnants of the sandy 
bed of the old lagoon, now intersected by many deep 
channels ; these channels, with their sides worn steep by 
the oceanic currents, uniting in the centre and forming 
the central deep expanse. By such means the Great 


Chagos bank — the most anomalous structure which 1 
have met M-ith— appears to have originated. 

If this bank should continue to subside, a mere 
wreck of an atoll would be left; for the corals are 
almost everywhere dead. Pitt's bank, situated not far 
southward, appears to be in this actual condition : it 
consists of a moderately level, oblong bank of sand, 
lying from 10 to 20 fathoms beneath the surface, with 
two sides protected by a narrow ledge of rock submerged 
between 5 and 8 fathoms. A Httle to the south of this 
ledge, at about the same distance as the southern rim of 
the Great Chagos bank lies from the northern rim, there 
are two other small banks with from 10 to 20 fathoms 
on them ; and not far eastward, soundings were struck 
on a sandy bottom with between 110 and 145 fathoms. 
The northern portion of Pitt's bank with its ledge-like 
margin, thus closely resembles any one segment of the 
Great Chagos bank between two of the deep-water 
channels, and the scattered banks southward and east- 
ward appear to be the last wreck of the less perfect 
portions of one great and now ruined atoll. 

I have examined with care the charts of the Indian 
and Pacific Oceans, and have now laid before the reader 
all the cases which I have met with, of reefs differing 
from the class to which they belong ; and I think it has 
been shown that they are all included in our theory, 
modified by occasional accidents, such jis might have 
been anticipated. We have thus seen, that in the lapse 
of ages encu-cling barrier-reefs are converted into atolls, 
—the term atoll being applicable as soon as the last 


pinnacle of encircled land sinks beneath the surface of 
the sea. We have seen that large atolls, during the pro- 
gressive subsidence of the areas in which they stand, 
sometimes become dissevered into smaller ones. At other 
times, when the reef-building polypifers perish, atolls 
are converted into atoll- formed banks of dead rock ; and 
these again, through further subsidence and the accu- 
mulation of sediment, pass into level banks with scarcely 
any distinguishing character. Thus may the history of an 
atoll be followed from its birth, through the occasional ac- 
cidents of its existence, to its death and final obliteration. 
Objections to our theory of the formation of Atolls 
and Barrier-reefs. — The vast amount of subsidence 
both in area and depth, necessary to have submerged 
every mountain, even the highest, throughout the 
immense spaces of ocean now interspersed with atolls, 
will probably strike most persons as a formidable objec- 
tion to the theory. But as continents, as large as the 
spaces supi^osed to have subsided, have been raised 
above the level of the sea, — as whole regions are now 
rismg, for instance, in Scandmavia and South America, 
— and as no reason can be assigned why subsidence 
should not have occurred in some parts of the earth's 
crust on as great a scale as elevation, this objection has 
little force. The remarkable point is, that a subsiding 
movement to such an extent and amount should have 
taken place within a period, during which the corals 
have contmued to add matter to the same reefs. An- 
other and less obvious objection to the theory may 
perhaps be advanced, namely, that, although atolls and 


barrier-reefs are supposed to have gone on subsiding 
for a long period, yet that their lagoons and lagoon- 
channels have only rarely come to exceed 40 and 
never 60 fathoms in depth. But if our theory is worth 
consideration, we already admit that the rate of sub- 
sidence has not ordinarily exceeded that of the upward 
growth of the massive corals which live on the margins 
of the rcofs, so that we have only further to suppose 
that the rate has never exceeded that at which lagoons 
and lagoon-channels are filled up by the growth of the 
delicate corals which hve there, and by the accumula- 
tion of sediment. As the fillmg-up process, in the case 
of barrier-reefs lying far from the land, and of the larger 
atolls, must be an extremely slow one, we are led to 
conclude that the subsiding movement has always been 
equally slow. And this conclusion accords well with 
what is known of the rate of recent movements of 

It has, I think, been shown in this chapter, that 
subsidence explains both the normal structure and 
the less regular forms of those two great classes of 
reefs which have justly excited the astonishment of all 
the naturaHsts who have sailed through the Pacific and 
Indian Oceans. The necessity, also, that a fomidation 
should have existed at the proper depth for the growth 
of the corals over certain large areas, almost compela 
us to accept this theory. But further to test its truth, 
a crowd of questions may be asked. Do the different 
kinds of reefs which have been produced by the 
Bame kind of movement, generally lie within the 


Rame or closely adjoining areas ? How are such reefs 
related to each other in form and position, — for 
instance, do neighbouring groups of atolls, and the 
separate atolls in each group, hear the same relation 
to each other as do ordinary islands ? Although coral- 
reefs \Yhich have just begun to re-grow, after having 
been killed by too rapid a subsidence, would at first 
belong to the fringing class, yet, as a general rule, 
reefs of this class indicate that the land has either 
long remained at a stationary level, or has been up- 
raised. Of a stationary level it is hardly possible 
to find any evidence except of a negative kind ; but 
of recent elevation, upraised marine remains afford 
a sure proof : it may therefore be asked, do fringed 
coasts often afford such evidence ? Do the areas 
which have subsided, as shown by the presence of 
atolls and barrier-reefs, and the areas which have 
either remained stationary or have been upraised, 
as indicated by fringing-reefs, bear any determinate 
relation to each other ? Is there any relation between 
the areas of recent subsidence or elevation, and the 
presence of active volcanic vents? These several 
questions will be considered in the following chapter.' 

' I may take this opportunity of briefly considering the appear- 
ance which would probably be presented by a vertical and deep 
section across a coral formation (referring chiefly to an atoll) formed 
by the upward growth of coral during successive subsidences. This 
is a subject worthy of attention, as a means of comparison with 
ancient coral strata. The circumferential parts would consist of 
massive species in a vertical position, with their interstices tilled up 
with detritus ; but this would be the part most subject to subsequent 
denudation and removal. It is useless to speculate how large a 


proportion of the exterior annular reef would consist of upright coral, 
and how much of fragmentary rock, for this would depend on many 
contingencies,— such as on the rate of subsidence occasionally allow- 
ing a fresh growth of coral to cover the whole surface, and on the 
breakers having force suHicient to throw fragments over this samo 
space. The conglomerate which composes the base of the islets, 
would (if not removed by denudation together with the exterior reef 
on wliich it rests) be conspicuous from the size of the fragments, — 
the dillercnt degrees in which they have been rounded, — the presence 
of ragments of conglomerate torn up rounded and re-cemented, — 
and from the oblique stratification. The corals which lived in the 
lagoon-reefs at each successive level, would be preserved upright, 
and they would consist of many kinds, generally much branched. 
In this part, however, a very large proportion of the rock, and in 
some cases nearly all of it, would be formed of sedimentary matter, 
being in an excessively fine or moderately coarse state, with the par- 
ticles almost blended together. The conglomerate which was formed 
of rounded pieces of the branched corals on the shores of the lagoon, 
would differ from that formed on the islets and derived from the 
outer coast; although both might have been accumulated very near 
each other. The stratification, taken as' a whole, would be hori- 
zontal : but the conglomerate beds resting on the exterior reef, and 
the beds of sandstone on the shores of the lagoon and on the ex- 
ternal flanks of the reef, would probably be divided (as at Keeling 
atoll and at Mauritius) by numerous layers dipping at considerable 
angles in different directions. The calcareous sandstone and coral 
rock would almost necessarily contain innumerable shells, echini, 
and the bones of fish, turtle, and perhaps of birds : possibly, also, 
the bones of small saurians, as these animals find their way to 
islands far remote from any continent. The large shells of some 
species of Tridacna would be found vertically imbedded in the solid 
rock, in the position in which they lived. We might expect, also, 
to find a mixture of the remains of pelagic and littoral animals in 
the strata formed in the lagoon, for pumice and the seeds of plants 
are floated from distant countries into the lagoons of many atolls ; 
on the outer coast of Keeling atoll near the mouth of the lagoon, the 
shell of a pelagic Pteropodous animal was brought up on the arming 
of the sounding-lead. All the loose blocks of coral on Keeling atoll 
were burrowed by vermiform animals ; and as every cavity, no doubt, 
ultimately becomes filled with spathose limestone, slabs of the rock 
would, if polished, probably exhibit the excavations of such burrow- 
ing animals. The conglomerate and fine-grained beds of coral-rock 
would be hard, sonorous, white, and composed of nearly pure cal 


careous matter ; in some few parts, judging from the specimens at 
Keeling atoll, they would probably contain a small quantity of iron. 
I have seen a conglomerate now forming on the shores of the Mal- 
diva atolls, resembling conglomerate limestone from Devonshire. 
Floating pumice and scorire, and occasionally stones transported in 
the roots of trees (see my Naturalist's Voyage, p. 461) appear the 
only sources through which foreign matter is brought to coral-for- 
mations standing in the open ocean. The area over which sediment 
is transported from coral-reefs must be considerable; Captain 
Moresby informs me that during the change of monsoons, the sea 
is discoloured to a considerable distance off the Maldiva and Chagos 
atolls. The sediment off fringing and barrier coral-reefs must bo 
mingled with the mud which is brought down from the land, and is 
transported seaward through the breaches which occur in front of 
almost every valley. If the bed of the ocean were to be upraised 
and converted into land, the atolls of the larger archipelagoes would 
form flat-topped mountains, varying in diameter from a few to sixty 
miles — for the smallest atolls would probably be worn quite away ; 
and from being horizontally stratified and of similar composition, 
they would, as Sir C. Lyell has remarked, falsely appear as if they 
had originally been united into one vast continuous mass. Such 
great strata of coral-rock would rarely be associated with erupted 
volcanic matter, for this could only take place, as may be inferred 
from what follows in the next chapter, when the area in which they 
were situated, commenced to rise, or at least ceased to subside. 
During the enormous period necessary to effect an elevation of tho 
kind just alluded to, the surface would necessarily be greatly de- 
nuded ; hence it is highly improbable that any f ringing-reef, or even 
any barrier-reef, at least those encircling small islands, would be 
preserved to a distant period. From this same cause, the strata 
which were formed within the lagoons of atolls and the lagoon- 
channels of barrier-reefs, and which must consist in a large part of 
sedimentary matter, would more often be preserved to future ages, 
than the exterior solid reef composed of massive corals in an upright 
position ; although it is on this exterior part that the present exist- 
ence and further growth of atolls and barrier-reefs depend. 




Description of the coloured map — Proximity of atolls and barrier' 
reefs — Relation inform and position of atolls with ordinary islands 
— Direct evidence of subsidence difficult to b,i detected — Proofs of 
recent elevation where fringing -reefs occur — Oscillations of level 
— Absence of active volcanoes in the areas of subsidence — Immen- 
sity of the areas which have been elevated and have subsided — 
Their relation to the present distribution of the la-nd — Areas of 
subsidetice elongated — Their intersection and alternation with 
tliose of elevation — Amount, and slow rate of the req_uisite sub- 
sidence — Recapitulation. 

It will be convenient first to give a sliort account of 
the appended map of the Pacific and Indian Oceans 
(Plate III.) ; a fuller one, with the data for colouring 
each spot, is reserved for the Appendix, and every 
place there referred to may be found in the Index. 
A larger chart would have been desirable ; but, small 
as the adjoined one is, it is the result of maiiy months* 
labour. I have consulted, as far as I was able, every 
original voyage and map ; and the colours were first 
laid down on charts on a large scale. The same blue 
colour, with merely a difference in the tint, is used 
for atolls or lagoon-islands, and for barrier-reefs ; 


these being in all essential respects closely related. 
Fringing-reefs, on the other hand, have been coloured 
dull red, for there is an important distinction between 
them and barrier-reefs and atolls with respect to the 
depth beneath the surface, at which, as we must believe, 
their foundations lie. The two distinct colours, there- 
fore, mark two great types of structure. 

The dark blue colour represents atolls and sub- 
merged annular reefs with deep water in their centres. 
I have coloured a few low and small coral-islands 
as if they had been atolls, although not including a 
lagoon ; but tliis has been done only when it clearly 
appeared that they had originally contained one. 
When no such evidence exists they have been left 

The pale blue colour represents barrier-reefs. The 
most obvious character of reefs of this class is the 
broad and deep-water moat within the reef ; but this, 
like the lagoon of a small atoll, is liable to become 
filled up with detritus and with reefs of delicately- 
branched corals. When, therefore, a reef round the 
entire circumference of an island extends far into a 
profoundly deep sea, so that it can hardly be con- 
founded with a fringing-reef which must rest on a 
foundation of rock within a small depth, it has been 
coloured pale blue, although it does not now include 
a deep-water moat. But this has been rarely done, 
and each case is distinctly mentioned in the Appendix. 

The red colour represents reefs which fringe the 
land closely where the sea is deep, and extend to 


a moderate distance from it where the bottom ia 
j;ently inclined ; but they never include a deep-water 
moat or lagoon-like channel running parallel to Lb a 
shore. It must, however, be remembered that 
fringing-reefs are frequently breached by deep-water 
channels, where mud has been deposited in front of 
rivers and valleys. 

In all cases, a space of 30 miles in width has 
been coloured round or in front of the reefs of each 
class, in order that the colours might be made con- 
spicuous in a map on so small a scale. 

The vermilion spots and streaks represent vol- 
canos now in action, or historically known to have 
been so. They are laid down chiefly from Von Buch's 
work on the Canary Islands ; and my reasons for 
making a few alterations are given in the note below.' 

• I have also made considerable use of the geological part of 
Berghaus' Physical Atlas. Beginning at the eastern side of the 
Pacific, I have added to the number of the volcanos in the southern 
part of the Cordillera, and have coloured Juan Fernandez according 
to observations collected during the voyage of the Beagle (Geol. 
Trans, vol. v. p. 601). I have added a volcano to Albemarle Island, 
one of the Galapagos Archipelago (see my Journal of Researches, 
p. 457). In the Sandwich group there are no active volcanos, except 
at Hawaii; but the Rev. W. Ellis informs me there are streams of 
lava apparently modern on Maui, having a very recent appearance, 
which can be traced to the craters whence they flowed. The same 
gentleman informs me that there is no reason to believe that any 
active volcano exists in the Society Archipelago ; nor are there any 
kno^vTi in the Samoa or Navigator group, although some of the 
streams of lava and craters there appear recent. In the Friendly 
group, the Rev. J. Williams says (Narrative of Missionary Enter- 
prise, p. 20) tliat Toofoa and Proby Islands are active volcanos. I 
infer from Hamilton's Voyage in the Pandora (p. 9")), that Proby 
Island is synonymous with Onouafou, but I have not ventured to 


The uncoloured ixirts consist, first and chiefly, of 
coasts where no coral-reefs, or quite insignificant ones, 

colour it. There can be no doubt respecting Toof oa ; and Captain 
Edwards (Von Buch, p. 386) found the lava of a recent eruption 
at Amargura still smoking. Berghaus marks four active volcanos 
actually within the Friendly group ; but I do not know on what 
authority; I may mention that Maurelle describes Latte as having 
a burnt-up appearance ; I have marked only Toofoa and Armagura. 
SouLh of the New Hebrides lies Matthews Rock, which is described 
as an active crater in the voyage of the Astrolabe. Between it and 
the volcano on the eastern side of New Zealand lies Brimstone 
Island, which from the high temperature of the water in the crater 
may be ranked as active (Berghaus Vorbemerk, II. Lief. S. 56). 
Malte Brun, vol. xii. p. 231, says that there is a volcano near Port 
St. Vincent, in New Caledonia : I believe this to be an error, arising 
from smoke seen on the opposite coast by Cook (2nd voyage, vol. ii. 
p. 23), which smoke went out at night. The Mariana Islands, 
especially the northern ones, contain many craters (see Freycinet's 
Hydrog. Descript.) which are not active. Von Buch, however, states 
(p. 462), on the authority of La Peyrouse, that there are no less than 
seven volcanos between these islands and Japan. Gemelli Careri 
(Churchill's Collect, vol. iv. p. 45S) says there are two active volcanos 
in lat. 23° 30' and in lat. 24° ; but I have not coloured them. From 
the statements in Beechey's Voyage (p. 518, 4to edit.) I have coloured 
one in the northern part of the Bonin group. M. S. Julien has 
clearly made out from Chinese manuscripts not very ancient (Comptes 
Eendus, 1840, p. 832), that there are two active volcanos on the 
eastern side of Formosa. In the map appended to the first edition 
I marked an active volcano in Torres Straits, and gave my authority ; 
but Mr. Jukes informs me that there certainly is no volcano there ; 
a wooded island on fire having been mistaken for one. Mr. M'Clel- 
land (Report of Committee for Investigating Coal in India, p. 39) 
has shown that the volcanic band which passes through Barren 
Island musk be extended northwards. It appears by an old chart, 
that Cheduba was once an active volcano (see also Silliman's North 
American Journal, vol. xxsviii. p. 385). In Berghaus' Phys. Atlas, 
1840 (No. 7 of Geological Part) a volcano on the coast of Pondi- 
cherry is said to have burst forth in 1757. Ordinaire (Hist. Nat. 
dcs Volcans, p. 218) says that there is one at the mouth of the 
Persian Gulf, but I have not coloured it, as he gives no particulars. 
A volcano in Amsterdam, or St. Paul's, in the southern part of the 


exist. Secondly, of coasts where the sea is extremely 
shallow ; and the reefs in this case generally lie far 
from the land, and are very irregular, so that they 
cannot always be classed. Thirdly, reefs which appear 
merely to coat submerged banks of rock or of sedi- 
ment ; for such reefs differ in some essential respects 
from those which owe their whole thickness to the 
growth of corals. Fourthly, in the Eed Sea, and 
within some parts of the East Indian Archipelago 
(if the imperfect charts of the latter can be trusted), 
there are many scattered reefs of small size, repre- 
sented by mere dots, which rise out of deep water ; 
and these have likewise been left uncoloured. In the 
Eed Sea, however, some such reefs seem once to have 
formed parts of a continuous barrier. There exist, 
also, scattered in the open ocean, some linear and 
irregularly-formed reefs which are jDrobably, as shown 
in the last chapter, remnants of atoUs ; but as they 
cannot safely be placed in this class, they have not 
been coloured ; they are, however, few in number, and 
of insignificant dimensions. Lastly, some reefs have 
been left uncoloured from the want of sufficient in- 
formation ; and some because they are intermediate 
in character between barrier and fringing-reefs. The 

Indian Ocean, has been seen (Naut. Mag. 1838, p. 842) in action. 
Dr. J. Allan, of Forres, informs me in a letter that, when he was at 
Joanna, he saw flames at night, apparently volcanic, issuing from 
the Chief Comoro Island, and that the Arabs assured him that they 
v.'ere volcanic, adding 'that the volcano burnt more during the wet 
season : I have marked this as a volcano, though with some hesita- 
tion, as the flames may have arisen from gaseous sources. 


value of the map is lessened, in proportion to the 
nnmber^ of reefs which I have thus been obliged to 
leave uncolourcd ; but their number is not very great, 
as will be seen by comparing the map with the state- 
ments in the Appendix. I have experienced more 
difficulty in colouring fringing-reefs than in colouring 
barrier-reefs, as the former, from their small size, 
have not much attracted the attention of navigators. 
As I have had to seek my information from all kinds 
of sources, I do not venture to hope that the map is 
free from errors. Nevertheless, I trust it will give 
an approximately correct view of the general distri- 
bution of the coral-reefs throughout the world, (with 
the exception of some fringing-reefs on the coast of 
Brazil, not included within the limxits of the map,) and 
of their arrangement into the three great classes which, 
though necessarily ill-defined from the nature of the 
objects classified, have been adopted by most voyagers. 
I may further remark, that the dark-blue colour repre- 
sents land entirely composed of coral-rock ; the pale 
blue, land with a wide and thick border of coral-rock ; 
and the red, land with a mere narrow fringe of coral-rock. 
Looking now at the map under a theoretical point 
of view, the two blue tints signify that the foundations 
of the reefs thus coloured have largely subsided, and 
that the rate of subsidence has been less than the up- 
w^ard growth of the corals. It is also probable that in 
many cases the foundations are still subsiding. The red 
signifies that the shores thus coloured support fringing- 
reefs ; and they have not, as a general rule, recently 


subsided, at least to any considerable amount, for the 
effects of subsidence on a small scale would hardly be 
distinguishable. Such shores must either have remained 
stationary since the period when they were first fringed ; 
or they may have been repeatedly upraised, with new 
lines of reefs successively formed round them. If, how- 
ever, coral-reefs became attached for the first time to a 
shore which was subsiding, or if a barrier-reef was de- 
stroyed and submerged with a new reef re-attached 
to the shore, this would necessarily belong at first 
to the fringing class, and would be coloured red, 
although the land was sinking. So it would be with 
a subsiding shore, if it plunged at a very high angle 
beneath the sea, for in this case the reef would remain 
closely attached to the land as it grew upwards, and 
would resemble in all respects a fringing-reef. This 
source of doubt applies especially to atolls which 
h.ave been upraised (such as Metia and Ehzabeth 
Islands), for from the steepness of their sub-marine 
flanks, a reef growing up during a subsequent 
period of subsidence round them, would still continue 
closely to skirt the land, and would therefore be 
coloured red. Well-characterised atolls or encircling 
reefs, where several occur together in a group, or a 
smgle barrier-reef if of large dimensions, clearly indicate 
a movement of subsidence. The evidence from a single 
atoll, or from a single encircling-reef, must be received 
with caution, for the former may be based upon a sub- 
merged crater or bank, and the latter on a submerged 
margin of sediment or of worn-down rock. 


On the distributio7i of the different classes of reejs. 
' — Having made the foregoing preliminary remarks, I 
■will now consider how far the distribution of the dif- 
ferent kinds of coral-islands and reefs corroborates our 
theory. A glance at the map shows that the reefs 
which are coloured blue and red, and which are believed 
to owe their origin either to widely different move- 
ments, or in the case of the red to a stationary condition, 
are not indiscriminately mingled together. Atolls and 
barrier-reefs, as may be seen by the two blue tints, 
generally lie near each other ; and this would be the 
natural result of both having been produced by the same 
movement of subsidence. Thus, all the Society Islands 
are encircled by barrier-reefs ; and to the N.W. and 
S.E. there are several scattered atolls. To the eastward 
lies the great Paumotu or Low Archipelago consisting 
entirely of atolls ; and still further to the N.E., we meet 
with the Mendana or Marquesas Islands, which, from 
their abrupt and deeply indented shores, Dana ^ be- 
lieves have probably subsided ; though hardly any coral- 
reefs exist there, which might have afforded additional 
evidence of subsidence. In the midst of the Caroline 
atolls, there are three fine encircled islands. The 
northern point of the barrier-reef of New Caledonia 
apparently forms, as before remarked, a great atoll. 
The Australian barrier is described as including both 
atolls and small encircled islands. Captain King ^ 

■ ' Corals and Coral Islands, 1872, p. 325. 
^ Sailing Directions, appended to vol. ii. of his Surveying Voyage 
to Australia. 


mentions many atoll-formed and encircling coral-reefs, 
some of which Ho within the barrier, and others may be 
said (for instance, between lat. 16° and 13°) to form 
joart of it. Flinders ' has described an atoll-formed 
reef in lat. 10°, seven miles long and from one to three 
broad, resembling a boot in shape, and apparently in- 
cluding a deep lagoon. Eight miles westward of this, 
and forming part of the barrier, lie the Murray Islands, 
which are high and are encircled. In the Corallian sea, 
between the two great barrier-reefs of Australia and 
New Caledonia, there are many low islets and coral- 
reefs, some of which are annular, or like a horse- shoe. 
Bearing in mind the smallness of the scale of our map 
(the lines of latitude being 900 miles apart), we see that 
none of the larger groups of reefs and islands which are 
coloured blue, and which are supposed to have been 
produced by long-continued subsidence, lie near exten- 
sive lines of coast coloured red ; these latter having 
either long remained stationary, or having been upraised 
with new reefs re-formed on them. Where red and blue 
circles do occur near each other, I am able, in several 
instances, to show that there have been oscillations of 
level ; subsidence having preceded the elevation of the 
red spots ; and elevation having preceded the subsidence 
of the blue spots ; and in this case the juxtaposition of 
reefs belonging to the two great types of structure is 
little surprising. We find, therefore, that atolls and 
barrier-reefs, which both owe their origin to subsidence, 
lie near together and are as a general rule separated 

' Voyage to Tena Australis, vol. ii. p. 336. 


from fringing-reefs, which show that the land is sta- 
tionary or rising ; and all this holds good to the full 
extent which might have been anticipated by our 

As atolls have been formed during the sinking of 
the land by the upward growth of the reefs which 
primarily fringed the shores of ordinary islands ; so we 
might expect that these rings of coral, hke so many 
rude outline charts, would still retain traces of the 
general form, or at least of the general range, of the 
islands round which they were first modeUed. That this 
is the case with the atolls in the Southern Pacific, as 
far as their range is concerned, seems highly probable, 
when we observe that the principal groups are directed 
in nearly N.W. and S.E. lines, and that nearly all the 
mountainous islands and shores in the S. Pacific range 
in this same direction ; namely, N. -Eastern Australia, 
New Caledonia, the northern half of New Zealand, the 
New Hebrides, Saloman, Navigator, Society, Marquesas, 
and Austral Archipelagoes. In the Northern Pacific, 
the Caroline atolls almost abut against the N.W. Ime of 
the Marshall atolls, much in the same manner as the 
E. andW. line of islands extending from Ceramto New 
Britain abuts against New Ireland. In the Indian 
Ocean the Laccadive and Maldiva atolls extend nearly 
parallel to the western mountains of India. There is 
also a close resemblance between atohs and ordmary 
islands in the manner in which they are grouped, as 
well as in their shapes. Thus the outline of all the 
larger groups of atolls is elongated; and the atolla 


themselves are generally elongated in the same direc- 
tion with the group. The Chagos group is less elon- 
gated than is usual, and the individual atolls in it are 
likewise but little elongated ; this is strikingly seen by 
comparing them with the neighbouring Maldiva atolls. 
In the Marshall and Maldiva archipelagoes, the atolls 
are ranged in two parallel lines, like a great double 
mountain-chain. Some of the atolls in the larger archi- 
pelagoes stand so near to each other, and have such 
an evident relationship, that they compose little sub- 
groups ; in the Caroline Archipelago, one such sub-group 
consists of Pouynipete, a lofty island encircled by a 
barrier-reef, and separated by a channel only four miles 
and a half in width from Andeema atoll, with a second 
atoll a little further removed. 

On the direct evidence of the blue spaces in the map 
having subsided during the upivard growth of the reefs 
thus coloured, and of the red spaces having remained 
stationary, or having been upraised. — With respect to 
subsidence, we cannot expect to obtain in semi-civil- 
ised countries proofs of a movement which tends to 
conceal its own evidence. But on coral-islands we see 
plain signs of a round of decay and renovation — on 
some, the last vestiges of land — its first commence- 
ment on others : we hear of storms washing away and 
desolating the islets to an extent which astonished the 
inhabitants ; we know by the great fissures with which 
some of these islands are traversed, and by the earth- 
quakes felt under others, that subterranean disturbances 
are in progress. All these appearances accord well with 

Ch. vt. distkibution of coral-reefs. 1C3 

the belief that these islands have recently subsided ; 
though not proving the fact. At Keehng atoll, however, 
I have described certain appearances, which seem 
directly to show that the surface subsided there during 
the late earthquakes. In the Caroline Archipelago, 
the island of Poujaiipete (Plate I. fig. 7), from being en- 
circled by a great barrier-reef, must have subsided, in 
accordance with our theory; and in the New South 
Wales Lit. Advert. Feb. 1835, there is an account of 
this island, (subsequently confirmed by Mr. Campbell,) 
in which it is said, ' At the N.E. end, at a place called 
Tamen, there are ruins of a town, now only accessible 
by boats, the waves reaching to the steps of the houses.' 
Hence it would appear that the island must have sub- 
sided since these houses were built. Mr. Hales also states, 
from information acquired during tlie U.S. Exploring 
Expedition, that certain buildings on this island are 
now in the water: 'what were once paths are now 
passages for canoes, and when the walls are broken down 
the water enters the enclosures.' ' Yanikoro, according 
to the Chevalier Dillon, is often violently shaken by 
earthquakes, and there, the unusual depth of the channel 
between the shore and the reef, the wall-like structure 
on the inner side of the reef, the smah quantity of low 
alluvial land at the foot of the mountains, and the almost 
entire absence of islets on the reef, all seem to show 
that this island has not remained long at its present 
level.2 At the Society Archipelago, on the other hand, 

' Professor Dana also concludes from these facts that the island 
is subsiding ; see Corals and Coral Islands, 1872, p. 330. 

2 See Captain Dillon's Voyage in search of La Peyrouse. M. 


where a slight tremor is only rarely felt, the shoalness 
of the lagoon-channels round some of the islands, the 
number of islets formed on the reefs of others, and the 
Lroad belt of low land at the foot of the mountains, all 
indicate that those islands have not undergone for a 
long period, any movement of subsidence, although 
their encircling reefs must on our theory have been 
originally produced through sulxsidonce.' 

Although Dana admits that atolls and barrier-reefs 
must have been originally formed by the subsidence of 
their foundations, he believes that a large number of 
atolls, situated between the Paumotu or Low group to the 
east and the Feejees to the west, and northward nearly 
as far as the equator, have recently been uplifted to 
the height of a very few feet."-^ Mr. Couthouy came to a 
similar conclusion during the same expedition with re- 
spect to. many of the Paumotu atolls. These observers 
ground their belief chiefly from having found the great 
shells of the Tridacna vertically embedded in coral- 
rock, at a height at which they cannot now exist. Mr. 
Couthouy also states that he found corals standing on 

Cordier, in his Report on the Voyage of the Astrolabe (vol. i. p. cxi.), 
speaking of Vanikoro, says the shores are surrounded by reefs of 
madrepore, ' qu'on assure Hre dc formation tout-d-fait viodciiie.' 

' Mr. Couthouy states (J'emarks, p. 44) that at Tahiti and Einieo 
the space between the reef and the shore has been nearly filled up 
by the extension of coral-reefs of the kind which within most barrier- 
reefs merely fringe the land. From this circumstance, he arrives at 
the same conclusion as I have done, namely, that the Society Islands 
have remained stationary during a long period. 

« Corals and Coral Islands, 1872, pp. 199, 345. See also Mr. 
Coufhouy's Remarks on Coral Formations. [See Wilkes' Explormg 
Expedition, vol. i. chap, xv.] 


the shores and in the midst of the lagoons, from 12 to 
30 inches ahove the sea-level, with the tips of their 
branches dead. He also refers to masses of coroA-rocls 
which he thinks could not have been carried into their 
present ijositions and subsequently been water-worn, 
whilst the land stood at its present level. Nevertheless 
it might, I think, have been anticipated that many 
atolls would have presented the above appearance, if 
they had long remained at a stationary level. The sea, 
after the land had at some former period subsided a few 
feet, would have continued for a long time brciiking 
over the whole reef, even after the living corals had 
grown up to their full height on the outer margin. The 
waters of the lagoon would thus have been disturbed and 
raised, so that shells and corals, from being bathed by 
the troubled waters, could have existed at a greater 
height than that at which they could exist after the 
reef had been raised by the agglutination of fragments 
and sand, and after islets had been formed on its sur- 
face. Even the mere outward growth of a reef, and the 
consequent increase of its breadth, by checking the 
inward rush of the breakers, would tend to lower the 
level in the lagoon at which corals and shells can live. 
We have seen that at the Keeling Islands there are 
fields of rotten coral with the tips of their branches pro- 
jecting above the surface of the lagoon, — the result of 
the tides not rising so high as formerly (as is said to be 
the case by the inhabitants), from the closing of the 
channels between the islets on the outer reef, and from 
the lagoon being partially choked up by the growth 


of the corals. Here, so far from there liaving been 
any recent elevation of the land, we have reason to be- 
lieve that there has been subsidence. Messrs. Dana 
and Couthouy's observations relate chiefly to the 
Paumotu atolls, and here again some facts indicate 
recent subsidence rather than elevation : I refer to the 
manner in which Chain atoll suffered during a storm, 
and to Sir E. Belcher's statement,' that after an interval 
of fourteen years, a well-known islet had disappeared, 
and the lagoon at a particular spot had become deeper 
than it was before. 

There are other causes of change which might, as it 
appears to me, easily lead to a mistaken belief in the 
recent elevation of low coral formations. We must re- 
member that the outer and living margin of the reef 
grows up to a height determined by the constant break- 
ing of the waves. Outside this margin there is a sloping 
surface also covered with living corals, but belonging to 
species which do not grow to the surface ; and beyond 
this, there is a much steeper slope, consisting of coral- 
sand. Now after a somewhat rapid subsidence of, for 
instance, one or two fathoms, w-e may feel almost sure 
that the corals on the outer margin would grow up 
quickly to the surface and form a nearly vertical wall. 
This would be succeeded outside by a steeply sloping 
surface of Hving corals, which would likewise sooner or 
later grow up to their former level ; but outside this, 
the much steeper slope, formed by the slow accumula- 
tion of fine detritus, would not recover for a very long 
» Voyage Round the World, vol. i. 1843, p. 382. 


time its former angle relatively to the upper bank of 
living corals. Now it seems highly probable that a 
change of any kind in the outer submarine slope of 
an island would influence the height to which the living 
corals on the margin would be constantly bathed by the 
surf, and to which they would consequently be able to 
grow. Again, it seems possible that if during one season 
of the year the currents of the sea and the prevalent 
winds coincided in direction, the waves would then reach 
to a higher level and the corals grow higher, than at 
another season when the currents and the winds did not 
coincide in direction. The result would be that the 
corals which during the one season had grown to their 
full height, would at the other season expose their dead 
summits, and give the appearance of the land having 
been slightly elevated. I have referred to these possi- 
bilities merely to show how difficult it must ever be to 
judge whether low coral formations have really been 
raised to a height of only two or three feet, as Dana 
believes to have been the case with several groups of 
atolls. To me it seems more probable that all the above- 
mentioned a|:)pearances merely indicate that the atolls 
in question have long remamed at the same level. If, 
however, the conclusion arrived at by so excellent an 
observer as Professor Dana, should hereafter be con- 
firmed, the question will arise, seeing how immense an 
area has been thus affected, whether those geologists 
are not right who believe that the level of the ocean 
is subject to secular changes from astronomical 



Evidence that many coasts fringed with coral-reefs 
and coloured red on the map, have been recently ele~ 
rated. — As the areas which have slowly subsided with- 
in the period of existing corals are many and large, 
we might have expected that such movements would 
have been counterbalanced by the recent elevation of 
other equally large areas ; and this, as we shall see, 
apparently holds good. Corals attached to a rising 
coast would necessarily form a fringing-rccf ; and this 
reef would be upraised at each successive elevation, with 
a new one formed on the coast at a lower level. Such 
reefs would differ only by their smaller breadth from 
those attached to a shore which had long remained 
stationary ; for they would not have had sufficient time 
to form a foundation of their own detritus and grow far 
outwards. Fringing-reefs indicate as a general rule 
that the land to which they are attached has not re- 
cently subsided. But they do not tell us whether the 
land is rising or stationary. Nevertheless, the crust 
of the earth seems liable to such incessant changes of 
level that a long-continued stationary condition ap- 
parently is rare. We may infer that this is so from 
the number of cases, within the limits of our map, in 
which upraised corals or other organic remams have 
been found on the shores which are fringed with reefs, 
and are, therefore, coloured red. It may be mentioned 
as bearing on this subject, that I was much surprised 
on first reading a memoir on coral formations by 
MM. Quoy and Gaimard,* by finding that their de- 
' Annales des Sciences Nat, torn. vi. p. 279, Ac. 


Bcriptions applied only to reefs of the fringing class, for 
I knew that they had crossed both the Pacific and Indian 
Oceans ; but my surprise ended in satisfaction, when I 
discovered that all the islands which they had visited, 
though several in number — namely, Mauritius, Timor, 
New Guinea, the Mariana and Sandwich Archipelagoes 
— could be shown by their own statements to have 
been elevated within a recent geological period. 

I will now enter on some details, showing how many of 
the islands and coasts which from being fringed with reefs 
are coloured red on our map, have been recently upraised. 

Sandwich Islands. — Several of these islands are frniged 
with reefs, though Dana found very few corals at Hawaii ; 
and almost every naturalist who has visited them has there 
observed upraised corals and shells, apparently identical 
with living species. The Eev. W. Ellis informs me that 
he noticed round several parts of Hawaii, beds of coral 
detritus, about twenty feet above the level of the sea, and 
where the coast is low they extend far inland. Upraised 
coral-rock forms a considerable part of the borders of Oahu ; 
and at Elizabeth Island ' it composes three strata, each 
about ten feet thick. Nihau, which forms the northern, as 
Hawaii does tbe southern end of the group (350 miles in 
length), likewise seems to consist of coral and volcanic 
rocks. Mr. Couthouy ^ has lately described several upraised 
beaches and ancient reefs with their surfaces perfectly pre- 
served, as well as beds of recent shells and corals, at the 
Islands of Maui, Morokai, Oahu, and Tauai (or Kauai), all 
in this group. Mr. Pierce, an intelligent resident at Oahu, 
is convinced, from changes which have taken place within 
bis memory during the last sixteen years, ' that the eleva- 

' Zoology of Captain Beechey's Voyage, p. 176. See also MM. 
Quoy and Gaimard in Annales des Sciences Nat. torn. vi. 
- Eemarlcs on Coral Formations, p. 51. 


tion is at present going forward at a very perceptible rate.' 
The natives at Kauai state that the land is there gaining 
rapidly on the sea ; and Mr. Couthouy has no doubt, from 
the natui'e of the strata, that this is the result of elevation. 

Elizabeth Island, in the southern part of the Low or 
Paumotu Archipelago, and Mctia in the northern part, 
consist of upraised coral-rock, closely fringed by living 
reefs. ^ In cases like these, where islands have the appear- 
ance which one of the smaller surrounding atolls with a 
shallow lagoon would present if elevated, we are led to con- 
clude that the elevation has taken place at an epoch not 
geologically remote ; for it is improbable that such small 
and low fabrics should have resisted for an immense period 
all the many destroying agents of nature. When the sur- 
face of an ordinary island is strewed with marine remains, 
from the beach to a certain height, and not above that 
height, it is exceedingly improbable that these remains, 
although they may not have been specifically examined, 
should belong to any very ancient period. It is necessary 
to bear these remarks in mind in considering the evidence 
of the elevatory movements in the Pacific and Indian 
Oceans, as it does not often rest on specific determinations, 
and therefore should be received with caution. Six of the 
Cook and Austral Islands (S.W. of the Society group) are 
fringed ; of these, five were described to me by the Piev. J. 
"Williams, as formed of coral-rock (associated with some 
basalt in ]\Iangaia), and the sixth as lofty and basaltic. 
Mangaia is nearly 300 feet high with a level summit ; and, 
according to Mr. S. Wilson,^ is an upraised reef; 'and 
there are in the central hollow, formerly the bed of the 
lagoon, many scattered patches of coral-rock, some of them 
raised to a height of forty feet.' These knolls of coral-rock 

' Beechey's Voyage in the Pacific, p. 46, 4to edit. Dana, Corals 
and Coral Islands, p. 193. "Wilkes, U.S. Exploring Expedition, vol. i. 
p. 337. 

- Couthouy's Eemarka, p. 34. 


were evidently once reefs within the lagoon of an atoll. 
Mr. Martens, at Sydney, informed me that this island is 
surrounded by a terrace-like plain at about the height of 
100 feet, which probably marks a pause in its elevation. 
From these facts we may infer that the Cook and Austral 
Islands have been upheaved at a not very remote period. 

Savage Island (S.E. of the Friendly group) is according 
to Forster about 40 feet in height, and according to 
Williams about 100 feet. Forster ' describes the plants as 
already growing out of the dead but still upright and 
spreadmg trees of coral ; and the younger Forster ^ believes 
that an ancient lagoon is now represented by a central 
plain : here we cannot doubt that the elevatory forces have 
recently acted. The same conclusion may be extended to 
the islands of the Friendly Group, which have been well 
described in the second and third voyages of Cook, and 
recently by Dana. The surface of Tongatabou is low and 
level, but with parts 50 or 60 feet high ; the whole consists 
of coral-rock, * which yet shows the cavities and irregular- 
ities worn mto it by the action of the tides.' ^ On Eoua 
the same appearances were noticed at an elevation of 
between 200 and 300 feet. Vavao, also, at the opposite or 
northern end of the group, consists, according to the Eev. 
J. Williams, of coral-rock. Tongatabou, with its northern 
extensive reefs, resembles either an upraised atoll with one 
half originally imperfect, or one unequably elevated ; and 
Anamouka, an atoll equably elevated. This latter island 
contains* in its centre a salt-water lake, about a mile and a 
half in diameter, without any communication with the sea, 
and around it the land rises gradually like a bank : the 
highest part is only between twenty and thirty feet; but 

* Observations made during Voyage Round the World, p. 147. 
' Voyage, vol. ii. p. 163. 

• Cook's Third Voyage (4to edit.), vol. i. p. 314. 
« Ibid. vol. i. p. 235. 


on this part, as well as on the rest of the land, (which, as 
Cook observes, rises above the height of a true lagoon- 
island,) coral-rock like that on the beach was found. In the 
Navigator or Samoan Archipelago, Mr. Couthouy ' found 
on Manua many large fragments of coral at the height of 
eighty feet, ' on a steep hill-side, rising half a mile inland 
from a low sandy plain abounding in marine remains.' 
The fragments were embedded in a mixture of decomposed 
lava and sand. It is not stated whether they were accom- 
panied by shells, or whether the corals resembled recent 
species ; as these remains were embedded, they possibly 
may belong to a remote epoch ; but I presume this was 
not the opinion of Mr. Couthouy. On the other hand, !Mr. 
Dana says expressly in one place, that ' no satisfactory 
evidences of elevation were detected about these islands ; ' 
and in another place he says (p. 326) that some of the 
islands have probably subsided. Earthquakes are very 
frequent in this archipelago. 

Still proceeding westward we come to the New Hebrides. 
On these islands, Mr. G. Bennett (author of Wanderings 
in New South Wales) informs me that he found much coral 
at a great altitude, which he considered of recent origin.^ 
The Loyalty Islands are situated west of the New Hebrides, 
and not far from New Caledonia ; and one of these islands 
has been clearly shown by the Eev. W. B. Clarke (Journal 
of Geolog. Soc. 1847, p. 61) to consist wholly of coral-rock, 
and to have been raised within a recent period by at least 
two distinct elevations to the height of 250 feet. The 
shores are now fringed by reefs. Respecting Santa Cruz 
and the Saloman Archipelago ^ I have no information ; but 
at New Ireland, which forms the northern point of the 

' Remarks on Coral Formations, p. 50. 

' [Prof. Moseley, Notes of a Naturalist in the Challenger, speaks ol 
indications of elevation to an extent of about 5 feet.] 
' [See Mr. Guppy's description, Appendix II.] 


latter chain, both Labillardiere and Lesson have described 
large beds of an ajjparently very modern madreporitic rock, 
with the form of the corals little altered. The latter 
author ^ states that this formation composes a newer line 
of coast, modelled romid an ancient one. There only re- 
mains to be described in the Pacific, that curved line of 
fringed islands, of which the Marianas form the main part. 
Of these Guam, Kota, Tinian, Saypan, and some islets 
farther north, are described by Quoy and Gaimard,^ and 
Chamisso,^ as chiefly composed of madreporitic limestone, 
which attains a considerable elevation, and is in several 
cases worn into successively rising cliffs : the two former 
naturalists seem to have compared the corals and shells 
with the existing ones, and state that they are of recent 
species. Peel Island, one of the Benin or Arzobispo group, 
between the Marianas and Japan, has fringing-reefs ; and 
it has clearly been upraised to a height of at least 50 feet, 
as shown by the ridges of corals and shells extending 
uniformly at this level.^ Fais, which hes in the prolonged 
line of the Marianas, between this group and the Pellews, 
is fringed by reefs ; it is 90 feet high, and consists entirely 
of madreporitic rock.-'' 

In the East Indian Archipelago, many authors have 
recorded proofs of recent elevation. M. Lesson" states that 
near Port Dory, on the north coast of New Guinea, the 
shores are flanked, to the height of loO feet, by madre- 
poritic strata of a modern date. He mentions similar for- 
mations at Waigiou, Amboina, Bourou, Ceram, Sonda, and 
Timor: at this latter place, MM. Quoy and Gaimard^ have 

' Voyage de la Coc[uille, Part. Zoolog. 

2 Freycinet's Voyage aiitour du Monde. See also the Hydro- 
graphical Memoir, p. 215. ^ Kotzebue's First Voyage. 
^ P. W. Graves, Journal of Geological Soc. 1855, p. 532. 

* LutkS's Voyage, vol. ii. p. 304. 

« Partio Zoolog. Voyage de la Coquille. 

* Ann. des Scien. Nat., torn. vi. p. 281. 


likewise described tlie primitive rocks, as coated to a con- 
siderable height with coral. Some small islets eastward of 
Timor are said in Kolff's Voyage ' to resemble small coral 
islets upraised some feet above the sea. Dr. Malcolmson 
informs me that Dr. Hardie found in Java an extensive 
formation, containing an abmidance of shells, of which the 
greater part appear to be of existing species. Dr. Jack ^ 
has described some upraised shells and corals, apparently 
recent, on Pulo Nias off Sumatra ; and ]\Iarsden relates in 
his history of this great island, that the names of many 
pi'omontories show that they were originally islands. On 
part of the west coast of Borneo and at the Sooloo Islands, 
the form of the land, the nature of the soil, and the water- 
washed rocks, present appearances ^ (although it is doubt- 
ful whether such vague evidence is worthy of mention) of 
having recently been covered by the sea ; and the inhabi- 
tants of the Sooloo Islands believe that this has been the 
case. Mr. Cuming, who has lately investigated with so 
much success the mollusca of the Phillippines, found near 
Cabagan, in Luzon, about 50 feet above the level of the 
E. Cagayan and 70 miles from its mouth, a large bed of 
fossil shells : these, as he informs me, are certainly of the 
same species with those now existing on the shores of the 
neighbouring islands. From the accounts given by Captam 

' Translated by Windsor Earl, chaps, vi. and vii. 

- Geolog. Transact. '2nd series, vol. i. p. 403. On the Peninsula 
of jMalacca, in front of Penang, 5° 30' N., Dr. Ward collected some 
shells which Dr. Malcolmson informs me, although not compared 
with existing species, had a recent appearance. Dr. Ward describes 
in this neighbourhood (Trans. Asiat. Soc. vol. xviii., part 2, p. 166) a 
single water-worn rock, with a conglomerate of sea-shells at its base, 
eituated six miles inland, which, according to the traditions of the 
natives, was once surrounded by the sea. Captain Low has also 
described (ibid. Part i. p. 131) mounds of shells lying two milea 
inland on this line of coast. 

» Notices of the East Indian Arch., Singapore, 1828, p. 6, and 
Append, p. 43. 


Basil Hall and Captain Beecbey ^ of tlie lines of inland 
reefs, and walls of coral rock worn into caves, above tbe 
present reacb of tbe waves, at tbe Loo Choo Islands, tbere 
can be little doubt tbat tbey bave been upraised at no very 
remote period. 

Dr. Davy ^ describes tbe nortbern province of Ceylon as 
being very low, and composed of a limestone witb sbells 
and corals of very recent origin ; be adds, tbat it does not 
admit of a doubt tbat tbe sea bas retired from tbis district 
even witbin tbe memory of man. Tbere is also some 
reason for believing tbat tbe eastern sbores of India, nortb 
of Ceylon, bave been upraised witbin tbe recent period.^ 
On tbe opposite side of tbe Gulf of Bengal, Captain Hal- 
stead everyvv^bere found during bis survey of tbe Burmese 
coast (as be informed Sir C. Lyell), proofs of recent eleva- 
tion in upraised beacbes and beds of sbells and corals. In 
tbe Indian Ocean Mauritius bas been recently upraised, 
as I bave sbown in tbe cbapter on fringing-reefs. Tbe 
nortbern extremity of Madagascar is described by Captain 
Owen * as formed of madreporitic rock, as likewise are tbe 

' Captain B. Hall, Voyage to Loo Choo, Append, pp. xxi. and xxv. 
Captain Beecbey's Voyage, p. 49G. 

^ Travels in Ceylon, p. 13. This madreporitic formation is men- 
tioned by M. Cordier in his report to tbe Institute (May 4, 1839) on 
the voyage of the Chevrette, as one of immense extent, and belonging 
to the latest tertiary period. 

^ Dr. Benza, in his Journey through the N. Circars (the Madras 
Lit. and Scient. Journal, vol. v.), has described a formation with 
recent freshwater and marine shells, occurring at the distance of 
three or four miles from the present shore. Dr. Benza, in conver- 
sation with me, attributed their position to a rise of the land. Dr. 
Malcolmson, however, (and there cannot be a higher authority on 
the geology of India,) informs me that he suspects that these beds 
may have been formed by the mere action of the waves and currents 
accumulating sediment. From analogy I should much incline to 
Dr. Benza's opinion. 

* Owen's Africa, vol. ii. p. 37, for Madagascar ; and for S. Africa, 
wol. i. pp. 412 and 426. Lieut. Boteler's narrative contains fuller 


shores and outlying islands along an immenso space of ' 
Eastern Africa, from a little north of the equator for 900 
miles southward. Nothing can be more vague than the 
expression 'madreporitic rock ;* but at the same time it is, 
I think, scarcely possible to look at the chart of the linear 
islets running in front of the coast from the equator far 
southward, and rising to a greater height than can be ac- 
counted for by the growth of coral, without feeling con- 
vinced that a line of fringing-reefs has been elevated at a 
period so recent, that no great changes have since taken 
place on the surface of this part of the globe. Some, also, 
of the higher islands of madreporitic rock on this coast, for 
instance Pemba, are singularly shaped, apparently showing 
the combined eifect of the growth of coral on submerged 
banks, together with their subsequent upheaval. Dr. Allan 
informs me that he never observed any elevated organic 
remains on the Seychelles, which come under our fringed 

The nature of the formations round the shores of the 
Bed Sea, as described by several authors, proves that the 
whole of this large area has been elevated within a very 
recent tertiary epoch. A part of this space in the ap- 
pended map is coloured blue, indicating the presence of 
barrier-reefs ; on which circumstance I shall presently 
make some remarks. Eilppell ' states that the tertiary 
formation, of which he has examined the organic remains, 
forms a fringe along the shores with a uniform height of 
from 30, to 40 feet, from the mouth of the Gulf of Suez to 
about lat. 26° ; but that south of 26°, the beds attain only 
the height of from 12 to 15 feet. This, however, can 
hardly be quite accurate ; although possibly there may be 

particulars regarding the coral rock, vol. i. p. 174, and vol. ii. pp. 41 
and 54. See also Euschenberger's Voyage round the World, vol. i 
p. 60. 

' Riippell, Eeise in Abyssinien, Band i. s. 141. 


a decrease in the elevation of the shores in the middle parts 
of the Ked Sea, for Dr. Malcolmson informs me that he 
collected shells and corals, apparently recent, from the 
clififs of Camaran Island (lat, 15° 30' N.) at a height of 
between 30 and 40 feet ; and Mr. Salt (Travels in Abyssinia) 
describes a similar formation a little southward on the op- 
posite shore at Amphila. Moreover, near the mouth of the 
Gulf of Suez, although on the coast opposite to that on 
which Dr. Eiippell says that the modern beds attain a 
height of only 30 to 40 feet, Mr. Burton^ found a deposit 
replete with existing species of shells, at the height of 200 
feet. In an admirable series of drawings by Captain 
Moresby, I could see how continuously the cliff-bounded, 
low, tertiary plains extended with a nearly equable height, 
both on the eastern and western shores. The southern 
coast of Arabia seems to have been subjected to the same 
elevatory movement, for Dr. Malcolmson found at Sahar 
low cliffs containing shells and corals apparently of recent 

The Persian Gulf abounds with coral-reefs ; but as in 
this shallow sea it is difficult to distinguish reefs from 
sandbanks, I have coloured only some near the mouth. 
Towards the head of the gulf Mr. Ainsworth'^ says that the 
land is worn into terraces, and that the strata contain 
organic remains of existing forms. 

The West Indian Archipelago of ' fringed islands ' alone 
remains to be mentioned : evidence of an elevation within 
a late tertiary epoch of nearly the whole of this great area, 
may be found in the works of almost all the geologists who 
have visited it. I will give some of the principal references 
in a note.^ 

' Lyell's Principles of Geology, 5th edition, vol. iv. p. 25. 

^ Ainsworth's Assyria and Babylon, p. 217. 

' These references only relate to works published before 1842 
the date of the first edition of this book. On Florida and the north 
shores of the Gulf of Mexico, Rogers' Report to Brit. Ascioc. vol. iii. 


On reviewing the above details it is impossible not 
to be struck with the number of eases in which upraised 
organic remains, apparently belonging to the recent 
period, have been found on the shores now frmged by 
reefs, and which are coloured red on our map. It may, 
however, be thought that similar proofs of elevation 
could be found on the coasts coloured blue, and which 
we have good reason to believe have recently subsided ; 
but such proofs cannot be found, with the few follow- 
ing and doubtful exceptions. 

The entire area of the Eed Sea appears to have been 
upraised within a late tertiary period ; nevertheless I 
have been compelled, though on unsatisfactory evidence 
(given in the Appendix), to class the reefs in the middle 
part of the coast, not as fringing, but as barrier-reefs. 
If, however, the statements should prove accurate re- 
spectmg the less height of the tertiary beds in the 
middle, compared wdth the northern and southern 
districts, we might well suspect that the former had 
subsided subsequently to a general elevation by which 
the whole area had previously been upraised. Several 
authors • have observed shehs and corals high up on the 

p. 14. — On the shores of Mexico, Humboldt, Polit. Essay on New 
iSpain, vol. i. p. 62. (I have also some corroborative facts with 
respect to the shores of Mexico.) — Honduras and the Antilles, Lyell's 
Principles, 5th ed. vol. iv. p. 22. — Santa Cruz and Barbadoes, Prof. 
Hovey, SiUiman's Journ. vol. xxxv. p. 74.— St. Domingo, Courro- 
joUes Jour, de Phys. torn. liv. p. 106.--Bahamas United Service 
Journ. No. Ixxi. pp. 218 and 224.— Jamaica, De la Beche, Geol. Man 
p. 142.— Cuba, Taylor in Lond. and Edin. Phil. Mag. vol. xi. p. 17. 
Dr. Daubeney also at a meeting of the Geolog. Soc. orally described 
Borae very modern beds lying on the N.W. parts of Cuba. I might 
have added many other less important references. [See Appendix II.] 
' Eilis, in his Polynesian liescarches, was the first to call atten- 


mountains of the Society Islands, — a group of islanda 
encircled by barrier-reefs, and which, therefore, must 
have recently subsided. Thus at Tahiti, Mr. Stutchbury 
found on the summit of one of the highest mountains, 
between 5,000 and 7,000 feet above the level of the sea, 
* a distinct and regular stratum of semi-fossil coral ; * 
but we cannot infer from such evidence as this that the 
island has been elevated within the recent period ; and 
on the other hand, several naturalists, including Mr. 
Dana and myself, have in vain searched near the coast 
for upraised shells and corals, where if present they 
could not have been overlooked.^ Two of the Harvey 

tion to these remains (vol. i. p. 38) and the tradition of the natives 
concerning them. See also Williams, Nar. of Miss. Enterprise, p. 
21 ; also Tyerman and G. Bennett, Journ. of Voyage, vol. i. p. 21.^ ; 
also Mr. Conthouy's Remarks, p. 51 ; but his principal fact, namely, 
that there is a mass of upraised coral on the narrow peninsula ol 
Tiarubu, is from hearsay evidence ; also Mr. Stutclibury, West oi 
England Journ. No. 1, p. 54. There is a passage in Von Zach, 
Corres. Astronom. vol. x. p. 2G6, inferring an uprising at Tahiti, 
from a footpath now used, which was formerly impassable ; but I 
particularly enquired from Feveral native chiefs, whether they knew 
of any change of this kind, and they were unanimous in giving me an 
answer in the negative. 

' [Some of the mountains rise to 7,000 feet. A depth of 25 to 35 
fathoms, which is the limit of the growing corals, is reached at from 
100 to 150 fathoms from the edge of the reef. The slope then steepens 
rapidly to 160 and 180 fathoms, which depth is reached at a distance 
of 225 to 250 fathoms from the edge of the reef ; to 100 fathoms the 
Blope is about 45°, thence to about 200 it is about 30°, and then it 
eases off. From 35 to 150 fathoms sponges, alcyonarians, corals, and 
other invertebrates were obtained ; beyond the latter, coral-sand with 
volcanic minerals and pelagic shells. Inside the lagoons the reefs 
were fringed with living corals, sloped downwards and outwards for 
a few feet, then plunged at once to depths of 10 and 16 fathoms. The 
deposit in the lagoons was m some places a coral-sand, in others a 
volcanic mud. There 13 evidence of some amount of upheaval. 


Islands, namely, Aitutald and Manouai, are formed of 
upraised coral rocks, and have probably been elevated 
within a recent period ; nevertheless they are encircled 
by reefs extending so far from the land, that I have 
coloured them blue, though with much hesitation, as 
the space within the reef is shallow, and the enchcled 
land is not abrupt. If these reefs really belong to the 
barrier class, we have here another instance of sub- 
sidence having followed elevation, both movements 
having been effected apparently within the recent 
period. There arc also many cases of coral-forma- 
tions, such as Elizabeth Island, Metia, Mangaia, 
Beveral of the Friendly and one of the Loyalty Islands, 
which it can hardly be doubted once existed as atolls, 
and were originally formed during subsidence, but 
have since been elevated, and are now surrounded by 
fringing-reefs. We have, however, no reason to feel 
surprise at occasional or even frequent alternations of 
level of the above two kinds. 

On the absence of active Volcanos in the areas of 
subsidence, a7id on their frequc7it presence in the areas 
of elevation.^— TliQ absence of active volcanos through- 
out the great areas of subsidence on our map, as 
shown by the pale and dark blue tints, — namely, in 
the central parts of the Indian Ocean, in the China 

Mr. Murray (p. 781) regards this reef as favouring his theory. Nar- 
rative of Challenger Voyage, p. 778.] 

' It may be well here lo state that all the roofs on the map were 
coloured either red or blue before the vermilion spots and streaks, 
showing the position of the active volcanos and volcanic chains, 
were added ; and indeed before I knew of the existence of several of 


Sea, in tlie sea between the barriers of Australia and 
New Caledonia, in the Caroline, Marshall, GUbert, and 
Low Archipelagoes, — is a very striking fact. So is the 
presence of active volcanic vents and chains on or near 
many of the shores coloured red on our map, and which 
are fringed with reefs ; for, as we have just seen, these 
fringed coasts have been recently upheaved in a large 
number of cases. Active volcanos likewise coincide 
with proofs of recent elevation on or near several 
other long lines of coast within the limits of our map, 
where there are no reefs of living corals, and which 
consequently are not coloured red. It must be here 
remarked, with regard to the proofs of both subsidence 
and elevation, that I do not judge by the absence or pre- 
sence or nature of the Coral-reefs round the volcanos 
themselves ; for, as Dana repeatedly insists, the corals 
may have been there destroyed or injured by the heat 
or exhalations. Nor have I taken into account the 
presence of upraised organic remains on the flanks of 
the volcanos themselves. I judge from the position of 
the active volcanic vents in relation to neighbouring 
islands and coasts, situated at too great a distance for 
any corals grov/ing there to be injured by the eruptions; 
and where, from the presence of atoll-formed or barrier- 
reefs, or of upraised marine remains, we have reason 
to believe that either subsidence or elevation has 
occurred within a recent period. 

The following cases offer a few partial exceptions to 
the rule that active volcanos are situated at a distance 
from the areps of subsidence. The Great Comoro 


Island probably contains a volcano, and it is only 
twenty miles distant from the barrier-reef of Mohilla. 
Ambil volcano, in the Phillippine Archipelago, is dis- 
tant only a little more than sixty miles from the atoll- 
formed Appoo reef: and there are two other volcanos 
on the map within ninety miles of circles colom'ed 
blue. But there is not a shiglc active volcano within 
several hundred miles of a group, even a small group, of 
atolls ; and it is clear that a group of atolls, surmount- 
ing a number of islands now all sunk beneath the level 
of the sea, implies a much greater amount of subsi- 
dence, than does a single atoll or a single encu'cling 
barrier-reef. It is a striking fact that two volcanos aro 
known to have been in recent action in the Friendly 
Archipelago ; and the islands have here been formed 
by the recent elevation of a group of atolls. Again, 
extinct craters and well-preserved streams of lava occur 
on many of the encircled islands in the Pacific, and 
these by our theory have subsided at no very remote 
period ; but although thus plainly formed of volcanic 
matter, they do not offer a single active volcano. In 
these cases the volcanos seem to have come into action 
or to have been extinguished, in accordance with the 
latest movements of elevation or subsidence. 

Within the limits of our map, active volcanos occur 
on or near other coasts besides those which are fringed 
with reefs and coloured red; and some of these coasts 
are known to have been upraised within the recent 
period. Thus I have shown in my Geological Observa- 
tions on S. America (184G) that the whole western shore 


of this great continent, for a space of between 2,000 and 
3,000 miles south of the equator, has undergone an up- 
ward movement during the period of existing marine 
shells ; and the Andes here form the grandest volcanic 
chain in the world. The islands on the north-western 
side of the Pacific, forming the second grandest volcanic 
chain, are very imperfectly known ; but Luzon, in the 
Phillippines, and the Loo Choo islands, have been re- 
cently elevated ; and at Kamtschatka ' there are exten- 
sive tertiary beds of modern date. The co-existence 
in other, parts of the world, of active volcanos with 
upraised beds of a modern origin, will occur to every 
geologist. Nevertheless, until it could be shown that 
volcanos were absent or inactive in subsiding areas, 
the conclusion that their distribution depended on the 
nature of the su'bterranean movements in progress, 
would have been hazardous. But now, viewing the 
appended map, it may, I think, be considered as 
almost established, that volcanos are often present in 
the areas which have lately risen or are still rising, 
and are invariably absent in those which have lately 
subsided or are still subsiding ; and this, I think, is 
the most important generalisation to which the study 
of coral-reefs has indirectly led me.^ 

On the dimensions and relative positions oj the 

' Namely, at Sedanka, in lat. 58'' N. (Von Buch's Descript. cleg 
Isles Canaries, p. 455). 

2 We may infer from this rule, that at any place where an old 
formation contains interstratilied beds of erupted matter, the surface 
of the land or the bed of the sea formed, at the period of eruption, a 
rising, at least not a subsiding area. 



svhsiding areas on our map, as indicated hy the pre- 
sence of atolls and harrier-reefs ; and of the rising or sta- 
tionary areas, as hioion hy upraised organic remains, 
or inferred from the presence of fringinc) -reefs. — The 
immense sm-faces seen on the map, which accord- 
ing to our theory, or from the plain evidence of up- 
raised remains, have undergone either a downward 
or upward change of level within a geologically late 
period, is a highly remarkable fact. The existence of 
continents shows that the areas which have been up- 
raised are immense. With respect to South America 
we may feel sure, and with respect to the western shores 
of the Indian Ocean we have reason to suspect, that 
this rising is either now actually in progress, or has 
taken place quite recently. By our theory, it may 
safely be inferred that the areas whicli have lately sub- 
sided are likewise immense ; or, judging from the earth- 
quakes now occasionally felt there, and from other ap- 
pearances, are still subsiding. The smallness of the 
scale of our map should not be overlooked ; each square 
on it containing 810,000 square miles. If we take the 
space of ocean from near the southern end of the Low 
Archipelago to the northern end of the Marshall Archi- 
pelago, — a length of 4,500 miles, we see that, as far as 
known, every island, excepting Metia, is atoll-formed. 
The eastern and western boundaries of our map are 
continents, and they are rising : the central areas of 
the great Indian and Pacific Oceans, are mostly sub- 
siding ; between them, north of Austraha, lies the most 
broken land on the globe, and there the rising parts 


are surrounded and penetrated by areas of subsidence ; ' 
80 that the prevaihng movements now in progress, seem 
to accord with the present state of the great terrestrial 
and oceanic divisions of the world. 

The blue spaces on the map are nearly all elongated ; 
Buch as the great north and south line of atolls in the 
Indian Ocean, the space between the barrier-reefs of 
Australia and New Caledonia, the Caroline Archi- 
pelago, &c. Whether adjoining elongated spaces, run- 
ning in different directions, have subsided by one com- 
mon movement, or independently of each other, we do 
not know. In the case of the Caroline and Marshall 
Archipelagoes, situated near each other, but extending 
in different directions, it seems probable that they have 
subsided independently of each other ; for the McAskill 
Islands,^ ly^ng towards the eastern end of the Caroline 
Archipelago, are formed of upraised coral-rock; and we 
thus see that the above two areas of subsidence have 
been at one time interrupted by an area of upheaval. 
The curved line of elevation formed by the Mariana 
Islands, seems to cross a former line of subsidence pro- 
longed from the Caroline Archipelago ; for the island 
of Fais, apparently an upraised atoll, is situated nearly 
at the point of intersection of the two lines. The 
Sandwich Archipelago is 530 miles in length, from 
Hawaii to the westernmost rocky islet, but is pro- 

1 I suspect that the Arm and Timor-laut Islands present an in- 
chided small area of subsidence, like that of the China Sea ; but I 
have not ventured to colour them blue, owing to the want of suilieiont 
information. See Aj^pendix. 

^ Dana, Corals and Coral Islands, p. 306. 


longed by numerous reefs to a point 2,000 miles dis- 
tant from Hawaii. The south-eastern end of this long 
line is one of elevation and of volcanic activity; 
"whereas the north-western end, judging fi*om the 
structm-e of the reefs, though these are imperfectly 
knowai, is one of subsidence.^ So that here we ap- 
parently have opposite movements in progress towards 
the two extremities of the same long line. The com- 
monest case seems to be a tendency to alternation 
between the areas of subsidence and elevation, as 
if the sinking of one had counterbalanced the rising 
of another. 

The existence in many parts of the w'orld of lofty 
table-land, proves that large surfaces have been upraised 
in mass to a great height above the level of the sea ; 
although in almost every country the highest points 
consist of upturned strp,ta, or of erupted matter : and 
from the wide spaces over which atolls are scattered, 
although not one pinnacle of land now remains above 
the level of the sea, we may conclude that immense 
areas have subsided to an amount sufficient to bury not 
only any formerly existing lofty table-land, but even the 
heights formed by fractured strata and erupted matter. 
The effects left on the land by the later elevatory 
movements, namely, successively rising cliffs, succes- 
sive lines of erosion, and great beds of shells and 
pebbles, all requh-ing time for their production, prove 
that these movements have been extremely slow. And 

' Dana, Corals and Coral Islands, pp. 307, 355. See also mj 


with respect to the whole amount of subsidence neces- 
sary to have produced the many atolls widely scattered 
over immense spaces, the movement, as already shown, 
must either have been uniform and exceedingly slow, 
or effected by small steps separated from each other by 
long intervals of time, so as to have allowed the reef-con- 
atructing polypifers to bring up their solid frameworks 
to the surface ; and this is one of the most interesting 
conclusions to which we are led by the study of coral- 
formations. We have little means of judging whether 
many considerable oscillations of level have usually 
occurred during the elevation of large areas ; but we 
know from clear geological evidence, such as trees still 
standing upright at successive levels and covered by 
marine strata, that this has frequently been the case ; 
and we have seen on our map, that some of the same 
islands after having subsided, have been upraised ; and 
that others after having been uplifted, have subsided. 
We may therefore conclude that the subterranean 
changes which cause some areas to rise and others to 
sink, have generally acted in a closely similar manner. 

Eecapitulation. — In the three first chapters, the 
principal kinds of coral-reefs were described in detail, 
and they were found to differ little, as far as relates 
to the actual surface of the reef. An atoll differs from 
an encircling barrier-reef only in the absence of land 
within its central expanse ; and a barrier-reef differs 
from a fringing-reef only in being placed, relatively 
to the probable inclmation of its submarine foundation, 

1 04 KKCAriTULATION. Cii. VI, 

at a much greater distance from the land, and in 
the presence of a deep lagoon-hke space within tha 
reef. In the fourth chapter the growing powers of tho 
reef-constructing polypifers were discussed ; and it was 
shown that they cannot flourish heneath a very limited 
depth. In accordance with this limit, there is no diffi- 
culty respecting the foundation on which a fringmg- 
reef is based ; whereas, with harrier-reefs and atolls, 
there is the greatest difficulty on this head ; — in bar- 
rier-reefs from the improbability of rock or banks of 
sediment having extended, in every instance, so far 
seaward within the required depth ;— and in atoUs, 
from the immensity of the spaces over which they are 
interspersed, and the apparent necessity for believing 
that they are all based on mountam-summits, which, 
although rising very near to the surface of the sea, in 
no one instance rise above it. To escape this latter 
admission, which implies the existence of submarine 
chains of mountains of almost exactly the same height 
extending over many thousand square miles, there is 
but one alternative ; namely, the prolonged subsidence 
of the foundations on which the atolls first became 
attached, together with the upward growth of the 
reef-constructing corals. On this view every difficulty 
vanishes : fringing-reefs are thus easily converted into 
barrier-reefs ; and barrier-reefs into atolls, as soon as 
the last pinnacle of land sinks beneath the surface of 
the sea. 

The wall-hke structure on the inner sides of atoll3 
and barrier-reefs— the basin or ring-like shape of the 


marginal and central reefs in the Maldiva atolls — the 
union of some atolls as if by a ribbon — the apparent 
disseverment of others — the ordinary outline of groups 
of atolls and their forms — are all thus explained. We 
thus understand the occurrence in both atolls and 
barrier-reefs of portions, or of the whole, in a dead and 
submerged condition, though still retaining the outline 
of a living reef. The existence of breaches through 
barrier-reefs in front of valleys, though separated from 
them by wide spaces of deep water, can be similarly ex- 
plained. It confirms our theory that we find the two 
kinds of reefs formed through subsidence generally situ- 
ated near each other and at a distance from the spaces 
where fringing-reefs abound. On searching for other 
evidence of the movements assumed by the theory, 
we find marks of change in atolls and in barrier-reefs, 
and of subterranean disturbances beneath them ; but 
from the nature of things, it is scarcely possible to 
find direct proofs of subsidence, although some appear- 
ances are strongly in favour of it. On the fringed 
coasts, however, the frequent presence of upraised 
marine remains belonging to a recent epoch, plainly 
shows that these coasts have been lately elevated. 

Finally, when the two great types of structure, 
namely barrier-reefs and atolls on the one hand, 
and fringing-reefs on the other, are laid down on a 
map, they offer a grand and harmonious picture of 
the movements which the crust of the earth has 
undergone within a late period. We there see vast 
areas rising, with volcanic matter every now and then 


bursting forth. We see otlicr \Yide spaces sinldng with- 
out any volcanic outbursts ; and we may feel sure that 
the movement has been so slow as to have allowed the 
corals to grow up to the surface, and so widely extended 
as to have buried over the broad face of the ocean 
every one of those mountains, above which the atolls 
now stand like monuments, marking the place of their 





In the beginning of the last chapiter I stated the principles 
on which the map has been coloured. There only remains 
to be said, that it is an exact copyof one by M. C. Gressier, 
published by the Depot General de la Marine, in 1835. 
The names have been altered into English, and the longi- 
tude has been reduced to that of Greenwich. The colours 
were first laid down on accurate charts, on a large scale. 
The data, on which the volcauos historically known to 
have been in action, have been marked with vermilion, 
were given in a note to the last chapter. I will commence 
my description on the eastern side of the map, and will 
describe each group of islands consecutively, proceeding 
westward across the Pacific and Indian Oceans, and ending 
with the West Indies. 

The Western Shores of America appear to be 
entirely without coral-reefs : south of the equator the survey 
of the Beagle, and north of it the published charts show 
that this is the case. Even in the Bay of Panama, where 
corals flourish, there are no true coral-reefs, as I have been 
informed by Mr. Lloyd. There are no coral-reefs in the 
Galajjagos archipelago, as I know from personal inspection ; 
and I believe there are none on the Cocos, Bevilla-gigedo, 
and other neighbouring islands. Clipperto7i^ rock, 10° N., 

' [Undoubtedly an atoll, according to Sir J. Belcher's chari.— 
Captain Wharton.] 


109° W., from a drawing appended to a MS. plan in the 
Admiralty, does not appear to bo an atoll, but Sir E. 
Belcher (Voyage romid the World, vol. i. 1813, p. 255) 
speaks of it as of coral-formation, with deep water within 
the lagoon ; left micoloured. The eastern part of the 
Pacific presents an enormous area without any islands, 
except Easter and Gomez, which do not appear to be sur- 
rounded by reefs. 

The Low or Paujiotu Archipelago. — This group 
consists of about 80 atolls : it would be quite superfluous 
to refer to descriptions of each. In D'Urvillc and Lottin'a 
chart, one island {Wolchonshy) is written with a capital 
letter, signifying, as explained in a former chapter, that it 
is a high island ; but this must be a mistake, as the orgina.l 
chart by Bellingshausen shows that it is a true atoll. 
Captain Beechey says of the 32 groups which he examined 
(of the greater number of which I have seen beautiful MS. 
charts in the Admiralty), that 29 now contain lagoons, and 
lie believes the other three orginally did so. Bellingshausen 
(see an account of this Russian voyage, in the Biblioth. des 
Voyages, 1834, p. 443) says that the 17 islands which he 
discovered resembled each other in structure, and he has 
given charts on a large scale of all of them. Kotzebue has 
given plans of several ; Cook and Bligh mention others ; a 
few were seen during the voyage of the Beagle ; and notices 
of other atolls are scattered through several publications. 
The Actaon group in this archipelago has lately been dis- 
covered (Geograph. Journ., vol. vii. p. 454) ; it consists of 
three small and low islets, one of which has a lagoon. 
Another lagoon-island has been discovered (Naut. Mag. 
1889, p. 770) in 22° 4' S. and 13G° 20' W. Dana, in his 
work on Corals and Coral Islands, gives a full account of 
this archipelago. Towards the S.E. there are some islands 
of a diiferent nature : Elizabeth Island is described by 
Beechey (p. 46, 4to edit.) as fringed by reefs, at the distance 


of between two and three hundred yards ; coloured red, 
Fitcairn Island, m the immediate neighbourhood, according 
to the same authority, has no reefs of any kind, althougli 
numerous pieces of coral are thrown up on the beach ; the 
sea close to its shore is very deep (see Zool. of Beechey's 
Voyage, p. 164) ; left uncoloured. Gavihier Islands (see 
Plate I. fig. 8) are encircled by a barrier-reef; the greatest 
depth within is 38 fathoms ; coloured pale blue. Mctia or 
Aurora Island lies N.E. of Tahiti, close to the large spaco 
coloured dark blue in the map ; it has been already de- 
scribed as an upraised atoll ; as it is said by Captain Wilkes 
(Narrative of U.S. Exploring Expedition, vol. i. p. 337) to 
be surrounded by fringing-reefs, in one part 500 feet in 
width, it has been coloured red. But I must remind the 
reader of the discussion in the sixth chapter, showing that 
if an upraised atoll were to subside again, the reef would 
probably retain for a long time or for ever, its fringing 
character, owing to the steepness of the submarine flanks. 
The Society Aechipelago is separated by a narrow 
space from the Low Archipelago ; and in their parallel 
direction they manifest some relation to each other. I 
have already described the general character of the reefs 
of these encircled islands. In the atlas of the Coquille's 
Voyage there is a good general chart of the group, and 
separate plans of some of the islands. TaJiiti, the largest 
island in the group, is almost surrounded, as seen in Cook's 
chart, by a reef h'om half a mile to a mile and a half from 
the shore, with from 10 to 30 fathoms within it. Some 
considerable submerged reefs, lying parallel to the shore, 
with a broad and deep space within, have lately been dis- 
covered on the N.E. coast of the island, (Naut. Mag. 1836, 
p. 26-1,) where none are laid down by Cook. At Eimeo the 
reef, ' which like a ring surrounds it, is in some places one 
or two miles distant from the shore, in others united to 
the beach ' (Ellis, Polynesian Kesearches, vol. i. p. 18, 12mo. 


edit.). Cook found deep water (20 fathoms) in some of the 
harbours within the reef. Mr. Couthouy, however, states 
(Kemarks, p. 45) that both at Tahiti and Eimeo, the space 
between the barrier-reef and the shore has been almost 
lillod up, — ' a nearly continuous fringing-reef surrounding 
the island, and varying from a few yards to rather more 
than a mile in width, the lagoons merely forming canals 
between this and the sea-reef,' that is the barrier-reef. 
Tapamanoa is surrounded by a reef at a considerable 
distance from the shore ; from the island being small, 
it is breached, as I am informed by the Rev. W. Ellis, 
only by a narrow and crooked boat-channel. This is the 
lowest island in the group, its height probably not ex- 
ceeding 500 feet. A little way north of Tahiti, the low 
coral islets of Teturoa are situated ; from the description of 
them given me by the Eev. J. Williams (tlie author of the 
Karrative of Missionary Enterprise), I should have thought 
that they formed a small atoll, and likewise from the de- 
scription given by the Eev. D. Tverman and G. Bennett 
(Journ.ofVoy. and Travels, vol. i. p. 183), who say that ten 
low coral islets * are comprehended within one general reef, 
and separated from each other by interjacent lagoons ; ' but 
as Mr. Stutchbury (West of England Journal, vol. i. p. 54) 
describes it as consisting of a mere narrow ridge, I have 
left it uncoloured. Maitea, eastward of the group, is classed 
by Forster as a high encircled island ; but from the account 
given by the Rev. D. Tyerman and G. Bennett (vol. i. p. 57) 
it appears to be an exceedingly abrupt cone rising from 
the sea without any reef ; left iincoloured. It would be 
Buperfluous to describe the northern islands in this group, 
as they may be well seen in the chart accompanying the 
4 to. edition of Cook's Voyages, and in the atlas of the 
Coqiiille's Voyage. Mmmia is the only one of the northern 
islands in which the water within the reef is not deep, being 
only 41 fathoms ; but the great width of the reef, stretching 



tliree miles and a half southward of the land (which is 
represented in the drawing in the atlas of the Coquille's 
Voyage as descending abruptly to the water), shows, on tlie 
principle explained in the beginning of the last chapter, 
that it belongs to the barrier class. I may here mention, 
from information communicated to me by the Eev. W. 
Elhs, that on the N.E. side of Hiiaheine there is a bank of 
sand, about a quarter of a mile wide, extending parallel to 
the shore, and separated from it by an extensive and deep 
lagoon : this bank of sand rests on coral-rock, which un- 
doubtedly was once a hving reef. North of Bolabola lies 
the atoll of Toubai (Motou-iti of the Coquille's atlas), which 
is coloured dark blue ; all the islands which are surrounded 
by barrier-reefs are coloured pale blue : three of them are 
represented in figures 3, 4, and 5, in Plate I. There are 
three low coral-groups lying a Httle W. of the Society 
Archipelago, and almost forming part of it, namely, Bell- 
ingshausen, which is said by Kotzebue (Second Voyage, 
vol. ii. p. 255) to be a lagoon-island ; Mopeha, which from 
Cook's description (Second Voyage, book iii. chap, i.) no 
doubt is an atoll ; and the Scilly Islands, which are said 
by Wallis (Voyage, chap, ix.) to form a group of loio islets 
and shoals, and which, therefore, probably compose an 
atoll : the two former have been coloured blue, but not the 

Mendana or Marquesas Group. — These islands are 
almost entirely destitute of reefs, as may be seen in 
Krusenstern's Atlas, making a remarkable contrast with 
the adjacent group of the Society Islands. Mr. F. D. 
Bennett has given some account of this group, in tho 
seventh volume of the Geograph. Journ. He informs me 
that all the islands have the same general character, and 
that the water is very deep close to their shores. He 
visited three of them, namely, Dominicana, Christiana, 
and Boapoa, their beaches are strewed with rounded masses 


of coral, and although no regular reefs exist, yet the shore 
is in many places lined by coral rock, so that a boat 
grounds on this formation. Hence these islands ought 
perhaps to come within the class of fringed islands and be 
coloured red ; but as I am determined to err on the cautious 
side, I have left them uncoloured. Dana mfers (Corals 
and Coral Islands, p. 325), from their steepness and deeply 
indented outline, that they have subsided. 

Cook ob Haiivey and Austral Islands. — Palmerston 
Island is minutely described as an atoll by Captain Cook 
during his voyage in 1774 ; it is coloured blue. Aitutaki 
■was partially surveyed by the Beagle (see map accompany- 
ing Voyages of Adventure and Beagle) ; the land is hilly, 
sloping gently to the beach ; the highest point is 300 feet ; 
on the southern side, the reef projects five miles from the 
land : off this point the Beagle found no bottom with 270 
fathoms : the reef is surmounted by many low coral-islets. 
I am informed by the Eev. J. Williams, that within the 
reef the water is exceedingly shallow, not being more than 
a few feet deep ; nevertheless, from the great extension of 
the reef into a profoundly deep ocean, this island probably 
belongs, on the principle lately adverted to, to the barrier 
class, and I have coloured it pale blue, although with much 
hesitation. — Manouai or Harvey Island : the highest point 
is about 50 feet : the Eev. J. "Williams informs me that 
although the reef lies far from the shore, it is less distant 
than at Aitutaki, but the water within the reef is rather 
deeper : I have likewise coloured this island pale blue, but 
with many doubts. — Round Mitiaro Island, as I am in- 
formed by Mr. WiUiams, the reef is attached to the shore ; 
coloured red. — Matiki, or Maouti : the reef round this 
island (under the name of Parry Island in the Voyage of 
H.M.S. Blonde, p. 209) is described as a coral flat, only 
50 yards wide, and two feet under water. This statement 
has been corroborated by Mr. Williams, who calls the reef 


attached; coloured red. — Atiu, ov Wateeo: a moderately 
elevated, hilly island, like the others of the group ; the 
reef is described in Cook's Voyage as attached to the shore, 
and about 100 yards wide ; coloured red. — Fenoua-iti : 
Cook describes this island as very low, not more than six 
or seven feet in height (vol. i. book ii. chap. iii. 1777) ; in 
the chart pubhshed in the Coquillc's atlas, a reef is en- 
graved close to the shore : this island is not mentioned in 
the list given by Mr. Wilhams (p. IG) in the Narrative of 
Missionary Enterprise ; nature doubtful ; but as it lies so 
near Atiu, it has been unavoidably coloured red. — Baro- 
tonga: Mr. Williams informs me that this is a lofty 
basaltic island, with an attached reef; coloured red. — ■ 
There are three other islands, Bourouti, Boxburgh, and 
Hull, of which I have not been able to obtain any account, 
and have left them imcoloured, Hicll Island, in the French 
chart, is written with small letters as being low. — 
Mangaia : height about 300 feet ; ' the surrounding reef 
joins the shore ' (Williams's Narrative, p. 18) ; coloured red. 
—Bimetara : Mr. Williams informs me that the reef is 
rather close to the shore ; but, from information given me 
by Mr. Ellis, the reef does not appear to be quite so closely 
attached to it as in the foregoing cases : the island is about 
300 feet high (Naut. Mag. 1839, p. 738) ; coloured rod.— 
Burutu: Mr. Wilhams and Mr. Ellis inform me that this 
island has an attached reef ; coloured red. It is described 
by Cook under the name of Oheteroa : he says it is not 
surrounded like the neighbouring islands, by a reef ; but 
he must mean a distant reef. — Toubouai : in Cook's chart 
(Second Voyage, vol. ii. p. 2) the reef is laid down in a 
part at the distance of one mile, and in another part at the 
distance of two miles from the shore ; Mr. Ellis (Polynes. 
Ees. vol. iii. p. 381) says the low land round the base of 
the island is very extensive ; and this gentleman informs 
me that the water within the reef appears deep ; coloured 



bine. — Baivaivai, oi Vivitao : Mr. Williams informs me 
that the reef is hero distant from the shore ; Mr. Ellis, 
however, says that this is certainly not the case on oiio 
Bide of the island ; and he believes that the water within 
the reef is not deep ; hence I have left it uncoloured. — 
Lancaster Eecf, described in Naut. Mag. 1833 (p. G93), as 
an extensive crescent-formed coral-reef, has not been 
coloured, — Bcqya, or Oparree : from the accounts given of 
it by Ellis and Vancouver, there does not appear to be any 
reef. — I. de Bass is an adjoining island, of which I cannot 
find any account. — Eeviin Island : Krusenstern seems 
hardly to know its position, and gives no further par- 

Islands between the Loio ajid Gilbert Archipelagoes 

Caroline Island (10° S., 150° W.) is described by Mr. 
F. D. Bennett (Geograph. Journ, vol. vii. p. 225) as con- 
taining a fine lagoon ; coloured blue. Westward of Caroline 
Island, a small lagoon-island is described in the U.S. 
Exploring Expedition in lat. 10° S. and 152° 22' W. long. ; 
coloured hlue.— Flint Island (11° S., 151° W.) : Kruson- 
stern believes that it is the same with Percgrino, which is 
described by Quiros (Burney's Chron. Hist. vol. ii. p. 283) 
as ' a cluster of small islands connected by a reef, and 
forming a lagoon in the middle ; ' coloured blue. — Wostock 
is an island a little more than half a mile in diameter, and 
apparently quite flat and low, discovered by Bellingshausen; 
it is situated a little west of Caroline Island, but it is not 
placed on the French charts ; I have not coloured it, 
although I entertain little doubt, from the chart of Bellings- 
hausen, that it originally contained a small lagoon. — Pen- 
rhyn Island (9° S., 158° W.) : a plan in the atlas of tho 
First Voyage of Kotzebue, shows that it is an atoll, which 
according to Wilkes (U.S. Exploring Expedition, vol. iv. 
p. 277) JK nine miles in length ; coloured blue. — Starbuck 


Island (5° S., 15G° W.) is described in Byron's Voyaj^e in 
the Blonde (p. 206) as formed of a flat coral-rock, with no 
trees ; the height not given ; not coloured. — Maiden Island' 
(4° S., 154° W.) : in the same Voyage (p. 205) this island 
is said to be of coral formation, and no part above 40 feet 
high ; I have not ventured to colour it, although from being 
of coral formation, it is probably fringed ; in which case it 
should be red. — Jarvis, or Bunker Island (0° 20' S., 1C0° 
W.) is described by Mr. F. D. Bennett (Geograph. Journ. 
vol. vii. p. 227) as a narrow, low strip of coral formation ; 
not coloured. — Broohis a small, low island between the two 
latter; its position, and perhaps even existence is doubtful; 
not Golomed.—Pcscado and Humphrey Islands : I can find 
out nothing about these islands, except that the latter ap- 
pears to be small and low ; not coloured. — Eearson, or Grand 
Duke Alexander's (10° S., 161° W.) : an atoll, of which a 
plan is given by Bellingshausen ; blue. — Souvorojf Islands 
(13° S., 163° W.) : Admiral Krusenstern, in the most 
obliging manner, obtained for me an account of these islands 
from Admiral Lazareff, who discovered them. They con- 
sist of five very low islands of coral formation, two of which 
are connected by a reef, with deep water close to it. Tliey 
do not surround a lagoon, but are so placed that a line 
drawn through them includes an oval space, part of which 
is shallow ; these islets, therefore, probably once (as is the 
case with some of the islands in the Caroline Archipelago) 
formed a single atoll ^ ; but I have not coloured them. — 
Danger Island (10° S., 166° W.) : described as low by 
Commodore Byron, and more lately surveyed by Bellings- 
hausen ; it is a small atoll with three islets on it ; blue. — • 
C7amzcc Island (9° S., 172° W.) : discovered in the PaJzcZora 
(G. Hamilton's Voyage, p. 75) : it is said, ' In running along 
the land, we saw several canoes crossing the lacjowiis ; ' aa 

' [Sfarb2(cJc and Maiden Islands are fringed.— Captain Wharton.] 
» [Suvcroff is a complete atoll (French chart). -Captain Wharton.] 


lliis island is in the close vicinity of other low islands, and 
as it is said that the natives make reservoirs of water in 
old cocoa-nut trees (which shows the nature of the land), I 
have no doubt it is an atoll, and have coloured it blue. — 
York Island (8° S., 172° W.) is described by Commodore 
Byron (chap: x. of his Voyage) as an atoll; blue. — Sydney 
Island (4° S., 172° W.) is about three miles in diameter, with 
its interior occupied by a lagoon (Captain Tromelin, Annal. 
Marit. 1829, p. 297) ; coloured blue. — H^dl Island is situ- 
ated GO miles to the west of Sydney Island, and is described 
by Wilkes (U.S. Exploring Expedition, vol. iii. p. 3G9) as a 
lagoon-island ; coloured blue. — Phonnix Island (4° S., 171° 
W.) is nearly circular, low, sandy, not more than two miles 
in diameter, and very steep outside (Tromelin, Annal. Marit. 
1829, p. 297) : it may be inferred that this island originally 
contained a lagoon, but I have not coloured it. — New Nan- 
tucket (0° 15' N., 174° W.) : from the French chart it must 
be a low island ; I can find nothing more about it, or about 
Mary Island ; both uncoloured. — Gardiier Island (5° S., 
174° W.), from its position, is certainly the same as Kemhi 
Island, and is described (Krusenstern, p. 435, Appen. to 
Mem. published 1827) as having a lagoon in its centre ; 
coloured blue. 

Islands south of the Sandivich Archipelago. 

Christmas Island (2° N., 157° W.) : Captain Cook, in his 
Third Voyage (vol. ii. chap, x.), has given a detailed account 
of this atoll. The breadth of the islets on the reef is un- 
usually great, and the sea near it does not deepen so sud- 
denly as is generally the case. It has more lately been 
visited by Mr. F. D. Bennett (Geograph. Journ. vol. \ii. 
p. 22G) ; and he assures me that it is low and of coral 
formation : I particularly mention this, because it is en- 
graved with a capital letter, signifying a high island, in 
D'Urville and Lottin's chart. Mr. Couthouy, also, has 


given some account of it (Remarks, p. 46) from the 
Hawaiian Spectator ; he believes it has lately undergone a 
small elevation, but his evidence does not appear to me 
satisfactory ; the deepest part of the lagoon is said to be 
only ten feet ; nevertheless, I have coloured it blue. — Fan- 
ning Island (4° N., 158° W.), according to Captain Tromelin 
(Ann. Maritim. 1829, p. 283), is an atoll : his account, as 
observed by Krusenstern, differs from that given in Fan- 
ning's Voyage (p. 224), which, however, is far from clear ; 
coloured blue. — Washington Island (4° N., 159° W.) is en- 
graved as a low island in D'Urville's chart, but is described 
by Fanning (p. 22G) as having a much greater elevation than 
Fanning Island, and hence I presume it is not an atoll ; 
not coloured. — Palmyra Island (6° N., 1G2° W.) is an atoll 
divided into two parts (Krusenstern's Mem. Suppl. p. 50, 
also Fanning's Voyage, p. 233) ; blue. — Smyth's, or John- 
ston's Islands (17° N., 170° W.) : Captain Smyth, R.N., 
has had the kindness to inform me that they consist of two 
very low small islands, with a dangerous reef off the east 
end of them ; Captain Smyth does not recollect whether 
these islets, together with the reef, surrounded a lagoon ; 

Sandwich Archipelago. — Hawaii : in the chart in 
Freycinet's Atlas small portions of the coast are fringed 
by reefs ; and in the accompanying Hydrog. Memoir, reefs 
are mentioned in several places, and the coral is said to 
injure the cables ; but Dana saw hardly any reefs here.' 

' [Prof. Dana, noticing this remark in Silliman's Amer. Jour., Dec. 
1874, states the result of further enquiries on his part from the Eev. 
Mr. M'Coan, long a resident of Hilo : — ' With respect to your enquiry 
whether there is any elevated coral-reef rock around the shores of 
Hawaii, I would reply that I think not. . . . Honolulu, on the island of 
Oahu, is built much of it upon the elevated coral-reef rock, and there 
Bre large areas in the district of Waiana and other portions of the Oahu 
shores : but there is nothing of this kind on Hawaii. You are awara 
that corals, even under the water, are on the weather [eastern] side 


On one side of the islet of Kohailiai there is a bank of 
sand and coral with five feet of water on it, running parallel 
to the shore, and leaving a channel of about fifteen feet 
deep within. I have coloured this island red, but it is 
very much less perfectly fringed than others of the group. 
— Maui : in Frcyciuet's chart of the anchorage of Eaheina, 
two or three miles of coast are seen to be fringed ; and in 
the Hydrog. Memoir ' banks of coral along shore ' are 
spoken of. Mr. F. D. Bennett informs me that the reefs, 
on an average, extend about a quarter of a mile from the 
beach ; the land is not very steep, and outside the reefs the 
sea does not become suddenly deep ; coloured red. — Morotoi, 
I presume, is fringed : Freycinet speaks of the breakers 
extending along the shore at a little distance from it. 
From the chart, I believe it is fringed ; coloured red. — Oahu: 
Freycinet, in his Hydrog. Memoir, mentions some reefs. 
Mr. F, D. Bennett informs me that the shore is skirted for 
forty or fifty miles in length. There is even a harbour for 
ships formed by the reefs, but it is at the mouth of a valley ; 
red. — A tool, in La Peyrouse's charts, is represented as 
fringed by a reef, in the same manner as Oahu and 
Morotoi ; and this, I am informed by Mr. Ellis, is of coral- 
formation on part at least of the shore ; the reef does not 
leave a deep channel within ; red. — Oneehoio : Mr. Ellis 
believes that this island is also fringed by a coral-reef : 
considering its close proximity to the other islands, I have 
ventured to colour it red. I have in vain consulted the works 
of Cook, Vancouver, La Peyrouse, and Lisiausky for any 
satisfactory account of the small islands and reefs which lie 
scattered in a N.W. line prolonged for a great distance from 
the Sandwich group, and hence have left them uncoloured, 
with one exception ; for I am indebted to Mr. F. D. Bennett 
for informing me of an atoll-formed reef, in lat. 28° 22' N., 

of this island not abuiulant.' In the Narrative of the Challenger 
Voyage (p. G99) reefa are mentioned as occurring at Honolulu.] 


long. 178° 30' W., on which the Gledstanes was wrecked 
in 1837. It is apparently of large size, and extends in a 
N.W. and S.E. line : very few islets have been formed on it. 
The lagoon seems to be shallow ; at least, the deepest part 
which was surveyed was only three fathoms. Mr. Couthouy 
(Bemarks, p. 38) describes this island under the name of 
Ocea7i Island. Considerable doubts should be entertained 
regarding the nature of a reef of this kind, with a very 
shallow lagoon, and standing far from any other atoll, on 
account of the possibility of a crater or flat bank of rock 
lying at the proper depth beneath the surface of the water, 
having afforded a foundation for a ring-formed coral- 
reef. I have, however, thought myseK compelled, from 
its large size and symmetrical outline, to colour it blue. 
Some information and references are given by Dana (Corals 
and Coral Islands, pp. 324, 3G5) with respect to the reefs 
and islets extending for 2,000 miles in a N.W. line from 

Samoa oe Navigator Geoup. — Kotzebue, in his 
Second Voyage, contrasts these islands with many others 
in the Pacific, in not having harbours for ships, formed by 
distant coral-reefs. The Rev. J. Williams, however, informs 
me that coral-reefs do occur in irregular patches on the 
shores ; but that they do not form a continuous band as round 
Mangaia, and other such perfect cases of fringed islands. 
From the charts accompanymg La Peyrouse's Voyage, it 
appears that the north shore of Savaii, Maouna, Orosenga, 
and Manua are fringed by reefs. La Peyrouse, speaking 
of Maouna (p. 126), says that the coral-reef surrounding its 
shores almost touches the beach, and is breached in front 
of the little coves and streams, forming passages, for canoes, 
and probably even for boats. Further on (p. 159) he ex- 
tends the same observation to all the islands which he 
visited. — Mr. Williams in his Narrative, speaks of a reef 
going round a small island attached to Oyolava, and return- 


ing again to it : all tliose islands have been coloured red.— 
A chart of Base Island, at the extreme [east] end of tlie 
group, is given by Freycinet, from which I should have 
thought that it had been an atoll ; ' but according to Mr. 
Couthouy (Eemarks, p. 43) it consists of a reef, only a 
league in circuit, surmounted by a very few low islets ; the 
lagoon is very shallow, and is strewed with numerous large 
boulders of volcanic rock. This island, therefore, probably 
consists of a bank of rock, a few feet submerged, with the outer 
margin fringed with reefs ; hence it cannot be properly 
classed with atolls, in which, as we have reason to believe, 
the foundations always lie at a depth greater than that at 
which the reef-constructing polypifers can live ; not coloured. 

Beveridge Keef, 20° S., 107° W., is described in the 
Naut. Mag. (May 1833, p. 4-12) as ten miles long in a N. 
and S. line, and eight wide ; ' in the inside of the reef, there 
appears deep water ; ' there is a passage near the S.W. 
corner : this therefore seems to be a submerged atoll, and 
is coloured blue. 

Savage Island, 19° S., 170° W., has been described by 
Cook and Forster. The younger Forster (vol. ii. p. 1G3) 
says it is about 40 feet high : he suspects that it contains 
a low plain, which formerly was the lagoon. The Kev. J. 
Williams gives 100 feet as its height, and he informs me 
that the reef fringing its shores resembles that romad 
Mangaia ; colovu-ed red. 

Feiendly Archipelago. — Pylstaart Island: judging 
from the chart in Freycinet's Atlas I should have supposed 
that it had been regularly fringed ; but as nothing is said in 
the Hydrog. Memoir (or in the Voyage of Tasman, the dis- 

' [It is an atoll. — Capt. Wharton. Eose Island Las a lagoon six 
to twelve fathoms deep and an entrance to it of four fathoms. 
Excej)t two small banks, one su2:)i)orting a group of trees, it is under 
water at high tide. — Letter from Prof. Dana to Mr. Darwin, July 21 


coverer) about coral-reefs, I have left it uncoloured. — 
Tongatahoii : in the atlas of the Voyage of the Astrolabe, 
the whole south side of the island is represented as 
narrowly fringed by the same reef which forms an exten- 
Bive platform on the northern side. The origin of this 
latter reef, which might have been mistaken for a barrier- 
reef, has already been attempted to be explained, when 
giving the proofs of the recent elevation of this island. — In 
Cook's charts the little outlying island of Eoaigec is repre- 
sented as fringed ; coloured red. — Eoua : I cannot make 
out from Captain Cook's charts and descriptions that this 
island has any reef, although the bottom of the neighbour- 
ing sea seems to be covered with corals, and the island 
itself is formed of coral-rock. Forster, however, distinctly 
(Observations, p. 14) classes it with the high islands 
having reefs, but it certainly is not encircled by a barrier- 
reef; and the younger Forster (Voyage, vol. i. p. 42G) says, 
that * a bed of coral rocks surrounded the coast towards the 
landing-place.' I have therefore classed it with the fringed 
islands, and coloured it red. Dana also shows (Corals and 
Coral Islands, p. 337) that most of the islands of this group 
are formed of upra^ised coral-rock. The several islands 
lyi]ig N.W. of Tongatabou, namely Anamouka, Komango, 
Kotou, Lefouga, Foa, &c., are seen m Captain Cook's chart 
to be fringed by reefs, and several of them are connected 
together. From the various statements in the first volume 
of Cook's Third Voyage, and especially in Chapters IV. and 
VI., it appears that these reefs are of coral, and certainly 
do not belong to the barrier class ; coloured red. — Toufoa 
and Kao, forming the western part of the group, according 
to Forster, have no reefs ; the former is an active volcano. 
— Vavao : there is a chart of this singularly-formed island, 
by Espinoza : according to Mr. Williams it consists of 
coral-rock : the Chevalier Dillon informs me that it is not 
fi-inc;cd ; not coloured. Nor are the islands of Lattc and 


Amargura colonrod, for I have not seen plans of them on a 
large scale, and I do not know whether they are fringed : 
Amargura is said (Athenreum, 1848, p. 40) to have been 
lately in violent eruption. 

Niouha, 16° S., 174° W., or Kejypel Island of Wallis, or 
Cocos Island : from a view and chart of this island, given 
in WaUis's Voyage, (4to edit.) it is evidently encircled by 
a reef; coloured blue. It is, however, remarkable that 
Boscaioen Island, immediately adjohiing, has no reef of 
any kind ; uncoloured. 

Wallis Island, 13° S., 17G° W.: a chart and view of 
this island in Wallis's Voyage (4to edit.) shows that it is 
encircled.^ A view of it in the Naut. Mag. July 1833, p. 37G, 
shows the same fact. Nine islands, most of them high, 
are said in "Wilkes U.S. Exploring Expedition (vol. ii. 
p. 157) to be enclosed within the same reef, through 
which, it is asserted, ships can enter ; coloured blue. 

Alloi.[fatou,ov Horn Island, Ononafu, or Proi?/ Island,'' 
and Himter Islands, lie between the Navigator and Fidji 
groups. I can find no distinct accounts of them. 

EiDJi OB Feejee or Viti Geoup.^ — Until lately the 

• [Wallis Island is encircled. There are eleven islands, seven of 
which are on the outer reef. — Kote sent to Mr. Darwin by Lieut. 
Chas. Smith, H.M.S. Fawn.] 

- ['Horn Islands, comprising Fotuna andAlofa; each has a distinct 
fringing-reef. Fotuna is about 2,500 feet and Alofa 1,200 feet high, 
I can give no infoi-mation regarding the depth of water, except that 
there is a deep ship-channel between the islands, no soundings being 
obtainable with the hand-line. The channel is hardly a mile broad. 
Nina-fu, or Good Hope Island, which I presume to be the same as 
that call(!d Onouafu in Coral Eeefs, is entirely volcanic, and has no 
reef whatever.' — Note sent to Mr. Darwin by Lieut. Chas. Smith, 
H.M.S. Faim.'] 

* [This group contains every description of reef. — Capt. Wharton. 
Makata has a central volcanic peak, according to the Narrative of Chal- 
lenger ^oyo-rr^a (p. 487), and is surrounded by a barrier-reef, or one inter- 
mediate between that and a fringing-reef. The shore line suggests 


best chart of the numerous islands of this group was that 
in the atlas of the Astrolabe's Voyage ; but now the islands 
have been surveyed during the U.S. Exploring Expedition, 
and full information respecting them and the reefs has 
been given by Dana. Many of the islands are bold asd 
mountainous, and are surrounded by reefs, lying far from 
the land, and outside the ocean appears very deep. The 
Astrolabe sounded with 90 fathoms in several places about 
a mile from the reefs, and found no bottom. It is evident 
that the water within many of the encircling reefs is deep : 
as indeed I was formerly assured was the case by Dillon. 
Beyond the high and encircled islands there are numerous 
atoll-formed reefs. Hence the whole group has been 
coloured blue. In the S.E. part lies Batoa, or Turtla 
Island of Cook (Second Voyage, vol. ii. p. 23, and chart ; 
4to edit.), surrounded by a coral-reef, 'which in some 
places extends two miles from the shore ; ' within the reef 
the water appears to be deep, and outside it is unfathom- 
able ; coloured pale blue. At the distance of a few miles, 
Captain Cook {ibid. p. 24) found a circular coral-reef, four 
or five leagues in circuit, with deep water within ; ' in 
short, the bank wants only a few little islets to make it 
exactly like one of the half-drowned isles so often men- 
tioned,' — namely, atolls. South of Batoa Hes the high 
island of Ono, which appears in Bellingshausen's Atlas to 
be encircled ; as do some other small islands to the south ; 
coloured pale blue : near Quo, there is an annular reef, 
quite similar to the one just described in the words of 
Captain Cook ; coloured dark blue. 

subsidence, and the reef is breached opposite to the principal inlet on 
the land. At Ngaloa harbour, Kandava, the map of this part of the 
island suggests subsidence, but there are also indications of slight 
upheaval. There is a barrier-reef. The soundings 150 fathoms 
from the edge of the reef were 80 fathoms, the slope for the first 05 
fathoms from the shore being 1 in l.then 1 in 1-4, diminishing to 1 in 
2 till 300 fathoms was reached.] 


Botouviah, 13° S., 179° E.— From the chart in 
Duperrcy's Atlas, I tliought that this island was encircled, 
but the Chevalier Dillon assures me that the reef is only a 
shore or fringing one ; coloured red.' 

• Independence Island, 10° S., 179° E., is described by 
Ih. G. Bennett (United Service Journ. 1831, part ii. p. 197) 
as a low island of coral formation ; it is small, and does 
not appear to contain a lagoon, although an opening 
through the reef is referred to. A lagoon probably once 
existed, and has since been filled up ; left uncoloured. 

Ellice Group. — Oscar, Pcysier, and Ellice Islands 
are figured in Arrowsmith's Chart of the Pacific (corrected 
to 1832) as atolls, and are said to be very low ; blue.^ — 
Nederlandisch Island : I am greatly indebted to the kind- 
ness of Admiral Krusenstem for sending me the original 
documents concerning this island. From the plans given 
by Captains Eeg and Khremtshenko, and from the detailed 
account given by the former, it appears tliat it is a narrow 
coral-island, about two miles long, containing a small 
lagoon. The sea is very deep close to the shore, which is 
fronted by sharp coral-rocks. Captain Eeg compares the 
lagoon with that of other coral-islands ; and he distinctly 
says, the land is ' very low.' I have therefore coloured it 
blue. Admiral Krusenstern (Supplement au Recueil des 
Memoires Hydrographiques publics en 182G et 1827, &c. &c. 
St. Petersburg, 1835) states that its shores are 80 feet high ; 
this probably arose from the height of the cocoa-nut trees, 
with which it is covered, being mistaken for land. — Grand 
Cocal is said in Kruscnstern's j\Icmoir to be low, and to be 
sm-rounded by a reef; it is small, and therefore probably 

' [' There is an extinct volcano on the island.' — Note written in 
Mr. Darwin's copy of this work.] 

■■' [' Mitchell Island, to the south of the Ellice group, is a very low 
atoll with about ten small islands on the reef. We were unable to 
discover any entrance to the lagoon.' — Lieut. Chas. Smith, H.M.S. 


once contained a lagoon; uncoloured.' — St. Augnstin'. 
from a chart and view of it, given in the atlas of the 
Coquille^s Voyage, it appears to be a small atoll, with its 
lagoon partly filled up ; coloured blue. 

GiLBEKT Group. — The chart of this group, given in the 
atlas of the Coquille's Voyage, shows that it is composed 
of ten well-characterized, but very irregularly shaped atolls. 
In D'Urville and Lottin's chart, Sydenham is written with 
a capital letter, signifying that it is high ; but this certainly 
is not the case, for it is a perfectly characterized atoll, and 
a sketch, showing how low it is, is given in the Coquille's 
atlas. Some narrow strip-like reefs project from the south- 
ern side of Drummond atoll, and render it irregular. The 
southern island of the group is called Cliase (in some charts, 
BotcJies) ; of this I can find no account, but Mr. F. D. Ben- 
nett discovered (Geograph. Journ. vol. vii. p. 229) a low 
extensive island in nearly the same latitude, about three 
degrees westward of the longitude assigned to Eotches ; 
and this probably is the same island. Mr. Bennett informs 
me that the man at the masthead reported an appearance 
of lagoon- water in the centre ; and, therefore, considering its 
position, I have coloured it blue. — Pitt Island, at the ex- 
treme northern point of the group, is left uncoloured, as 
neither its exact position nor nature is known.— i?7/ro;i 
Island, which lies a little to the eastward, does not appear 
to have been visited since Commodore Byron's voyage, and 
it was then seen only from a distance of 18 miles : it is said 
to be low ; uncoloured. 

Ocean, Pleasant, and Atlantic Islands all lie considerably 
to the west of the Gilbert group : I have been unable to find 
any distinct account of them. Ocean Island is written with 

' [' Grand Cocal was searched for in vain by H.M.S. Basilisk, and as 
all the local traders deny its existence, I cannot think it exists. It 
has long been marked ' doubtful ' on the Admiralty charts, and tlie 
description leads me to suppose the island reiDorted to have been St. 
A.ugustin.' — Lieut. Chas. Smith, H.M.S. Fawn,] 

218 ArrENDix. 

small letters in the French chart, but in Krusenstem'8 
Memoir it is said to be high. 

JMAnsnALii Giioup. — We are well acqiiaintcd with thia 
group from the excellent charts of the separate islands, made 
diu'ing the two voyages of Kolzebue : a reduced one of the 
wliole group may be seen in Krusenstern's Atlas, and in 
Kotzebue's Second Voyage. The group consists (with the 
exception of two little islands which probably have had their 
lagoon filled up) of a double row of 23 large and well-cha- 
racterized atolls, from the examination of which Chamisso 
drew up his well-known account of coral formations. I in- 
clude in this group Gasioar Bico, or Cormvallislsland, which 
is described by Chamisso (Kotzebue's First Voyage, vol. iii. 
p. 179) ' as a low sickle-formed group, with mould only on 
the windward side.' Gaspard Island is considered by some 
geographers as a distinct island lying N.E. of the group, 
hut it is not entered in the chart by Krusonstern ; left un- 
coloured. In the S.W. part of this group lies Barincj Island, 
of which little is Imowai (see Krusenstern's Appendix, 1835, 
p. 149). I have left it uncoloured ; but Boston Island I 
have coloured blue, as it is described (ibid.) as consisting of 
14 small islands, which, no doubt, inclose a lagoon, as re- 
presented in a chart in the Coquille's atlas. — Three islands, 
Aur, Kmren and Gaspar Bico, are written in the French 
chart with capital letters ; but this is an error, for from the 
account given by Chamisso in Kotzebue's First Voyage, they 
are certainly low. The nature, position, and even existence 
of the shoals and small islands north of the Marshall group 
are doubtful. 

New Hebrides. — Any chart, on even a small scale, of 
these islands will show that their shores are almost without 
reefs,' presenting a remarkable contrast with those of Kew 

• [The New Hebrides have fringing-reefs in various parts. No 
barrier-reefs are yet known, but the charts are still very imperfect. 
— Capt. Wharton.] 


Caledonia on tlie one band, and the Fidji group on the other. 
Nevertheless, I have been assured by Mr. Gr. Bennett, tliat 
coral grows vigorously on their shores ; as, indeed, will be 
further shown in some of the following notices. As, there- 
fore, these islands are not encircled, and as coral grow? 
vigorously on their shores, we might almost conclude, with- 
out further evidence, that they are fringed, and hence I 
have applied the red colour with rather greater freedom 
than in other instances. — Matthew's Kock, an active volcano, 
some way south of the group (of which a plan is given in 
atlas of the Astrolaho's Voyage) docs not appear to have 
reefs of any kind about it. — Annatom, the southernmost of 
the Hebrides : from a rough woodcut given in the United 
Service Journal (I80I, part iii. p. 190), accompanying a 
paper by Mr. Bennett, it appears that the shore is fringed ; 
coloured red. — Tanna : Forster, in his Observations (p. 22), 
says Tanna has on its shores coral-rock and madrepores ; 
and the younger Forster, in his account (vol. ii. p. 2G9), 
speaking of the harbour, says the whole S.E. side consists 
of coral-reefs, which are overflowed at high water : part of 
the southern shore in Cook's chart is represented as fringed ; 
coloured red. — Immcr is described (United Service Journ. 
1831, part iii. p. 192) by Mr. Bennett as being of moderate 
elevation, with cliffs appearing like sandstone ; coral grows 
in patches on its shore, but I have not coloured it ; and 
I mention these facts because Immer might have been 
thought, from Forster's classification (Observations, p. 14), 
to have been a low island, or even an atoll. — Erromango 
Island : Cook (Second Voyage, vol. ii. p. 45, 4to edit.) 
speaks of rocks everywhere lining the coast, and tbc natives 
offered to haul his boat over the breakers to the sandy 
beach : Mr. Bennett, in a letter to the editor of the Singa- 
pore Chron., alludes to the reefs on its shores. It may, I 
think, be safely inferred from these passages that the shore 
is frhiged in parts by coral-reefs ; coloured red. — Sandivich 


Island : tlie east coast is said (Cook's Second Voyage, vol. ii. 
p. 41) to be low, and to be guarded by a chain of breakers. 
In the accompanying chart it is seen to be fringed by a reef; 
coloured red. — Mallicollo : Forster speaks of the reef- 
bounded shore : the reef is about 30 yards wide, and so 
sliallow that a boat cannot pass over it. Forster, also, (Ob- 
servat. p. 23,) says that the rocks of the sea-shore consist 
of madrepore. In tlie plan of Sandwich harbour, the head- 
lands are represented as fringed; coloured red. — Aurora 
and Pentecost Islands, according to Bougainville, apparently 
have no reefs ; nor has the large island of S. Esjnritu, nor 
Bligh Island, nor Batiks Islands, ^ which latter he to the 
N.E. of the Hebrides. But in none of these cases have I 
met with any detailed account of their shores, or seen plans 
on a large scale ; and it will be evident that a fringing-rcef 
of only thirty, or even a few hundred yards in width is of 
so little importance to navigation, that it will seldom be 
noticed, excepting by chance ; and hence I do not doubt 
that several of these islands, now left uncoloured, ought 
to be red. 

Santa-Cruz Guo-UF.—Vwiikoro (Fig. 1, Plate I.) offers 
a striking example of a barrier-reef : it was first described 
by the Chevalier Dillon, in his Voyage, and was surveyed 
in the Astrolabe ; coloured pale blue. — TikopianiidFataka 
Islands appear, from the descriptions of Dillon andD'Urville, 
to have no reefs : Anouda is a low, flat island, surrounded 
by cliffs, (Astrolabe, Hydrog. and Krusenstern Mem. vol. ii. 
p. 4o2) ; these are uncoloured. — Touiwua (Otooboa of 
Dillon) is stated by Captain Tromelin (Annales Marit. 1829, 
p. 289) to be almost entirely included in a reef, lying at the 
distance of two miles from the shore. There is a space of 
three miles without any reef, which, although indented 
•with bays, offers no anchorage from the extreme depth of 
the water close to the shore. Captain Dillon also speaks 
» [Banks Islands are fringed in parts.— Capt. Wharton.] 


of the reefs fronting this island ; coloured blue.' — Santa. 
Cruz : I have carefully examined the works of Carteret, 
Dentrecasteaux, Wilson, and Tromelin, and I cannot dis- 
cover any mention of reefs on its shores ; left uncoloured. 
— Tinakoro is a constantly active volcano without reefs. — ■ 
Mendana Isles (mentioned by Dillon under the name of 
Mavtmec, &c,) are said by Krusenstern to be low and 
intertwined with reefs. I do not believe they include a 
lagoon ; I have left them uncoloured. — Duff's Islands com- 
pose a small group directed in a N.W. and S.E. band ; they 
are described by Wilson (p. 296, Miss. Voy. 4to edit.) as 
formed by bold peaked land, with the islands surrounded 
by coral-reefs, extending about half a mile from the shore : 
at the distance of a mile from the reefs he found only seven 
fathoms. As I have no reason for supposing there is deep 
water within thesereefs, I have coloured them red. — Kennedy 
Island, N.E. of Duff's : I have been unable to find any 
account of it. 

New Caledonia. — The great barrier-reefs on the shores 
of this island have already been described (Fig. 5, Plate II.). 
They have been visited by Labillardiere, Cook, and the 
northern point by D'Urville ; this latter part so closely 
resembles an atoll. that I have coloured it dark blue. The 
Loyalty group is situated to the east of New Caledonia; 
some at least of the islands are formed of upraised coral- 
rock, and are fringed with living reefs ; see Rev. W. B. 
Clarke, in Journal of Geolog. Soc. 1847, p. 61 ; coloured 
red. North of this group there are some extensive low 
reefs (called Astrolabe and BeauprS"^), which do not seem 
to be atoll-formed : these are left uncoloured. 

> [This island has a barrier-reef, with a 4-f athom channel through 
It, which leads into a harbour in the island itself. There is also 
deep but uneven water generally inside thereof.— Lieut. Chas. Smith, 
H.M.S. Fmvn.] 

- [This is sn atoll.- Capt. Tfliarlon.] 



Australian Barkikr-Reep. — This great reef, wliiclihas 
already been described, has been coloured from the charts 
of Flinders and King. Jukes has given many details re- 
specting it in the Voyage of H.M.S. Fly (vol. i. 1847, chap, 
xiii.). In the northern parts, an atoll-formed reef, lying 
outside the barrier, has been described by Bligh, and is 
coloured dark blue. In the space between Australia and 
New Caledonia, called by Flinders the Corallian Sea, there 
are numerous reefs. Of these, some are represented in 
Krusenstern's Atlas as having an atoll-like structure ; ' 
namely, BamiHon Shoal, Frederic, Vine or Horse-shoe, 
and Alert Reefs ; these have been coloured dark blue. 

LouisiADE. — The dangerous reefs which front and 
surround the western, southern, and northern coasts of this 
so-called peninsula and archipelago, seem evidently to 
belong to the barrier class. The land is lofty, with a low 
fringe on the coast ; the reefs are distant, and the sea out- 
side them profoundly deep. Nearly all that is known of 
this group is derived from the labours of Dentrecasteaux 
and Bougainville : the latter has represented one continuous 
reef 90 miles long, parallel to the shore, and in places as 
much as 10 miles from it ; coloured pale blue. A little 
distance northward we have the Laugldan Islands, the 
reefs round which are engraved in the atlas of the Voyage 
of the Astrolabe, in the same manner as round the encircled 
islands of the Caroline Archipelago : the reef is, in parts, a 
mile and a half from the shore, to which it does not appear 
to be attached ; coloured blue. At some little distance from 
the extremity of the Louisiade lies Wells Reef, described 
in G. Hamilton's Voyage in H.M.S. Pandora (p. 100) : 
it is said, * We found we had got embayed in a double 
reef, which will soon be an island.' As this statement is 
only intelligible on the supposition of the reef being crescent 

' [There are many atolls in this sea. — Capt. Wharton.] 


or horse-shoe formed, like so many other submerged amiular 
reefs, I have ventured to colour it blue. 

Saloman Akchipelago. — The chart in Krusenstern's 
Atlas shows that these islands are not encircled ; and as 
coral appears, from the works of Surville, Bougainville, and 
Labillardiere, to grow on their shores, this circumstance, 
as in the case of the New Hebrides, is a presumption that 
they are fringed. I cannot find out anything from Dentre- 
casteaux's Voyage, regarding the southern islands of the 
group, so have left them uncoloured. — Malayta Island, in 
a rough MS. chart in the Admiralty, has its northern shore 
fringed. — Ysahel Island : the N.E. part of this island, as 
shown in the same chart, is also fringed : Mendana (Burney, 
vol. i. p. 280), speaking of an islet adjoining the northern 
coast, says it is surrounded by reefs : the shores, also, of 
Port Praslin appear regularly fringed. Choiseul Island : 
parts of the shores are fringed by coral-reefs, in Bougain- 
ville's chart of Choiseul Bay. — Bougainville Island : accord- 
ing to Dentrecasteaux, the western shore abounds with 
coral-reefs, and the smaller islands are said to be attached 
to the larger ones by reefs ; all the above-mentioned islands 
have been coloured red. — iiOiiA;a Islands : Captain Duperrey 
has kmdly informed me in a letter that he passed close 
round the northern side of this island (of which a plan is 
given in his atlas of the Coqttille's Voyage), and that it 
was ' garnie d'une bande de recifs a fleur d'eau adherentes 
au rivage ; ' and he infers, from the abundance of coral on 
the islands north and south of Bouka, that the reef pro- 
bably is of coral ; coloured red.' 

Off the north coast of the Saloman Archipelago there 
are several small groups which are little known : they 
appear to be low, and of coral formation ; and some of 
them probably have an atoll-like structure : the Chevalier 

' [Bouka, according to the best accounts, has a barrier-reef, but 
our information is still imperfect. — Capt. Wharton.] 

224 ArrEXDix. 

Dillon, however, informs me this is not the case with the 
Baxos de Canddaria.^ — Outong Java, according to the 
Spanish navigator, Maurelle, is thus characterized ; but this 
is the only one which I have ventured to colour blue. 

New Ireland. — The shores of the S.W. point of this 
island and some adjoining islets, are fringed by reefs, as 
may be seen in the atlases of the Voyages of the Coquilla 
and Astrolabe. M. Lesson observes that the reefs are open 
in front of each streamlet. The Duka of York's Island is 
also fringed ; but with regard to the other parts of Ncio 
Ireland, New Hanover, and the small islands lying north- 
ward, I have been unable to obtain any information. I 
will only add that no part of New Ireland appears to be 
fronted by distant reefs. I have coloured red only the 
above specified portions. 

New Britain and the Northern Shore of New 
Guinea. — From the charts in the Voyage of i]\Q Astrolabe, 
and from the Hydrog. Memoir, it appears that these coasts 
are entirely without reefs, as are the Scliouten Islands, 
lying close to the northern shore of New Guinea. The 
western and south-western parts of New Guinea will be 
treated of when we come to the islands of the East Indian 

Admiralty Group.^ — From the accounts given by 
Bougainville, Maurelle, Dentrecasteaux, and the scattered 

' [This is a perfect atoll. — Capt. Wharton.] 

2 [Narrative of CJmllenger Voyage, p. 699. Admiralty, or Bosco 
Islands. The main island rises to nearly 3,000 feet. The coast is low 
and indented with deep bays. There are many coral-reefs off the coast 
at varying distances, not forming a connected barier-reef. There is 
convenient anchorage within the reef, the soundings in the deeper part 
of the channel at Nares Harbour being from 25 to 34 fathoms, and very 
generally nearly or over 20 fathoms. Four other of the islands attain 
an elevation of from 600 to 800 feet ; the remainder are low and are 
situated on coral-reefs. The coast line of the main island is a plat- 
form of coral-sand rock, and the low outlying islands are tlie same, 
but tlie hills are presumed to be of volcanic rock.] 


notices collected by Horsburgli, it appears that some of the 
many islands composing it are high, with a bold outline ; 
and others are low, small, and interlaced with reefs. All 
the high islands appear to be fronted by distant reefs rising 
abruptly from the sea, and within some of which, there is 
reason to believe that the water is deep. I have therefore 
little doubt that they belong to the barrier class. In the 
southern part of the group, we have Elisabeth Island, 
which is surrounded by a reef at the distance of a mile ; 
and two miles eastward of it (Krusenstern, Append. 1835, 
p. 42) there is a little island containing a lagoon. Near 
here, also, lies Circular Keef (Horsburgli Direct, vol. ii. 
p. 796, 8th edit.), * three or four miles in diameter, having 
deep water inside with an opening at the N.N.W. part : the 
reef on the outside is steep to.' I have from these data, 
coloured the group pale blue, and Circular Keef dark blue. 
— The Anachorites, Ecliequier, and Herviites consist of in- 
numerable low islands of coral formation, which probably 
are atolls ; but not being able to ascertain this, I have not 
coloured them, nor Durour Island, which is described by 
Carteret as low. 

The Caeoline Aechipelago is now well known, chiefly 
from the hydrographical labours of Lutke : it contains 
about forty groups of atolls, and three encircled islands, 
two of which are engraved in Figs. 2 and 7, Plate I. 
Commencing with the eastern part, the encircling reef 
round Ualan appears to be only about half a mile from the 
shore ; but as the land is low, and covered with mangroves 
(Voyage autour du Monde, par F. Lutke, vol. i. p. 339), 
its margin has not probably been ascertained. The extreme 
depth in one of the harbours within the reef is 33 fathoms 
(see charts m Atlas of CoquilWs Voyage), and outside at 
half a mile distance from the reef, no bottom was obtained 
with 250 fathoms. The reef is surmounted by many 
islets, and the lagoon-like channel within is mostly shallow, 


and appears to have been much encroached on by the low 
land surrounding the central mountains ; these facts show 
that time has allowed much detritus to accumulate ; 
coloured pale blue. — PouynipMc or Seniavine. In the greater 
part of the circumference of this island, the reef is about 
one mile and three quarters from the shore ; but on the north 
side it is five miles distant from the included high islets. 
The reef is broken in several places ; and just within it, 
the depth in one place is 30 fathoms, and in another, 28, 
beyond which, to all appearance, there was ' un port vaste 
et sur ' (Lutke, vol. ii. p. 4). Coloured pale blue. — Hogoleu 
or Bong. This wonderful group contains at least 62 islands, 
and its reef is 135 miles in circuit. Of the islands, only a 
few, about six or eight (see Hydrog. Description, p. 428, of 
the Voyage of the Astrolabe, and the large accompanying 
chart taken chiefly from that given by Duperrey) are high, 
and the rest are all small, low, and formed on the reef. 
The depth of the great interior lake has not been ascer- 
tained ; but Captain D'Urville appears to have entertained 
no doubt about the possibility of taking in a frigate. The 
reef lies no less than 14 miles distant fi'om the northern 
coasts of the interior high islands ; seven miles from their 
western sides, and 20 from the southern : the sea is deep 
outside. This island resembles on a grand scale the 
Gambler group in the Low Archipelago. Of the low ' 
islands forming the chief part of the Caroline Archipelago, 
all those of larger size (as may be seen in the Atlas by 
Captain Lutke), and some even of the small ones of which 
plans are given in the Atlas of the Coqidlle's Voyage, are 
true atolls. There are, however, some low, small islands 
of coral formation, namely, Ollap, Tamatam, Bigall, 
Satahoual, which do not contain lagoons ; but it is probable 
that lagoons originally existed, but have since filled up : 

' In D'Ui-ville and Lottin's chart, Pescrare is wn-itten with capital 
letters ; but this evidently is an error, for it is one of the low islets on 
the reef of Namonouyto (see Lutk^'s charts), which is a regular atoll. 


Lutke (vol. ii. p. 304) seems to have thought that all the low 
islands, with only one exception, contained lagoons. The 
most southern island in the group, namely, Piguiram, is not 
coloured, because I have found no accomrt of it. Nongouor, 
or Monte Vcrdison, which was not visited by Lutke, is 
described and figured by IMr. Bennett (United Service 
Journal, Jan. 1832) as an atoll. All the before-mentioned 
islands have been coloured blue. It must, however, b^ 
stated that between Ualan and Pouynipete, the three 
McAskill Islands rise to a height of from 40 to 100 feet, 
and consist, according to Dana (Corals and Coral Islands, 
p. 30G), of coral-rock ; whether they are encircled or fringed 
by coral-reefs does not seem to be known. 

Westekn part of the Cakoline Archipelago. — Fais 
Island is 90 feet high, and is surrounded, as I have been in- 
formed by Admiral Lutke, by a narrow reef of living coral, 
of which the broadest part, as represented in the charts, is 
only 150 yards ; coloured red. — Pliilip Island, I believe, is 
low ; but Hunter, in his Historical Journal, gives no clear 
account of it ; uncoloured. Elivi : from the manner in 
which the islets on the reefs are engraved in the Atlas of 
the Aslrolabe's Voyage, I should have thought they were 
above the ordinary height ; but Admiral Lutke assures me 
that this is not the case : they form a regular atoll ; co- 
loured blue. Goiiap {Eap of Chamisso) is a high island 
with a reef (see Chart in Voyage of Astrolabe) in most parts 
more than a mile distant from the shore, and two miles in 
one part. Captain D'Urville thinks that there would be 
anchorage (Hydrog. Descript. Astrolabe Voyage, p. 430) 
for ships within the reef, if a passage could be found ; co- 
loured pale blue. — Goulou, from the chart in the Astrolabe's 
atlas, appears to be an atoll : D'Urville (Hydrog. Descript. 
p. 437) speaks of low islets on the reef ; coloured dark blue. 

Pelew Islands.^ — Krusenstern speaks of some of the 
islands being mountainous ; the reefs are distant from the 


eliorc, antl tlievo are spaces within them, not opposite to 
any valley, from 10 to 15 fathoms deep. According to a 
MS. chart of the group hy Lieut. Elmer in the Admiralty, 
there is a large space within the reef with decpish water : 
although the high land does not hold a central position 
with respect to the reefs, as is generally the case, I have 
little doubt that the reefs of the Pelew Islands ought to be 
^ranked in the barrier class, and I have coloured them pale 
blue. In Lieut. Elmer's chart there is a horse-shoe-formed 
shoal, 13 miles N.W. of Pelew, with 15 fathoms within the 
reef, and some dry banks on it ; coloured dark blue. — 
Spanish, Martires, Sanscrot, Pulo Anna and Mariere 
Islands are not coloured, because I know nothing about 
them, excepting that according to Krusenstern, the second, 
third, and fourth mentioned, are low, placed on coral-reefs, 
and therefore perhaps include a lagoon ; but Pulo Mariere is 
a little higher. Since the above remarks were written Prof. 
Semper has published an interesting article (Zeitschr. f. 
Wissensch. Zoologie, Bd. xiii. 18G3, p. 558) on these 
islands. He states that the southern islands consist of 
coral-rock, upraised to the height of from 400 to 500 feet ; 
and some of them, before their upheaval, seem to have ex- 
isted as atolls. They are now merely fringed by living reefs. 
The northern islands are volcanic, deeply indented by bays, 
and are fronted by barrier-reefs. To the north there are 
three true atolls. Prof. Semper doubts whether tlie whole 
group has subsided, partly from the fact of the southern 
islands being formed of upraised coral-rock ; but there seems 
to me no improbability in their having originally subsided* 
then having been upraised (probably at the time when the 
volcanic rocks to the north were erupted), and again having 
subsided. The existence of atolls and barrier-reefs inclose 
proximity is manifestly not opposed to my views. On the 
other hand, the presence of reefs fringing the southern 
islands is opposed to my views, as such reefs generally indi- 


cate that the land Las either long remained stationary, or 
has been upraised. It must, however, be borne in mind (as 
remarked in our sixth chapter) that when the land is pro- 
longed beneath the sea in an extremely steep slope, reefs 
formed there during subsidence will remain closely attached 
to the shore, and will be undistinguishable from fringing- 
reefs. Now we know that the submarine flanks of most atolls 
are very steep ; and if an atoll after upheaval and before the 
sea had eaten deeply into the land, and had formed a broad 
fiat surface, were again to subside, the reefs which grew to the 
surface during the subsiding movement, would still closely 
skirt the coast. After some hesitation, I have thought my- 
self justified in leaving these islands coloured blue. 

Mariana Archipelago, or Ladrones. — Guahan : 
almost the whole of this island is fringed by reefs, which 
extend in most parts about a third of a mile from the land. 
Even where the reefs are most extensive, the water within 
them is shallov/. In several parts there is a navigable 
channel for boats and canoes within the reefs. In Frey- 
cinet's Hydrog. Mem. there is an account of these reefs, 
and in the atlas, a map on a large scale ; coloured red. — • 
Bota : ' L'ile est presque entierement entouree de recifs ' 
(p. 212, Freycinet's Hydrog. Mem.). These reefs project 
about a quarter of a mile from the shore ; coloured red. — 
Tinian : the eastern coast is precipitous, and is without reefs ; 
but the western side is fringed like the last island ; coloured 
red. Say pan : the N.E . coast, and likewise the western shores 
appear to be fringed ; but there is a great, irregular, horn- 
like reef projecting far from this side ; coloured red.— 
Farallon de Medinilla appears so regularly and closely 
fringed in Freycinet's charts, that I have ventured to 
colour it red, although nothing is said about reefs in the 
Hydrographical Memoir. The several islands which form 
the northern part of the group are volcanic (with the excep- 
tion perhaps of Torres, which resembles in form the madre- 


pnritic island of Medinilla), and appear to be without 
reefs. — Mangs, however, is described (by Freycinet, p. 219, 
Ilydrog.) from some Spanish charts, as formed of small 
islands placed ' au milieu de nombreux recifs ; ' and as 
these reefs in the general chart of the group do not project 
so much as a mile ; and as there is no appearance from a 
double line, of the existence of deep water within, I have 
ventured, although with much hesitation, to colour thera 
red. Respecting Folger and Marshall Islands, which lie 
some way east of the Marianas, I can find out nothing, 
excepting that they are probably low. Kruscnstern says 
this of Marshall Island ; and Folger Island is written with 
small letters in D'Urville's chart ; uncoloured. 

BoNiN OR Aezobispo Gboup. — Peel Island has been 
examined by Captain Beechey, to whose kindness I am 
much indebted for giving me information regarding it : * at 
Port Lloyd there is a great deal of coral ; and the inner 
harbour is entirely formed by coral-reefs, which extend 
outside the port along the coast.' Captain Beechey, in 
another part of his letter to me, alludes to the reefs fring- 
ing the island in all directions ; but at the same time it 
must be observed that the surf washes the volcanic rocks 
of the coast in the greater part of its circumference. This 
island has certainly been elevated at least 50 feet within 
the recent period (see Journal of Geolog. Soc. 1855, p. 
532). I do not know whether the other islands of the 
archipelago are fringed ; I have coloured Peel island red. 
— Grampus Island, to the eastward, does not appear 
(Meare's Voyage, p. 95) to have any reefs, nor does 
Bosario Island (from Lutke's chart), which hes to the 
westward. Piespecting the few other islands in this part 
of the sea, namely the Sulphur Islands, with an active 
volcano, and those lying between Benin and Japan (situated 
near the extreme limit in latitude at which reel's can grow), 
I have not been able to find any clear account. 


West End op New Guinea.— Por^ Dory, from tlie 
charts iu the Voyage of the Coquille, it would appear that 
the coast in this part is fringed by coral-reefs ; M. Lesson, 
however, remarks that the corals are sickly ; coloured red. 
• — Waig'ou: a considerable portion of the northern shore 
of these islands are seen in the charts (on a large scale) 
in Freycinet's Atlas to be fringed by coral-reefs. Forrest 
(p. 21, Voyage to New Guinea) alludes to the coral-reefs 
lining the heads of Piapis Bay ; and Horsburgh (vol. ii. 
p. 599, 4th edit.), speaking of the islands in Dampier Strait, 
says, ' sharp coral-rocks line their shores ; ' coloured red. — 
In the sea north of these islands, we have Gucdcs (or 
Freewill, or St. David's), which from the chart given in 
the 4to edit, of Carteret's Voyage must be an atoll. 
Krusenstern says the islets are very low ; coloured blue. — - 
Carteret's Shoals, in 2° 53' N., are described as circular, 
with stony points showing all round, with deeper water in 
the middle ; coloured blue. — Aiott : the plan of this group, 
given in the atlas of the Voyage of the Astrolabe, shows 
that it is an atoll ; and, from a chart in Forrest's Voyage, 
it appears there is 12 fathoms within the circular reef; 
coloured blue. — The S.W. coast of New Guinea appears to 
be low, muddy, and devoid of reefs. The Amc, Timor- laut 
and Tenimher Groups have lately been examined by 
Captain Kolff, the MS. transla.tion of which, by Mr. W. 
Earl, I have been permitted to read, through the kindness 
of Captain Washington, E.N. These islands are mostly 
rather low, and are surrounded by distant reefs (the Ki 
Islands, however, are lofty, and, from Mr. Stanley's survey, 
appear without reefs) ; the sea in some parts is shallow, iii 
others profoundly deep, as near Larrat. From the imper- 
fection of the published charts, I have been unable to 
decide to which class these -reefs belong. From the dis- 
tance to which they extend from the land where the sea is 
very deep, I am strongly inclined to believe they ought to 


como wiihin tLe barrier class, and be coloured blue ; but I 
have been forced to leave them uncoloured. — The last- 
mentioned groups are connected with the east end of Ceram 
by a chain of small islands, of which the small groups of 
Ccram-laut, Goram, and Kcffing are surrounded by very 
extensive reefs, projecting into deep water, which, as in the 
last case, I strongly suspect belong to the barrier class ; 
but I have not coloured them. From the south side of 
Keffing, the reefs project five miles (Windsor Earl's Sailing 
Direct, for the Arafura Sea, p. 9). 

Ceram. — In various charts which I have examined, 
several parts of the coast are represented as fringed by 
reefs. — Manipa Island, between Ceram and Bourou, in an 
old MS. chart in the Admiralty, is fringed by a very irregu- 
lar reef, partly dry at low water, which I do not doubt is 
of coral formation ; both islands coloured red. — Bourou : 
parts of this island appear fringed by coral-reefs, namely, 
the eastern coast as seen in Freycinet's chart ; and Cajeli 
Bay, which is said by Horsburgh (vol. ii. p. 630) to be 
lined by coral-reefs, that stretch out a little way, and have 
only a few feet of water on them. In several charts, 
portions of the islands forming the Amboina Group ' are 
fringed by reefs ; for instance, Noessa, Harenca, and 
Ucaster, in Freycinet's charts. The above-mentioned 
islands have been coloured red, although the evidence is 
not very satisfactory. — North of Bourou the parallel line 
of the Xulla Isles extends : I have not been able to find 
out anything about them, excepting that Horsburgh (vol. 
ii. p. 543) says that the northern shore is surrounded by a 
reef at the distance of two or three miles ; uncoloured. — 
Mysol Group : the Kanary Islands are said by Forrest 
(Voyage, p. 130) to be divided from each other by deep 

' [At Amboina coral-reef rock occurs raised many hundred feet 
above sea level, forming a steep bill slope. Narrative of Challenger 
Voyage, vol. 1. p. 580. See also Moseley, Notes by a Naturalist, p. 389.] 


straits, and are lined with coral-rocks ; coloured red. — • 
Gnche, lying between Waigiou and Gilolo, is engraved as if 
fringed ; and it is said by Freycinet, that all the soundings 
under five fathoms were on coral; coloured red.— Gilolo : in 
a chart published by Dalrymple, the numerous islands on 
the western, southern {Batchian and the Strait of Patien- 
tia), and eastern sides appear fringed by narrow reefs ; 
these reefs, I suppose, are of coral, for it is said in Malte 
Brun (vol. xii. p. 15G), ' sur les cotes (of Batchian), comma 
dans la ])lupart des iles de cet archipel, il y a des rocs de 
madrepores d'une beaute et d'une variete infinies.' Forrest, 
also (p. 50), says Seland, near Batchian, is a little island 
with reefs of coral ; coloured red. — Morty Island (north of 
Gilolo) : Horsburgh (vol. ii, p. 506) says the northern coast 
is lined by reefs, projecting one or two miles, and having 
no soundings close to them ; I have left it uncoloured, 
although, as in some former cases, it ought probably to be 
pale blue. — Celebes. The western and northern coasts 
appear in the charts to be bold and without reefs. Near 
the extreme northern point, however, an islet in the 
Straits of Lvmhe, and part of the adjoining shore, appear 
to be fringed : the east side of the bay of Manado has deep 
water, and is fringed by sand and coral {Astrol. Voyage, 
Hydrog. Part, p. 453-4) ; this extreme point, therefore, I 
have coloured red. Captain Keppell, also, speaks (Expe- 
dition to Borneo, vol. i. p. 130) of the shore being in parts 
fringed with reefs ; he found upraised coral-reefs at the 
height of from 80 to 100 feet above the level of the sea. — 
Of the islands between the northern point of Celebes and 
the Philippines, I have not been able to find any account, 
except of Scrangani, which appears surrounded by narrow 
reefs; and Forrest (Voyage, p. 1G4) speaks of coral on 
its shores ; I have, therefore, coloured this island red. 
To the eastward of this chain lie several islands ; of which 
I cannot find any account, except of Karkalang, which ia 


said by Ilorsburgli (vol. ii. p. CO-l) to be lined by a dan- 
gerous reef, projecting several miles frour tbe nortliern 
shore ; not coloured. 

Islands near Timor. — Tbe account of tbe following 
islands is taken from Captain D. Kolff's Voyage in 182;3, 
translated by Mr. W. Earl from tbe Bnich.—Lcttc bas 
' reefs extending along sbore at tbe distance of balf a mile 
from tbe land.' — Moa bas reefs on tbe S.W. part. — Lakor 
bas a i*eef lining its sbore ; tbese islands are coloured red. 
— Still more eastward, Luan, differently fi'om tbe last- 
mentioned islands, bas an extensive reef ; it is steep out- 
side, and witbin tbere is a dcptb of 12 feet ; from tbese 
facts it is impossible to decide to wbicb class tins island 
belongs. — Kissa, off tbe point of Timor, bas its ' sbore 
fronted by a reef, steep too on tbe outer side, over wbicb 
small proabs can go at tbe time of bigli water ; ' coloured 
red. — Timor : most of tbe points, and some considerable 
spaces of tbe northern sbore, are seen in Freycinet's chart 
to be fringed by coral-reefs ; and mention is made of them 
in tbe accompanying Hydrog. Memoir ; coloured red. — ■ 
Savu, S.W. of Timor, appears in Hinders' chart to be 
tinged ; but I have not coloured it, as I do not know that 
the reefs are of coral. — Sandahvood Island bas, according 
to Horsburgb (vol. ii. p. 007), a reef on its southern shore, 
four miles distant from tbe land ; as the neighbouring sea 
is deep, and generally bold, this probably is a barrier-reef, 
but I have not ventured to colour it. 

N.W. Coast of Australia. — It appears, in Captain 
King's Sailing Directions (Narrative of Survey, vol. ii. 
pp. 325 to 369), that tbere are many extensive coral-reefs 
skirting, often at considerable distances, tbe N.W. shores 
and encompassing tbe small adjoining islets. Deep water 
in no instance is represented in tbe charts between tbese 
reefs and the land ; and, therefore, they probably belong 
ko tbe fringing class. But as they extend far into tbe sea, 


wLicli is generally shallow, even in places where the land 
seems to be somewhat precipitous, I have not coloured 
them. Houtman's Abrolhos (lat. 28° S. on west coast) 
have lately been surveyed by Captain Wickham (as described 
in Naut. Mag. 1841, p. 511) : they he on the edge of a 
steeply-shelving bank, which extends about 30 miles sea- 
ward, along the whole hne of coast. The two southern 
reefs, or islands, enclose a lagoon-like space of water, 
varying in depth from 6 to 15 fathoms, and in one spot 
with 23 fathoms. The greater part of the land has been 
formed on their inland sides, by the accumulation of frag- 
ments of corals ; the seaward face consisting of nearly bare 
ledges of rock. Some of the specimens, brought home by 
Captain Wickliam, contained fragments of marine shells, 
but others did not ; and these closely resembled a formation 
at King George's Sound, principally due to the action of 
the wind on calcareous dust, which I have described in my 
work on Volcanic Islands. From the extreme irregularity 
of these reefs with their lagoons, and from their position 
on a bank, the usual depth of which is only 30 fathoms, 
I have not ventured to class them with atolls, and hence 
have left them uncoloured. — Boiohy Shoals : these lie 
some way from the N.W. coast of Australia : according to 
Captain King (Narrative of Survey, vol. i. p. GO), they are 
of coral-formation. They rise abruptly from the sea, and 
Captain King found no bottom with 170 fathoms close to 
them. Three of them are crescent-shaped ; a third oval 
reef of the same group is entirely submerged (Lyell, 
Principles of Geolog., book iii. chap, xviii.) ; ' coloured blue. 
— Scott's Eeefs, lying north of Kowley Shoals, are briefly 
described by Captain Wickham (Naut. Mag., 1841, p. 440) 
as of great size, of a circular form, and ' with smooth 
water within, forming probably a lagoon of great extent.' 
There is a break on the western side, where there probably 
' [Book iii. ch. xlix. 11th edition.] 


is an entrance : the water ia very deep off tliese reefs ; 
coloured blue. 

Proceeding westward along the great volcanic chain of 
the East Indian or Malay Archipelngo, Solor Strait is 
represented as fringed in a chart published by Dalrymple 
from a Dutch IMS. ; as are parts of Floras, Adenara, and 
Solor. Horsburgh speaks of coral growing on these shores, 
and therefore I have no doubt that the reefs are of coral, 
and have coloured them red. We hear from Horsburgh 
(vol. ii. p. 602) that a coral flat bounds the shores of Sapj/ 
Bay. From the same authority it appears (p. GIO) that 
reefs fringe the island of Timor-Young, on the N. shore of 
Sumbawa ; and likewise (p. GOO) that Bally town in 
Lomhoch, is fronted by a reef, stretching along the shore 
at the distance of a hundred fathoms, with channels 
through it for boats ; these places, therefore, have been 
coloured red. — Bally Island : in a Dutch MS. chart on a 
large scale of Java, which was brought from that island 
by Dr. Horsfield, who had the kindness to show it me at 
the India House, its western, nortliern, and southern shores 
appear regularly fringed by a reef (see also Horsburgh, 
vol. ii. p. 593) ; and as coral is found abundantly there, I 
have no doubt that the reef is of coral, and therefore have 
coloured it red. 

Java. — My information regarding the reefs of this great 
island is derived from the chart just mentioned. The greater 
part of Madura is represented in it as regularly fringed, 
and Hkewise portions of the coast of Java immediately south 
of it. Dr. Horsfield informs me that coral is very abundant 
near Sourahaya. The adjoining islets, and parts of the N. 
coast of Java, west of Point Buang, or Japara, are fringed 
by reefs, said to be of coral. Lubech, or Bavian Islands, 
lying at some distance from the shore of Java, are regularly 
fringed by coral-reefs : Gurimon Java appears equally so, 
though it is not directly said that the reefs are of coral ; 


there is a depth of between 30 and 40 fathoms round these 
islands. Parts of the shore of Sunda Straits, wliere the 
water is from 40 to 80 fathoms deep, and the islets near 
Batavia appear in several charts to be fringed. In the 
Dutch chart the southern shore, in the narrowest part of 
the island, is in two places fringed by reefs of coral. West 
oi Sego.rrowodee Bay, and the extreme S.E. and E. shores 
are likewise fringed by coral-reefs ; all the above-mentioned 
places coloured red. 

Macassar Strait : the east coast of Borneo appears, in 
most parts, free from reefs, and where they occur, as on the 
coast of Pamaroong, the sea is very shallow ; hence no part 
is coloured. In Macassar Strait itself, in about lat. 2° S., 
there are many small islands with coral shoals pro- 
jecting far from them. There are also (old charts by Dal- 
rymple) numerous little flats of coral, not rising to the 
surface of the water, and shelving suddenly from five 
fathoms to no bottom with 50 fathoms ; they do not appear 
to have a lagoon-like structure. There are similar coral- 
shoals a little farther south ; and in lat. 4° 55' there are 
two, which are engraved from modern surveys, in a manner 
which may represent an annular reef with deep water 
inside : Capt. Moresby, however, who was formerly in this 
sea, doubts this fact, so that I have left them uncoloured : 
at the same time I may remark, that these two shoals make 
a nearer approach to the atoll-like structure than any other 
within the E. Indian Archipelago. Southward of these 
shoals there are other low islands and irregular coral-reefs ; 
and in the space of sea, north of the great volcanic chain, 
from Timor to Java, we have other islands, such as the 
Postillions, Kalatoa, ToJcan-Bessoes, &c., which are chiefly 
low, and are surrounded by very irregular and distant reefs. 
From the imperfect charts I have seen, I have not been 
able to decide whether they belong to the atoll or barrier 
class, or whether they merely fringe submarine banks, 
3 7 


and gently sloping land. In the Bay of Bonin, between 
the two southern arms of Celebes, there are numerous coral- 
reefs ; but none of them seem to have an atoll-like structure. 
I have, therefore, not coloured any of the islands in this 
part of the sea ; I think it, however, exceedingly probable 
that some of them ought to be blue. I may add that there 
is a harbour on the S.E. coast of Bouton, which, according 
to an old chart, is formed by a reef, parallel to the shore, 
with deep water within ; and in the Voyage of the Coquille, 
some neighbouring islands are represented with distant 
reefs, but I do not know whether with deep water within. 
I have not thought the evidence sufficient to permit me to 
colour them. 

Sumatra. — Commencing with the west coast and out- 
lying islands ; Engano Island is represented in the published 
chart as surrounded by a narrow reef, and Napier, in his 
Sailing Directions, speaks of the reef being of coral (also 
Horsburgh, vol. ii. p. 115) ; coloured red. Bat Island (3° 
51' S.) is surrounded by reefs of coral, partly dry at low 
water (Horsburgh, vol. ii. p. m).— Trieste Island (4° 2' S.) : 
the shore is represented, in a chart which I saw at the India 
House, as fringed in such a manner, that I feel sure the 
fringe consists of coral ; but as the island is so low that the 
sea sometimes flows quite over it (Dampier, Voyage, vol. i. 
p. 474) I have not coloured it. — Bulo Dooa (lat. 3°) : it is 
said in an old chart that there are chasms in the reef round 
the island, admitting boats to the watering-place, and that 
the southern islet consists of a mass of sand and coral. — 
Bulo Bioang : Horsburgh (vol. ii. p. 8G) says that the rocky 
coral-bank, which stretches about 40 yards from the shore, 
IS steep all round : in a chart, also, which I have seen, tha 
island is represented as regularly fringed. — Btdo Mintao is 
lined with reefs on its west side (Horsburgh, vol. ii. p. 107). 
— BuloBaniah : the same authority (vol. ii. p. 105), speaking 
of a part, says it is faced with coral-rocks. — Minguin (3" 


86' N.) ; a ccralreef fronts this place, and projects into the 
sea nearly a quarter of a mile (Notices of the Indian Archi- 
pelago, published at Singapore, p. 105). — Pulo Brassa (5° 
46' N.) : a reef surrounds it at a cable's length (Horsburgh, 
vol. ii. p. 60). I have coloured all the above specified points 
red. I may here add, that both Horsburgh and Mr. Moor 
(in the Notices just alluded to) frequently speak of the 
numerous reefs and banks of coral on the west coast of 
Sumatra ; but they nowhere have the structure of a barrier- 
reef, and ]\Iarsden (History of Sumatra) states that where 
the coast is flat, the fringing-reefs extend far from it. The 
northern and southern points, and the greater part of the 
east coast, are low, and faced with mud banks, and there- 
fore without coral. 

NicoBAK Islands. — The chart represents the islands 
of this group as fringed by reefs. With regard to Great 
Nicobar, Captain Moresby informs me that it is fringed by 
reefs of coral, extending between 200 and 300 yards from 
the shore. The Northern Nicobars appear so regularly 
fringed in the published charts, that I have no doubt the 
reefs are of coral. This group, therefore, is coloured red. 

Andaman Islands. — From an examination of the MS. 
chart, on a large scale, of these islands, by Captain Arch. 
Blair, in the Admiralty, several portions of the coast appear 
fringed; and as Horsburgh speaks of coral-reefs being 
numerous in the vicinity of these islands, I should have 
coloured them red, had not some expressions in a paper in 
the Asiatic Kesearches (vol. iv. p. 402) led me to doubt the 
existence of reefs ; uncoloured. 

The coast of Malacca, Tanasserim, and the coasts 
northward, appear in the greater part to be low and muddy : 
where reefs occur, as in parts of Malacca Straits, and near 
Singapore, they are of the fringing kind ; but the water ia 
Bo shoal, that I have not coloured them. In the sea, how- 
ever, between Malacca and the west coast of Borneo, where 


there is a greater depth from 40 to 50 fathoms, I have 
coloured red some of the groups, which are regularly 
fringed. The northern Natunas and the Anavihas Islands 
are represented in the charts on a large scale, published in 
the atlas of the Voyage of the Favourite [by La Place, 1831]» 
as fringed by reefs of coral, with very shoal water within them. 
Tumhelan and Bwioa Islands {1" N.) are represented in the 
English charts as surrounded by a very regular fringe. — 
St. Barhes (0^ 15' N.) is said by Horsburgh (vol. ii. p. 279) 
to be fronted by a reef, over which boats can land only at 
high water.— The shore of Borneo, at Tunjong Apec, is also 
fronted by a reef, extending not far from the land (Hors- 
burgh, vol. ii. p. 468). These places I have coloured red ; 
although with some hesitation, as the water is shallow. I 
might perhaps have added Pulo Leat, in Gaspar Strait, 
Lucepara and Carimata ; but as the sea is conlBned and 
shallow, and the reefs not very regular, I have left them 

The water deepens very gradually from the whole west 
coast of Borneo ; and I cannot make out that it has any 
reefs of coral. The islands, however, off the northern 
extremity, and near the S.W. end of Palawan, are fringed 
by very distant coral reefs : thus the reefs off Balabac are 
no less than five miles from the land ; but the sea, in the 
whole of this district, is so shallow, that the reefs might be 
expected to extend very far from the land. I have not, 
therefore, thought myself authorized to colour them. The 
N.E. point of Borneo, where the water is very shoal, is 
connected with ]\Iagindanao by a chain of islands called 
the Sooloo Archipelago, about which I have been able to 
obtain very little information ; Pangootaran, although ten 
miles long, entirely consists of a bed of coral-rock (Notices 
of E. Indian Arch. p. 58) : I believe from Horsburgh that 
the island is low ; not coloured. — Tahow Bank, in some 
old charts, appears Hke a submerged atoll ; not coloured. 


Forrest (Voyage, p. 21) states that one of tlie islands near 
Sooloo is surrounded by coral-rocks ; but there is no dis- 
tant reef. Near the S. end of Bassclan, some of the islets 
ill the chart accompanying Forrest's Voyage appear fringed 
with reefs ; hence I have coloured, though unwillingly, parts 
of the Sooloo group red. The sea between Sooloo and 
Palawan, near the shoal coast of Borneo, is interspersed with 
irregular reefs and shoal patches ; not coloured : but in the 
northern part of this sea there are two low islets, Cagayanes 
and Gavilli, surrounded by extensive coral-reefs ; the breakers 
round the latter (Horsburgh, vol. ii.p. 513) extend five or six 
miles from a sand-bank, which forms the only dry part ; these 
breakers are steep to outside : there appears to be an open- 
ing through the reef on one side, with four or five fathoms 
within : from this description, I strongly suspect that 
Cavilli ought to bO: considered an atoll ; but, as I have not 
seen any chart of it, even on a moderately large scale, I 
have not coloured it. The islets off the northern end of 
Palaioan are, like those off the southern end, fringed by 
reefs, some way distant from the shore, but the water is 
exceedingly shallow ; uncoloured. The western shore of 
Palawan will be treated of under the China Sea. 

Philippine Aechipelago. — A chart on a large scale 
oi A])iioo Shoal, which lies near the S.E. coast of Mindoro, 
has been executed by Captain D. Eoss : it appears atoll- 
formed, but with rather an irregular outline ; its diameter 
is about ten miles ; there are two well-defined passages 
leading into the lagoon ; close outside and all round the 
reef, there is no bottom with 70 fathoms ; coloured blue. — 
Mindorc : the N.W. coast is represented in several charts 
as fringed by a reef ; and Luban Island is said by Horsburgh 
(vol. ii. p. 43G) to be 'lined by a reef.' — Luzon: Mr. 
Cuming, who has lately investigated with so much success 
the Natural History of the Philippines, informs me that 
a length of about three miles of the shore northward of 


Point St. Jago is fringed by a reef; as are (Ilorsburgh, 
vol. ii. p. 437) tho Throe Friars off Silanguin Bay. Between 
Point Capones and Playa Honda, the coast is ' lined by a 
coral-reef, stretching out nearly a mile in some places' 
(Horsburgh) ; and Mr. Cuming visited some fringing-reefg 
on other parts of tho coast, namely, near Puebla, Iba, and 
Mansinglor. In the neighbourhood of Solon-solon Bay, 
the shore is lined (Horsburgh, vol. ii. p. 439) by coral-reefs, 
stretching out a great way : there are also reefs about the 
islets off Solamague ; and as I am informed by Mr. Cuming, 
near St. Catalina, and a little north of it. The same 
gentleman informs me that there are reefs on the S.E. 
point of this island in front of Samar, extending from 
Malalabon to Bulusan. These appear to be the principal 
fringing-reefs on the coasts of Luzon ; and they have all 
been coloured red. ]\lr. Cuming informs me that none of 
them have deep water within ; although it appears from 
Horsburgh that some few extend to a considerable distance 
from the land. Within the Philippine Archipelago, the 
shores of the islands do not appear to be commonly fringed, 
with the exception of the S. shore of Masbate, and nearly 
the whole of Bokol ; which are both coloured red. On tho 
S. shore of Macjindanao, Bunwoot Island is surrounded 
(according to Forrest, Voyage, p. 253) by a coral-reef, 
which in the chart appears one of the fringing class. "With 
respect to the eastern coasts of the archipelago, I have not 
been able to obtain any account. Prof. Semper has re- 
cently published a notice (Zeitschr. f. Wissensch. Zoologie, 
Bd. xiii. 18G3, p. 558) respecting the coral-reefs of this 
ai'chipelago. It appears that some of them come under the 
class of barrier-reefs ; but as I have not seen a chart on a 
large scale, and know nothing about the depth of the water 
outside tlie reefs, nor about the slope of the encircled land, 
I cannot judge whether they properly come under the barrier 


Babuyan Islands. — Horsburgli says (vol. ii. p. 442) 
coral-reefs line the shores of the harbour in Fuga ; and the 
charts show there are other reefs about these islands. 
Camiguin has its shore in parts lined by coral-rock (Hors- 
burgh, p. 443) ; and about a mile off shore the depth ia 
between 30 and 35 fathoms. The plan of Port San Pio Q uinto 
shows that its shores are fringed with coral; coloured 
red. — Bashee Islands : Horsburgh, speaking of the south- 
ern part of the group (vol. ii. p. 445), says the shores 
of both islands are fortified by a reef, and through some of 
the gaps in it the natives can pass in their boats in fine 
weather; the bottom near the land is coral-rock. From 
the pubhshed charts, it is evident that several of these 
islands are regularly fringed ; coloured red. The northern 
islands are left uucoloured, as I have been unable to find 
any account of them. — Foemosa : the shores, especially 
the western one, seem composed chiefly of mud and sand, 
and I cannot make out that they are anywhere lined by 
reefs, except in a harbour (Horsburgh, vol. ii. p. 449) at the 
extreme northern point: hence, of course, the whole of 
this island is left uncoloured. The small adjoining 
islands are in the same case. — Patchow, or Madjiko-sima 
Geoups : Patchuson : Captain Broughton says (Voyage to 
the N. Pacific, p. 191) that boats, with some difiiculty, can 
pass through the coral-reefs, which extend along the coast, 
nearly half a mile ofl' it. His boats were well sheltered 
within the reef ; but it does not appear that the water is 
deep there. Outside the reef the depth is very irregular, 
varying from 5 to 60 fathoms ; the form of the land is not 
very abrupt ; coloured xedi.—Taijpin-san : from the de- 
scription given by the same author (p. 195) it appears that 
a very irregular reef extends from the southern island to the 
distance of several miles ; but whether it encircles a space 
of deep water is not evident ; nor, indeed, whether these 
outlying reefs are connected with those more immediately 


adjoining the land ; left uncolourcd. I may here add that 
the shore of Kumi (lying west of Patchow) has a narrow 
reef attached to it in the plan, in La Peyrouse's Atlas ; hut 
in the account of the voyage it is not stated to he of coral ; 
uncoloured. — Loo Ciioo : the greater part of the coast of 
this moderately hilly island is skirted hy reefs, which do 
not extend far from the shore, and which do not leave a 
channel of deep water within them, as may be seen in the 
charts accompanying Captain B, Hall's Voyage to Loo Choc 
(see also remarks in Appendix, p. xxi. and xxv.) There are, 
however, some ports with deep water, formed by reefs, in 
front of the valleys, in the same manner as happens at 
Mauritius. Captain Beechey, in a letter to me, compares 
these reefs with those encircling the Society Islands ; but 
there appears to be a marked difference between them, in 
the less distance at which the Loo Choo reefs lie from the 
land with relation to the probable submarine inclination of 
the land, and in the absence of an interior deep-water 
channel. Hence I have classed these reefs, with fringing- 
reefs, and coloured them red. — Pescadoees (west of 
Formosa) : Dampier (vol. i. p. 41G) has compared the ap- 
pearance of these islands to the southern parts of England ; 
they are interlaced with coral-reefs ; but as the water is very 
shoal, and as spits of sand and gravel (Horsburgh, vol. ii. 
p. 450) extend far out from them, it is impossible to decide 
whether the reefs are of coral. 

China Sea.' — Proceeding from north to south, we first 
meet the Pratas Shoal flat. 20° N.), which, according to 
Horsburgh (vol. ii. p. 335), is composed of coral, is of a 
circular form, and has a low islet on it. The reef is on a 
level with the water's edge, and when the sea runs high, 

' [The China Sea has in it many atolls. They are of large size, 
though not complete, most of them having the rim submerged, with 
the reef awash and islands on parts. Some are wholly submerged 
like Macclesfield Bank.] 


there are breakers nearly all round ; ' the water within 
seems pretty deep in some places ; although steep in most 
parts outside, there appear to he several parts where a 
ship might find anchorage outside the breakers ; ' coloured 
blue. — The Paracells have been accurately surveyed by 
Captain D. Eoss, and charts on a large scale published : 
only a few low islets have been formed on these shoals, and 
this seems to be a general circumstance in the China Sea ; 
the sea close outside these reefs is deep ; several of them 
have a lagoon-like structure ; in other cases separate islets 
(Prattle, Bobert, Drummond, &c.) are so arranged round a 
moderately shallow space as to appear as if they had once 
formed one large atoll. — Bor'.hay Shoal (one of the Para- 
cells) has the form of an annular reef, and is ' apparently 
deep within ; ' it seems to have an entrance (Horsburgh, 
vol. ii. p. 332) on the wpb*. side ; it is very steep outside. — 
Discovery Shoal, also, is of an oval form, with a lagoon-like 
space within, and three openings leading into it, in which 
there is a depth fi-om 2 to 20 fathoms. Outside, at the dis- 
tance (Horsburgh, vol. ii. p. 333) of only twenty yards from 
the reef, soundings could not be obtained. The Paracells 
are coloured blue. — Macclesfield Bank is a coral bank of 
great size, lying east of the Paracells ; some parts are level, 
with a sandy bottom, but generally the depth is very 
irregular, and intersected by deep channels ; not coloured. 
— Scarborough Shoal : this coral shoal is engraved with 
a double row of crosses, forming a circle, as if there 
was deep water within the reef : close outside no bottom 
was found with a hundred fathoms ; coloured blue. — 
The sea off the west coast of Palawan and the northern 
part of Borneo is strewed with shoals : Sioalloio Shoal, 
according to Horsburgh (vol. ii. p. 431), ' is formed, like 
viost of the shoals hereabouts, of a belt of coral-rocks, with 
a basin of deeper water within.' — Half-Moon Shoal has a 
similar structure ; Captain D. Ross describes it as a narrow 


belt of coval-rock, ' with a basin of deeper water in tlie 
centre,' and deep sea close outside. — Bombay Shoal appears 
(Horsburgh, vol. ii. p. 432) ' to be a basin of smooth water 
surrounded by breakers.' I have coloured these three 
shoals blue. — The Paraquas Shoals are of a circular form, 
with deep gaps running through them ; not coloured. A 
bank, gradually shoaling to the depth of 30 fathoms, ex- 
tends to a distance of about 20 miles from the northern 
j)art of Borneo, and to 30 miles from the southern part of 
Palaioan ; near the land this bank appears tolerably free 
from danger, but a little further out is thickly studded 
with coral-reefs, which do not generally rise to the surface ; 
some of them are very steep, whilst others have a fringe of 
shoal- water roiind them. I should have thought that these 
shoals had level surfaces, had it not been for a statement 
made by Horsburgh, ' that most of the shoals hereabouts 
are formed of a belt of coral : ' I have not coloured them. — 
The coasts of China, Tonqiun, and Cochin-China, forming 
the western boundary of the China Sea, appear to be with- 
out reefs : with regard to the two last-mentioned coasts, I 
judge from an examination of the charts on a large scale in 
the atlas of the Voyage of the Favourite. 

Indian Ocean. — South Keeling atoll has been specially 
described in my first chapter. Nine miles north of it liea 
North Keeling, a very small atoll, surveyed by the Beagle, 
the lagoon of which is dry at low water. — Christmas Island,' 
lying to the east, is a high island, without, as I have been 

' [This island is described in letters to Nature byCaptains'Wharton 
and Maclcar (vol. xxxvi. pp. 12, 413, and vol. xxxvii. p. 203 ; cf. also 
p. 222). It is I'M miles from Java, the intervening ocean attaining 
a depth of 2,430 fathoms. It consists of coral limestone, no other 
rock being visible, which rises from the sea to a height of about 
1,200 feet. At the base is commonly a clifi about 30 feet high, and 
there are two upper tiers of ciiii's : one is described as being from 200 to 
300 feet high. A depth of 100 fathoms is found at one or two cables' 
length from the water's edge at the base of the lowest line of clilla.] 


informed, any reef. — Ceylon : a space of about 80 milea 
in length on the south-western and southern shores of these 
islands has been described by Mr. Twynam (Naut. Mag. 
1836, pp. 3G5 and 518) ; and parts appear to be regularly 
fringed by coral-reefs, which extend from a quarter to half 
a mile from the shore. These reefs are in places breached, 
and afford safe anchorage for the small trading craft. Out- 
side, the sea gradually deepens ; there is 40 fathoms about 
six miles off shore : I have coloured these reefs red. In the 
published charts of Ceylon, reefs also appear to fringe 
several parts of the south-eastern shores, coloured red. — At 
Venloos Bay the shore is likewise fringed. North of 
Trincomalee there are also reefs of the same character. 
The sea off the northern part of Ceylon is exceedingly 
shallow ; and therefore I have not coloured the reefs which 
partially fringe portions of the shores, and the adjoining 
islets, as well as the Indian promontory of Madura. 

Chagos, Maldiva, and Laccadive Archipelagoes.' — 
These three great groups of atolls and atoll-formed banks, 
have been often referred to in this volume, and are now 
well known from the admirable surveys of Captain Moresby 
and Lieut. Powell. Their published charts are worthy of 
the most attentive examination. In the Laccadive group, 
the atoll-like structure is less evident than in the Maldivas ; 
nevertheless the islands are all low, not exceeding the usual 
height of coral formations (see Lieut. Wood's account, 
Geograph. Journ. vol. vi. p. 29), and most of the reefs are 
circular ; within several of them, as I am informed by 
Captain Moresby, there is deepish water ; these, therefore, 
have been coloured blue. Directly north of the Laccadives, 
and almost forming part of the same group, there is along, 
narrow, slightly-curved bank, rising out of the depths of 
the ocean, composed of sand shells and decayed coral, with 
from 23 to 30 fathoms on it. I have no doubt that it haa 
' [Sec Appendix IL] 


had the same origin with the other atoll-hke banks ; but as it 
does not deepen towards the centre, I have not coloured it. 
I might have referred to other authorities regarding these 
three archipelagoes ; but after the publication of the charts 
by Captain Moresby (to whose personal kindness in giving 
nie much information I am exceedingly indebted), this 
would have been superfluous. 

The Sahia de Malha Bank consists of a series of narrow 
banks, with from 8 to 16 fathoms on them; they are 
arranged in a semi-circular manner, round a space about 
40 fathoms in depth, sloping to the S.E. to unfathomable 
depths ; they are steep on both sides, but more especially 
on the ocean-side. Hence this bank closely resembles in 
structure, and I may add from Captain Moresby's informa- 
tion in composition, Pitt's Bank in the Chagos group ; and 
Pitt's Bank must, from what we know about the great 
Chagos Bank, be considered as a sunken, half-destroyed 
atoll ; hence coloured blue. — Cargados Carajos Bank : its 
southern portion consists of a large, curved coral-shoal, 
with some low islets on the eastern edge, and likewise 
some on the western side, between which there is a depth 
of about 12 fathoms : northward, a great bank extends. I 
cannot (probably owing to the want of perfect charts) refer 
this reef and bank to any class ; therefore not coloured. — 
He de Sable is a httle island, lying west of C. Carajos, only 
some toises in height (Voyage of the Favourite, vol. i. 
p. 130) ; it is surrounded by reefs ; but its structure is un- 
intelligible to me. There are some small banks north of 
it, of which I can find no clear account. — Mauritius : the 
reefs round this island have been described in the chapter 
on fringing-reefs ; coloured red. — Bodriguez : the coral- 
reefs here are very extensive ; in one part they project even 
five miles from the shore. As far as I can make out, there 
is no deep-water channel witliin them ; and the sea outside 
does not deepen very suddenly. The outline, however, of 


the land appears to be (Life of Sir J. Mackintosli, vol. ii. 
p. 165) billy and rugged. I am unable to decide whetlier 
these reefs belong to the barrier class, as seems probable 
fi-om their great extension, or to the fringing class ; un- 
coloured.' — Bourbon : the greater part of the shores of this 
island are without reefs ; but Captain Carmichael (Hooker's 
Bot. Misc.) states that a portion, 15 miles in length, on the 
S.E. side, is imperfectly fringed with coral-reefs ; I have not 
thought this sufficient evidence for colouring the island. 

Seychelles. — The rocky islands of primary formation, 
composing this group, rise from a very extensive and toler- 
ably level bank, having a depth of between 20 and 40 
fathoms. In Captain Owen's chart, and in that in the atlas 
of the Voyage of the Favourite, it appears that the east 
side of Mahe. and the adjoining islets of St. Anna and Ccrf, 
are regularly fringed by coral-reefs. A portion of the S.E. 
part of Curieuse Island, the N. and part of the S.W. shore 
of PraSlin Island, and the whole west side of Digue Island, 
appear fringed. From a MS. account of these islands by 
Captain F. Moresby, in the Admiralty, it appears that 
Silhouette is also fringed ; he states that all these islands 
are formed of granite and quartz, that they rise abruptly 
from the sea, and that ' coral-reefs have grown round them, 
and project for some distance.' Dr. Allan of Forres, who 
visited these islands, informs me that there is no deep water 
between the reefs ?.nd the shore. The above specified 
points have been coloured red. Amirantes Islands : the 
small islands of this neighbouring group, according to the 
MS. account of them by Captain F. Moresby, are situated 
on an extensive bank ; they consist of the debris of corals 
and shells ; tbey are only about 20 feet in heigbt, and are 

' [There are fringing-reefs of a width of four and a half miles to 
leeward and of a few yards to windward. Outside them the water 
shoals gradually. The island is high and basaltic, with upraised 
coral in many places up to a height of about 50 feet above the sea. — 
Capt. Wharton. See Appendix II.] 


environed by reefs, some attached to the shore, and some 
rather distant from it. — I have taken pains to procure 
pLins and information regarding the several islands which 
lie to the S.E. and S.W. of the Seychelles ; from accounts 
given me by Captain F. Moresby and Dr. Allan, it appears 
that the greater number — namely, PZa^fc, AlpJionse, Coetivi, 
Galcga, Providence, St. Pierre, Astova, Assoviption, and 
Glorioso ' — are low, formed of sand or coral-rock, and irre- 
gularly shaped ; they are situated on very extensive banks, 
and are in connection with great coral-reefs. Galega is 
Baid by Dr. Allan to be rather higher than the others; and 
St. Pierre is described by Captain F. Moresby as beuig 
cavernous throughout, and as not consisting of either lime- 
stone or granite. These islands, as well as the Amirantes, 
certainly are not atoll-formed, and they seem to differ from 
all other groups ; I have not colom-ed them ; but probably 
the reefs belong to the fringing class. Their formation is 
attributed both by Dr. Allan and Captain F. Moresby, to 
the action of the currents, here exceedingly violent, on 
banks which no doubt have had an independent geological 
origin. They resemble in many respects some of the 
islands and banks in the West Indies, which owe their 
origin to a similar agency, in conjunction with an eleva- 
tion of the entire area. In close vicinity to the above 
several islands, there are three others of an apparently 
different nature ; first, Juan de Nova, which appears from 
some plans and accounts to be an atoll, but from others 
this does not appear to be the case ; ^ not coloured . Secondly* 
Cosvwlcdo : ' this group consists of a ring of coral, ten 
leagues in circumference, and a quarter of a mile broad in 

' [Platte, Coetivi, and Galega have narro-w f ringing-reefs.— Cap t. 

'' [Juan de Nova is an imperfect atoll. The islands on its eastern 
or weather side have been raised about 8 or 10 feet. The western part 
is submerged.— Capt. Wharton. A coral-bank, with 5 fathoms watei 
on it stretches ofi the southern end. — Lieut. Chas. Smith.] 


some places, inclosing a magnificent lagoon, into which 
there did not appear a single opening' (Horsburgh, vol. i. 
p. 151); coloured blue.' Thirdly, Aldahra: consists of 
three islets, about 25 feet in height, with red cliffs (Hors- 
burgh, vol. i. p. 170), surrounding a very shallow basin or 
lagoon. The sea is profoundly deep close to the shore. 
Viewing this island in a chart, it would be thought to be 
an atoll ; but the foregoing description shows that there 
is something different in its nature ; Dr. Allan also states 
that it is cavernous, and that the coral-rock has a vitrified 
appearance. Is it an upheaved atoll, or the crater of a 
volcano ? — uncoloured.^ 

CoMOKO Gboup. — Mayotta, according to Horsburgh 
(vol. i. p. 216, 4th edit.), is completely surrounded by a 
reef, which runs at the distance of three, four, and in some 
places even five miles from the land ; in an old chart, pub- 
lished by Dalrymple, a depth in many places of 30 and 38 
fathoms is laid down within the reef. In the same chart, 
the space of open water within the reef is in some parts 
even more than three miles wide : the land is bold and 
peaked ; this island, therefore, is encircled by a well- 
characterized barrier-reef, and is coloured pale blue. — 
Johanna : Horsburgh says (vol. i. p. 217), this island from 
the N.W. to the S.W. point, is bounded by a reef, at the 

> [The islands on the ring have been upraised about 10 feet. — 
Capt. Wharton.] 

'•^ [Aldabra is an upraised atoll 22 miles long ; the lagoon is nearly 
dry at low water. The height of the rock on the encircling islands 
is 20 feet, and it descends on both sides to the water for that distance 
in a cliff, though on the lagoon side the coral is much disintegrated by 
tlie mangroves. This is the only island in the Indian Ocean where 
the gigantic tortoise, of a distinct species, exists. — Capt. Wharton. 
Horsburgh 's account is misleading, as neither the red cliffs nor high 
forests were to be found. It is entirely composed of coral-rock with 
a fine growth of mangroves, inclosing an extensive but shallow lagoon. 
There is a narrow riband of o fathoms water running 3 miles into the 
^ajzoon from the N.W. corner. — Lieut. Chas. Smith.] 


distance of two miles from the shore ; in some parts, how- 
ever, the reef must be attached, since Lieut. Boteler (Narr. 
voL i. p. IGl) describes a passage through it, within which 
there is room only for a few boats. Its height, as I am 
informed by Dr. Allan, is about 3,500 feet ; it is very pre- 
cipitous, and is composed of granite, greenstone, and 
quartz ; coloured blue. — Mohilla : on the S. side of this 
island there is anchorage between a reef and the shore 
in from 30 to 45 fathoms (Horsburgh, vol. i. p. 214) ; it 
appears also encircled in Captain Owen's chart of Mada- 
gascar ; coloured blue. — Great Comoro Island is, as I am 
informed by Dr. Allan, about 8,000 feet high, and ap- 
parently volcanic ; ' it is not regularly encircled ; but reefs 
of various shapes and dimensions jut out from every head- 
land on the W., S., and S.E. coasts, inside of which reefs 
there are channels, often parallel with the shore, with deep 
water. On the N.W. coasts the reefs appear attached to 
the shore. The land near the coast is in some places bold, 
but generally speaking it is flat ; Horsburgh says (vol. i. 
p. 214), the water is profoundly deep close to the shore, 
from which expression I presume some parts are without 
reefs. From this description, I apprehend the reef belongs 
to the barrier class ; but I have not coloured it, as most of 
the charts which I have seen represent the reefs round it 
as very much less extensive than round the other islands 
of the group. 

MadagasCxVE. — ]\Iy information is chiefly derived from 
the pubhshed charts by Captain Owen, and the accounts 
given by him and by Lieut. Boteler. Commencing at the 
S.W. extremity of the island : towards the northern part of 
Star Bank (in lat. 25° S.) the coast for ten miles is fringed 
by a reef ; coloured red. The shore immediately S. of St. 

' [Great Comoro is volcanic and about 8,000 feet high. There is 
a little fringing-reef on the north and on the south-east side.— Lieut. 
Chas. Smith.] 


Augustin's Bay appears fringed; but Tullear Harbour, 
directly N. of it, is formed by a narrow reef ten miles long, 
extending parallel to the shore, with from 4 to 10 fathoms 
withm it. If this reef had been more extensive, it must 
have been classed as a barrier-reef ; but as the line of coast 
falls imvards here, a submarine bank perhaps extends 
parallel to the shore, which has offered a foundation for 
the growth of the coral ; I have left this part uncoloured. 
From lat. 22° IG' to 21° 37', the shore is fringed by coral- 
reefs (see Lieut. Botcler's Narrative, vol. ii. p. 106), less 
than a mile in width, and with shallow water within. 
There are outlying coral shoals in several parts of the offing, 
with about 10 fathoms between them and the shore, and 
the depth of the sea one mile and a half seaward, is only 
about 30 fathoms. The part above specified is engraved on a 
large scale ; and as in the charts on rather a smaller scale 
the same fringe of reef extends as far as lat. 23° 15', I 
have coloured the whole of this part of the coast red. The 
islands of Juan cle Nova (in lat. 17° S.) appear in the charts 
on a large scale to be fringed, but I have not been able to 
ascertain whether the reefs are of coral ; uncoloured. The 
main part of the west coast appears to be low, with outlying 
sand banks, which Lieut. Boteler (vol. ii. p. 106) says, ' are 
faced on the edge of deep water by a line of sharp-pointed 
coral-rocks.' Nevertheless I have not coloured this part, 
as I camiot make out by the charts that the coast itself is 
fringed. The headlands of Narrenda and Passandava Bays 
(14° 40') and the islands in front of Badama harhoiir are 
presented in the plans as regularly fringed, and have 
accordingly been coloured red. With respect to the East 
Coast of Madagascar, Dr. Allan informs me, that the whole 
line of coast, from Tamatave in 18° 12' to G. Amher at the 
extreme northern point of the island, is bordered by coral- 
reefs. The land is low, uneven, and gradually rises from 
the coast. From Captain Owen's charts, the existence of 


reefs, which evidently belong to the fringing class, N. of 
British Sound and near Ngoncy, might also have been 
inferred. Lieut. Boteler (vol. i. p. 155) speaks of ' thereof 
surrounding the island of St. Mary's at a small distance 
from the shore.' In a previous chapter I have described* 
from the information of Dr. Allan, the manner in which 
the reefs extend in N.E. lines from the headlands on this 
coast, thus sometimes forming rather deep channels within 
them : this seems caused by the currents, the reefs spring- 
ing up from the submarine prolongations of the sandy head- 
lands. The above specified portion of the coast is coloured 
red.' The remaining S.E. portions do not appear in any 
published chart to possess reefs of any kind ; and the Rev. 
W. Elhs believes that there are none. 

East Coast of Africa.— The northern parts appear, for 
a considerable space, to be without reefs. My information, 
I may observe, is derived from the survey by Captain Owen, 
together with his Narrative ; and that by Lieut. Boteler. 
At Mukdcesha (2° 1' N.) there is a coral-reef extending 
four or five miles along the shore (Owen's Nar. vol. i. 
p. 357), which in the chart lies at a distance of a quarter of 
a mile from the shore, and has within it from G to 10 feet 
of water : this then is a fringing-reef and is coloured red. 
From Juba, a little S. of the equator, to Lamoo (in 2° 20' 
S.) ' the coast and islands are formed of madrepore ' 
(Owen's Narrative, vol. i. p. 3GB). The chart of this part 
(entitled Dundas Islands) presents an extraordinary ap- 
pearance ; the coast of the mainland is quite straight, and 
is fVonted at the average distance of two miles, by exceed- 
ingly narrow, straight islets, fringed with reefs. Within 
this chain of islets, there are extensive tidal flats and 
muddy bays, into which many rivers enter : the depth of 
these spaces varies from one to four fathoms — the latter 

' [The northern end of Madagascar, of volcanic origin, has upraised 
coral, find is fringed with living coral.— Capt. Wharton.] 


depth not being common, and about 12 feet the average. 
Outside the chain of islets, the sea, at the distance of a 
mile, varies in depth from 8 to 15 fathoms. Lieut. Boteler 
(Nar. vol. i. p. 3G9) describes the muddy bay of Patta, 
"which seems to resemble other parts of the coast, as fronted 
by small, narrow, level islets formed of decomposing coral, 
the margin of which is seldom of greater height than 12 
feet, overhanging the rocky surface from which the islets 
rise. Knowing that the islets are formed of coral, it is 
I think scarcely possible to view the coast, and not at once 
conclude that we here see a fringing-reef, which has been 
upraised a few feet : the unusual depth of from two to four 
fathoms within some of these islets, is probably due to the 
mud of the rivers having prevented the growth of coral 
near the shore. As several parts of this line of coast are 
midoubtedly fringed by living reefs, I have coloured it 
red. — Malcenda (3° 20' S.) : in the plan of the harbour, 
the south headland appears fringed ; and in Owen's chart 
on a larger scale, the reefs are seen to extend nearly 30 
miles southward ; coloured red. Momhas (4° 5' S.) : the 
island which forms the harbour * is surrounded by cliffs of 
madrepore, capable of being rendered almost impregnable ' 
(Owen's Nar. vol. i. p. 412). The shore of the mainland, 
N. and S. of Mombas, is regularly fringed by a coral-reef 
at a distance from half a mile to a mile and a quarter from 
the land ; within the reef the depth is from 9 to 15 feet ; 
outside the reef the depth at rather less than half a mile 
is 30 fathoms. From the charts it appears that a space 
about 36 miles in length, is here fringed ; coloured red. — • 
Femba (5° S.) is an island of coral formation, level, and 
about 200 feet in height (Owen's Nar. vol. i. p. 425) ; it is 
35 miles long, and is separated from the mainland by a 
deep sea. The outer coast is represented in the charts as 
regularly fringed ; coloured red. The mainland in front of 
Pemba is likewise fringed. — Zanzibar resembles Pemba in 


most respects; its southern half on the western side ard 
the neighbouring islets are fringed ; coloured red.^ On the 

' [The following interesting account of Zanzibar is contained 
in a letter from Captain Wliarton, found among Mr. Darwin's 
papers : — 

* Zanzibar Bcems to me to have undergone several motions of sub- 
sidence and upheaval, the latter being the latest ; it appears now to 
have been for many years nearly stationary. 

' The island at present is surrounded with a nearly perfectly flat, 
dead, altered coral ledge, more or less dry at low water, without doubt 
the result of long action of the sea on the upheaved ancient and com- 
pressed coral of which the island is principally formed. This action 
has worn away the sea face of the land to the level of low water for a 
distance, in some instances, of 1^ miles inside the original high-water 
line, which now remains as a steep rim, dropping to 10 and 20 fathoms 
almost immediately, with (on the outside of the island) 100 fathoms 
within a quarter of a mile. I could see no sign of this ledge extend- 
ing seaward, though there is living coral on its steep face visible a 
few feet below at low water, but this is not abundant, as it is on some 
of the detached reefs off the island. 

' The present high-water line of the island at the back of this flat 
area is, for the major part of its perimeter, a cliff of the same old 
coral from 10 to 20 feet in height, undermined by the waves, and 
overhanging, in some places, to a marvellous extent, showing the 
hardness and cohesion of the material, and giving a notion of the 
long period of time necessary to wear it away. As a further proof 
of this is the fact of very few lately detached pieces being seen at the 
foot of the cliffs, though the blocks, when they do fall, must be large 
and not easily moved by the sea. 

' In most parts of the island the tops of these low cliffs run back 
from the sea nearly level for a greater or less distance, showing water- 
worn coral wherever the surface rock is exposed, and indicating 
another stationary period or one of very slow upheaval. Out of thifj 
level the higher lands of the island rise. 

' Zanzibar is intersected by what may be regaj-ded as 3 lines of hills 
running north and south, the highest of them being 4y0 feet above 
the sea. 

' I regret to add that I cannot say of what formation these hills 
may be ; I cannot call to mind any rock beyond the coral, which 
crops out at considerable heights (in one instance 250 feet), but there 
is a good deal of hardened clay or mudstone, ■which generally appears 


main land, a little S. of Zanzibar, there are some banks 
parallel to the coast, which I should have thought had been 

in the ravines, &c., and on the bare sides of the hills, but there may 
be other rock lying under this. 

' The valleys, or rather flat plains, between the ranges of hills, are 
mostly (particularly to the south) coral, worn and roughened, un- 
doubtedly by water. These are generally about 50 feet above the 
sea. Several isolated hills of coral stand on these plains, their bases 
being undermined and worn precisely as the present cliffs, and their 
flat summits present the same appearance. 

' The whole thickness of the coral of Zanzibar must be very great. 

' The coast of the mainland about Zanzibar is similar to the island, 
and, as far as I know them, Pemba, Moufia, and the coast far north 
and south are the same. 

' The outlying and detached reefs are of two kinds, those growing 
up with living coral, and those of dead coral, like the island washing 
gradually away. Of these latter many still have level islets and 
rocks on them, remnants of a former upheaval ; others afford a found- 
ation to coral sand-banks that are dry high at low water, and others 
are perfectly smooth and covered at high water, being just awash at 
low tide. Of the second of these, are the reefs referred to by you at 
page 258 as described by Lieut. Boteler as sand-banks. That descrip- 
tion is erroneous. 

• One island, mentioned in the beginning of the century, had by 
Capt. Owen's time (1825) been reduced to a sand-head always visible. 
Now (1874) even this has entirely disappeared, and the reef on which 
it stood is flat and bare. 

• This is the only instance in which I have been able to make any 
reliable comparison between Capt. Owen's chart and mine, as to 
reduction of reefs. 

' As to the perhaps still more interesting question of growing coral, 
I have been unable to make any such, as Owen's work was so cursory 
and hurried that it is impossible to know whether he struck the 
shoalest part of a reef. 

' There is, indeed, one instance that, if not isolated, might have been 
of use. He describes a particular shoal as being a ' knoll with deep 
water all round,' and in his chart, 7 fathoms is marked on it and 25 
fathoms around. That patch has now only 1^ fathoms on it and 20 
fathoms round. 

' This, altogether, looks like upheaval of the whole bottom ; but as 
in most instances our soundings agree remarkably well, I cannot think 


formed of coral, had it not been said (Boteler's Nar. vol. ii. 
p. 39) that they were composed of sand : not coloured. — 
Latham's Bank is a small island fringed by coral-reefs ; 
but being only 10 feet high it has not been coloured.— 
Monfcea is an island of the same character as Pemba : its 
outer shore is fringed, and its southern extremity is con- 
nected with Keelwa Point on the main land by a chain of 
islands fringed by reefs ; coloured red. The four last-men- 
tioned islands resemble in many respects some of the islands 
in the Bed Sea, which will presently be described. — 
Keelwa: in a plan of the shore, a space of 20 miles N. and 
S. of this place is fringed by reefs, apparently of coral ; 
these reefs are prolonged still further southwards in Owen'a 
general chart. In the plans of the rivers Linchj and Mong- 
Iwio (9° 59' and 10° 7' S.) the coast seems to have the same 
structure, coloured red. — Querimba Islands (from 10° 40' 
to 13° S.) : a chart on a large scale is given of these 
islands; they are low and of coral formation (Boteler's 
Nar. vol. ii. p. 54) ; and generally have extensive reefs pro- 
jecting from them, which are dry at low water, and which 
on the outside rise abruptly from a deep sea ; on the inside 
they are separated from the continent by a channel, or 
rather a succession of bays, with an average depth of 10 
fathoms. The small headlands on the continent also have 
coral banks attached to them ; and the Querimba islands 
and banks are placed on the line of prolongation of these 
headlands, and are separated from them by very shallow 
channels. It is evident that whatever cause, whether the 

that that can be so. On the other hand, the reef is so small and the 
bottom so clear, tliat it is difficult to understand how they could 
have missed the shoaler water if it existed then, as it is very plain 
to see. 

' Other reefs, with from 7 to 10 fathoms on them,^eem not to havo 

' W. J. S. Wharton, Ck>mmander E.N. 
• Mauritius : Sept. 16, 1874.] 


drifting of sediment or subterranean movements, produced 
the headlands, hkewise produced, as might have been ex- 
pected, submarine prolongations to them ; and these to- 
wards their outer extremities have since afforded a favour- 
able basis for the growth of coral-reefs, and subsequently for 
the formation of islets. As these reefs clearly belong to the 
fringing class, the Querimba Islands have been coloured 
xedL.—Monahila (13° 32' S.) : in the plan of this harbour, 
the headlands outside are fringed by reefs apparently of 
coral; coloured red. — Mozambique [15° S.) : the outer part 
of the island on which the city is built, and the neighbour- 
ing islands are fringed by coral-reefs ; coloured red. 
From the description given in Owen's Nar. (vol. 1. p. 162) 
the shore from Mozambique to Delagoa Bay appears to be 
low and sandy : many of the shoals and islets off this line 
of coast are of coral formation ; but from their small size 
and lowness, it is not possible, from the charts, to know 
whether they are truly fringed. Hence this portion of 
coast is left uncoloured, as are likewise those parts more 
northward, of which no mention has been made in the 
foregoing pages, from the want of information.^ 

Persian Gulf. — From the charts lately published on a 
large scale by the East Indian Company, it appears that 
several parts, especially the southern shores, are fringed by 
coral-reefs ; but as the water is very shallow, and as there 
are numerous sand-banks, which are difficult to distiu- 

' [The whole of the eastern coast of Africa, from the equator to 
Mozambique (at least) is of upraised coral, and so are the outlying 
islands. Fringing-reefs occur everywhere, partly formed by the 
action of the sea wearing back the upraised coral, and partly by 
living coral. In Zanzibar undoubted coralline limestone exists at 
100 feet, and a limestone of origin as yet undetermined at 300 feet. 
— W. From Wasin to Pangani (about lat. 5° S.) there is a barrier of 
large coral-reefs from 2 to 5 miles off shore with a deep channel inside, 
sometimes as much as 20 fathoms in depth.— Lieut. Chas. Smith.— ' 
From Mr. Darwin's papers.] 

260 ArPENDix. 

guish on tlie chart from reefs, I have not coloured the upper 
part red. Towards the mouth, however, where the water ig 
rather deeper, the islands of Orviuz and Larrack, appear 
so regularly fringed, that I have coloured them red. There 
are certainly no atolls in the Persian Gulf. The shores of 
Immauvi, and of the promontory forming the southern head- 
land of the Persian Gulf, seem to be without reefs. The 
whole S.W. part of Arabia Felix, except one or two small 
patches, and the shores of Socotra appear from the charts 
and the memoir of Captain Haines (Geograph. Journ. 1839, 
p. 125) to he without reefs. I believe there are no exten- 
sive coral-reefs on any part of the coasts of India, except on 
the low promontory of Madura (as already mentioned) in 
front of Ceylon. 

Eed Sea. — My information is chiefly derived from the 
admirable charts published by the East India Company in 
183G, from personal communication with Captain Moresby, 
one of the surveyors, and from the excellent memoir, 
* Ueber die Natur der Corallen-Biinkendes Eothen Meeres,' 
by Ehrenberg, The plains immediately bordering the Eed 
Sea seem to consist chiefly of a sedimentary formation of 
the newer tertiary period. The shore is, with the exception 
of a few parts, fringed by coral-reefs. The water is gener- 
ally profoundly deep close to the shore ; but this fact, which 
has attracted the attention of most voyagers, seems to have 
no necessary connection with the presence of reefs ; for 
Captain Moresby particularly observed that, in lat. 24° 10' 
on the eastern side, there is a piece of coast with very deep 
water close to it, without any reefs, but not differing in any 
other respect from the usual coast lino. The most remark- 
able feature in the Eed Sea is the chain of submerged banks, 
reefs, and islands lying some way from the shore, chiefly on 
the eastern side ; the space within being deep enough to 
admit safe navigation in small vessels. The banks are 
generally of an oval form, and some miles in width ; but 

RED SEA. 261 

Bome of them are very long in proportion to their widtli. 
Captain Moresby informs me that any one who had not 
made actual plans of them, would he apt to think that they 
were much more elongated than they really are. Many of 
them rise to the surface, but the greater number lie from 5 
to 30 fathoms beneath it, with irregular soundings on them. 
They consist of sand and living coral ; the latter in most 
cases, according to Captain Moresby, covering the greater 
part of their surface. They extend parallel to the shore, 
and are not unfrequently connected in their middle parts 
by short transverse banks with the main land. The sea is 
generally profoundly deep quite close to them, as it is near 
most parts of the coast of the main land ; but this is not 
universally the case, for between lat. 15° and 17° the water 
deepens quite gradually from the banks, both on the eastern 
and western shores. In many parts islands rise from the 
banks ; they are low, flat-topped, and consist of the same 
horizontally stratified formation with that forming the 
plain-like margin of the main land. Some of the smaller 
and lower islands consist of mere sand. Captain Moresby 
informs me that small masses of rock, the remnants of 
islands, are left on many of the banks where there is now 
no dry land. Ehrenberg also asserts that most of the 
islets, even the lowest, have a flat abraded basis, com- 
posed of the same tertiary formation as elsewhere: he 
believes that as soon as the surf wears down the protube- 
rant parts of the banks to just beneath the level of the sea, 
the surface becomes protected from further abrasion by the 
growth of coral, and he thus accounts for the existence of so 
many banks standing on a level with the surface of this sea. 
It appears that most of the islands are certainly decreasing 
in size. 

The banks and islands are curiously shaped in the parts 
just referred to, namely, from lat. 15*^ to 17°, where the 
sea deepens quite gradually : the Dhalac group, on the 


western coast, is surrounded by an intricate archipelago of 
islets and slioals ; the main island is irregular in outline, 
and includes a bay seven miles long, by four across, in 
which no bottom was found with 252 feet ; there is only 
one entrance into it, half a mile wide, and with an island 
in front. The submerged banks on the eastern coast, 
within the same latitudes, round Farsan Island, are, like- 
wise, penetrated by many narrow creeks of deep water ; 
one is twelve miles long, in the form of a hatchet, and close 
to its broad upper end, soundings were not struck with 
360 feet ; its entrance is only half a mile wide. In 
another creek of the same nature, but even with a more 
irregular outline, there was no bottom with 480 feet.' The 
island of Farsan itself, has as singular a form as any of its 
surrounding banks. The bottom of the sea round the 
Dhalac and Farsan Islands consists chiefly of sand and 
agglutmated fragments of coral, but, in the deep and narrow 
creeks, it consists of mi;d ; the islands consist of thin, 
horizontally stratified, modern tertiary beds, containing 
but little broken coral ; ^ their shores are fringed by living 

From the account given by Eiippell ^ of the manner in 
which Dhalac is rent by fissures, the opposite sides of 
which have been unequally elevated (in one instance to 
the amount of 50 feet), it seems probable that this irregular 
form, as well as that of Farsan, may have been partly 
caused by unequal elevation ; but, considering the general 
form of the banks, and of the deep-water creeks, together 
with the composition of the land, I think their configura- 

' [The islands of this group are of upraised coral, as is the fore- 
Bhore of the opposite coast of Abyssinia. In many parts of the Eed 
Sea coast the low coral cliffs give evidence of upheaval. There are, 
nevertheless, reefs which would be classed as barrier-reefa on both 
Bides of the central part of the Eed Sea.— Capt. "Wharton.] 

* Riippell, Reise in Abyssinie, Band. i. a 217. 

» Ibid. 8. 245. 

RED SEA. 263 

tion is more probably due in great part to currents having 
drifted sediment over an uneven bottom. It is almost 
certain that their form cannot be attributed to the growth 
of coral. The greater number of banks on the eastern 
side of the Eed Sea seems to have originated in nearly the 
same manner, whatever this may have been, as the Dhalac 
and Farsan archipelagoes. I judge of this from their 
similar configuration (in proof of which I may instance a 
bank on the east coast in lat. 22°) and from their similar 
composition. The depth, however, within the banks north- 
ward of lat. 17° is usually greater, and their outer sides 
shelve more abruptly (circumstances which seem to go 
together) than in the Dhalac and Farsan archipelagoes ; 
but this may have been caused by a stronger action of the 
currents during their formation : moreover, the greater 
abundance of living coral on the northern banks, tends to 
give them steeper margins. 

From this account, brief and imperfect as it is, we can 
see that the great chain of banks on the eastern side of the 
Eed Sea, and on the western side of the southern portion, 
differ greatly from true barrier-reefs, which are wholly 
formed by the growth of coral. Ehrenberg also concludes 
(Ueber die, &c. pp. 45 and 51) that these banks owe their 
origin in a quite secondary manner to the growth of coral. 
He remarks that the islands off the coast of Norway, if 
worn down level with the sea, and merely coated with 
living coral, would present a nearly similar appearance. 
It seems, however, from information given me by Dr. Mal- 
colmson and Captain Moresby, that Ehrenberg has rather 
under-rated the influence of corals on the formation of the 
tertiary deposits of the Eed Sea. 

The West Coast of the Bed Sea between Lat.l9° and 22°. 
— Eeefs exist hfere, which, if I had known nothing of the 
others in the Eed Sea, I should unhesitatingly have con- 
sidered as barrier-reefs. One of these reefs, in 20° 15', ia 

264 APrENDix. 

twenty miles long, less tliau a mile in width (but expanding 
at the northern end into a disk), slightly sinuous, and 
parallel to the main land at the distance of five miles from 
it, with very deep water inside, so that in one place sound- 
ings were not obtained with 205 fathoms. Some leagues 
further south, there is another very narrow reef, ten miles 
long, with other small portions of reef, north and south, 
almost connected Avith it ; and within this line of reefs (as 
well as outside) the water is profoundly deep. There are 
also some small linear and sickle-formed reefs, lying a 
little way out at sea. All these reefs are covered, as I am 
informed by Captain Moresby, by living corals. Here, then, 
we have all the characters of reefs of the barrier class, and 
some of the outlying reefs partially resemble atolls. My 
only source of doubt arises from the narrow'ness and 
straightness of the spits of sand and rock in the Dhalae 
and Farsan groups ; one of those spits in the former group 
is nearly fifteen miles long, only two broad, and is bordered 
on each side with deep water ; so that, if worn down by 
the surf, and coated with living corals, it would form a reef 
nearly similar to those within the space under considera- 
tion. Nevertheless I cannot believe that the many small, 
isolated, and sickle-formed reefs, as well as others long, 
nearly straight, and very narrow, with the water unfathom- 
ably deep close round them, could have been formed by 
corals merely coating banks of sediment or the abraded 
surfaces of irregularly shaped islands. It seems more pro- 
bable that the foundations of these reefs have subsided, and 
that the corals, during their upw-ard growth, have given to 
them their present forms. I have, therefore, with much 
hesitation coloured this part blue. 

The West Coast, from Lat. 22° to 24°.— This part of 
the coast (north of the space coloured blue\)n the map) is 
fronted by an irregularly shelving bank, from 10 to 30 
fathoms deep ; numerous little reefs, some of which have 



the most singular shapes, rise from this bank. Many of 
them may have been formed by the growth of coral on 
small abraded islets ; but some almost atoll-formed reefs 
rising from deep water near a promontory in lat. 24°, aro 
probably allied to the barrier class. I have not, however, 
ventured to colour this portion of coast blue. — On the locst 
coast, from lat. 19° to 17° (south of the space coloured blue 
on the map), there are many low islets of small dimensions 
not much elongated, and rising out of great depths at a 
distance from the coast: these cannot be classed either 
■with atolls, or barrier, or fringing-reefs. 

Eastern Coast. — There are many small outlying coral- 
reefs along this whole line of coast ; but as the greater 
number rise from banks not very deeply submerged, their 
origm, as we have seen, may be due simply to the growth 
of corals on an irregular abraded foundation. But between 
lat. 18° and 20° there are so many linear, elliptic and ex- 
tremely small reefs, rising abruptly out of profound depths, 
that the same reasons which led me to colour a portion of 
the west coast blue, have induced me here to do the same. 
There are some small outlying reefs on the east coast, north 
of lat. 20° (the northern limit coloured blue), which rise 
from deep water ; but as they are not numerous, and as 
scarcely any of them are linear, I have left them un- 

In the southern parts of the Eed Sea, considerable 
spaces of the main land, and some of the Dhalac islands, 
are skirted by reefs, which, as I am informed by Captain 
Moresby, are of hving coral, and have all the characters of 
the fringing class. As there are here no outlying linear 
or sickle-formed reefs, rising out of unfathomable depths, 
I have coloured these parts of the coast red. On similar 
grounds I have coloured the northern parts of the western 
coast (north of lat. 24° 30') red, and hkewise the shores of 


the chief part of the Gulf of Suez} In the Gulf of Acaha, 
as I aui informed by Captain Moresby, there are no coral- 
reefs, and the water is profoundly deep. 

"West Indies. — My information regarding the reefs of 
this area is derived from various sources, and from an ex- 
amination of numerous charts ; especially of those lately 
executed during the survey under Captain Owen, E.N. I lie 
under particular obligation to Captain Bird Allen, E.N., 
one of the members of the late survey, for many personal 
communications on this subject. As in the case of the 
Eed Sea, it is necessary to make some preliminary remarks 
on the submerged banks of the West Indies, which are in 
some degree connected with coral-reefs, and cause consider- 
able doubts in their classification. That large accumula- 
tions of sediment are in progress on the West Indian shores, 
will be evident to any one who examines the charts of that 
sea, especially of the portion north of a line joining 
Yucutan and Florida. The area of deposition seems les3 
intimately connegted with the debouchement of the great 
rivers, than with the course of the sea-currents ; as is 
evident from the vast extension of the banks from the pro- 
montories of Yucutan and Mosquito. 

Besides the coast-bauks, there are others of various 
dimensions which stand isolated ; these closely resemble 
each other ; they lie from 2 or 3 to 20 or 30 fathoms under 
water, and are composed of sand, sometimes firmly ag- 
glutinated, with nttle or no coral ; their surfaces are smooth 
and nearly level, shelving very gradually to the amount of 
a few fathoms all round towards their edges, where they 
plunge abruptly into the unfathomable sea. This steep 
inclination of their sides, which is likewise characteristic 
of the coast-banks, is very remarkable : I may give as an 

' [Wherever I have seen the coast of the Red Sea, it shows clear 
signs of upheaval in low coral clilfs. There are, nevertheless, reefa 
which would be classed as barrier on both sides of the central part 
of the Red Sea.— Capt. Wharton. For Masamarhu Island, see App. II.J 


instance, tlae Misteriosa Bank, on the edges of which the 
soundings change in 250 fathoms horizontal distance, from 
11 to 210 fathoms ; off the northern point of Old Provi- 
dence Bank, in 200 fathoms horizontal distance, the change 
is from 19 to 152 fathoms ; off the Great Bahama Bank, in 
160 fathoms horizontal distance, the inclination is in many 
places from 10 fathoms to no hottom with 190 fathoms. 
In all parts of the world, where sediment is accumulat- 
ing, something of the same kind may he ohserved ; the 
banks shelving very gently far out to sea, and then termin- 
ating abruptly. The form and composition of the banks in 
the middle parts of W. Indian sea, clearly show that their 
origin must be chiefly attributed to the accumulation of 
sediment ; and the only obvious explanation of their iso- 
lated position is the presence of a nucleus, round which 
the currents have collected fine drift matter. Any one who 
will compare the bank surrounding the hilly island of Old 
Providence, with the banks in its neighbourhood which 
stand isolated, will scarcely doubt that they surround sub- 
merged mountains. We are led to the same conclusion by 
examining the bank called Thunder Knoll, which is separ- 
ated from the Great INIosquito bank by a channel only seven 
miles wide, and 145 fathoms deep. There cannot be any 
doubt that the Mosquito bank has been formed by the ac- 
cumulation of sediment round the promontory of the same 
name ; and Thunder Knoll resembles the Mosquito bank, 
in the state of its surface submerged 20 fathoms, in the in- 
clination of its sides, in composition, and in every other 
respect. I may observe, although the remark is here irre- 
levant, that geologists should be cautious in concluding that 
all the outlyers of any formation have once been connected 
together, for we here see that deposits, doubtless of exactly 
the same nature, may be deposited with large valley-like 
spaces between them. 

Linear coral-reefs and small knolls project from many 


of the isolated, as well as from the coast banks ; sometimes 
they are irre^^ularly placed, as on the Mosquito bank, but 
more generally tlioy form crescents on the windward side, 
situated some little distance within the outer edge : — thus 
on the Serranilla bank they form an interrupted chain which 
ranges between two and three miles within the windward 
margin : generally they occur, as on Eoncador, Courtown 
and Ancgada banks, nearer the line of deep water. Their 
occurrence on the windward side is conformable to the 
general rule, of the efficient kinds of corals flourishing best 
where most exposed ; but I cannot explain their position 
some way within the line of deep water unless it be that a 
depth somewhat less than that close to the outer margin is 
most favourable to their growth. Where the corals have 
formed a nearly continuous rim, close to the windward 
edge of a bank some fathoms submerged, the reef closely 
resembles an atoll ; and if the bank surrounds an island (as 
in the case of Old Providence), the reef resembles an encir- 
cling barrier-reef. I should undoubtedly have classed some 
of these fringed banks as imperfect atolls, or barrier-reefs, 
if the sedimentary nature of their foundations had not been 
evident from the presence of other neighbouring banks, of 
similar forms and of similar composition, but without the 
crescent-like marginal reef. In the third chapter, I re- 
marked that some atoll-like reefs probably did exist, which 
had originated in the manner here supposed. 

Proofs of elevation within recent tertiary periods abound, 
as referred to in the sixth chapter, over nearly the whole 
area of the West Indies. Hence it is easy to understand 
the origin of the low land near those coasts where sedi- 
ment is now accumulating ; for instance, on the northern 
part of Yucutan, and on the N.E. part of Mosquito. Hence, 
also, the origin of the great Bahama banks, which are 
bordered on their western and southern edges by narrow 
long, singularly-shaped islands, formed of sand, shells and 


coral-rock, some of them being about a hundred feet in 
height, is easily explained by the elevatit. n of banks fringed 
on their windward sides by coral-reefs. On this view, how- 
ever, we must suppose either that the great Bahama sand- 
banks were all originally deeply submerged, and were 
brought up to their present level by the same elevatory 
action which formed the linear islands ; or that during the 
elevation of the banks, the superficial currents and swell 
of the waves wore them down, and kept them at a nearly 
uniform level. But this level is not quite uniform ; for in 
proceeding from the N.W. end of the Bahama group towards 
the S.E., the depth of the banks increases, and the area of 
land decreases, in a very gradual and remarkable manner. 
The view that these banks have been worn down by the 
currents and waves of the sea during their elevation, seems 
to me the most probable one. This view is also, I believe, 
applicable to many of the submerged banks, in widely dis- 
tant parts of the West Indian sea ; for, on any other view, 
the elevatory forces must have acted with astonishing uni- 

The shore of the Gulf of Mexico, for a space of many 
hundred miles, is formed by a chain of lagoons, from 1 to 20 
miles in breadth (Columbian Navigator, p. 178, &c.), con- 
taining either fresh or salt water, and separated from the 
sea by linear strips of sand. The shores of southern Brazil, 
and of the United States from Long Island (as observed by 
Professor Eogers,) to Florida, have the same character. 
Professor Eogers, in his report to the British Association 
(vol. iii. p. 13), speculates on the origin of these low, sandy, 
linear islets ; he states that the layers of which they are 
composed are too homogeneous, and contain too large i\, 
proportion of shells, to permit the common supposition of 
their formation being simply due to matter thrown up, 
where it now lies, by the surf: he considers these islands 
as upheaved bars or shoals, which were depositt^d in linca 


where opposed currents met. It is evident that these islands 
and spits of sand parallel to the coast and separated from 
it by shallow lagoons, have no necessary connection with 

Having now endeavoured to remove some sources of 
doubt in classifying the reefs of the West Indies, I will 
give my authorities for colouring such portions of coast as 
I have thought myself warranted in doing. Captain Bird 
Allen informs me that most of the islands on the Bahama 
Banks are fringed, especially on their windward sides, with 
living reefs ; and hence I have coloured those, which are 
thus represented in Captain Owen's cliart, red. The same 
officer informs me, that the islets along the southern part 
of Florida are similarly fringed ; coloured red. — Cuba : 
proceeding along the northern coast, at the distance of 40 
miles from the extreme S.E. point, the shores are fringed, 
by reefs, which extend westward for a space of IGO miles, 
with only a few breaks. Parts of these reefs are represented 
in the plans of the harbours on this coast by Captain Owen ; 
and an excellent description is given of them by Mr. Taylor 
(Loudon's Mag. of Nat. Hist. vol. ix. p. 449) ; he states 
tliat they enclose a space called the ' haxo,' from half to 
three-quarters of a mile in width, with a sandy bottom, and 
a little coral. In most parts people can wade, at low water, 
to the reef ; but in some parts the depth is between two 
and three fathoms. Close outside the reef, the depth is 
between six and seven fathoms : these well-characterized 
fringing-reefs are coloured red. — Westward of long. 77° 30', 
on the northern side of Cuba, a great bank commences, 
which extends along the coast for nearly four degrees of 
longitude. In its structure, and in the ' cays,' or low 
islands on its edge, there is a marked correspondence (as 
observed by Humboldt, Pers. Narr. vol. vii. p. 88) between 
it and the great Bahama and Sal Banks, which lie directly 
in front. Hence one is led to attribute the same origin to 


all these baaks ; namely, the accumulation of sediment, 
conjoined with an elevatory movement, and the growth of 
coral on their outer edges. The parts which are fringed 
by living reefs are coloured red. — Westward of these banks 
there is a portion of coast apparently without reefs, except 
in the harbours, the shores of which seem in the published 
plans to be fringed. — The Colorado Shoals (see Captain 
Owen's charts), and the low land at the western end of 
Cuba, correspond as closely in relative position and struc- 
ture to the hanks at the extreme point of Florida, as 
the banks above described on the north side of Cuba 
do to the Bahamas. The depth within the islets and 
reefs on the outer edge of the Colorados, is generally be- 
tween two and three fathoms, increasing to 12 fathoms in 
the southern part, where the bank becomes nearly open, 
without islets or coral-reefs ; the portions which are fringed 
are coloured red. — The southern shore of Cuba is deeply 
concave, and the included space is filled up with mud and 
sand-banks, low islands and coral-reefs. Between the 
mountainous Isle of Pines and the southern shore of Cuba, 
the general depth is only between two and three fathoms ; 
and in this part, small islands, formed of fragmentary rocks 
and broken madrepores (Humboldt, Pers. Narr. vol. vii. pp. 
61, 86 to 90, 291, 309, 320), rise abruptly, and just reach 
the surface of the sea. From some expressions used in the 
Columbian Navigator (vol. i. pt. ii. p. 94), it appears that 
considerable spaces along the outer coast of southern Cuba 
are bounded by cliffs of coral-rock, formed probably by the 
upheaval of coral-reefs and sand-banks. The charts re- 
present the southern part of the Isle of Pines as fringed by 
reefs, which the Columb. Navig. says extend some way from 
the coast, but have only from 9 to 12 feet of water on 
them ; these are coloured red. — I have not been able to pro- 
. cure any detailed description of the large group of banks and 
' cays ' further eastward on tlie southern side of Cuba ; 


within them there is a large expanse, with a muddy bottom, 
from 8 to 12 fathoms deep : although some parts on thi8 
line of coast are represented in the general charts of the 
West Indies, as fringed, I have not thought it prudent to 
colour them. The remaining portion of the south coast of 
Cuba appears to be without coral-reefs. 

YucuTAN. — The N.E. part of the promontory appears, 
in Captain Owen's charts, to be fringed ; coloured red. 
The eastern coast from 20° to 18° is fringed. South of hit. 
18°, there commences the most remarkable reef in the 
West Indies : it is about 130 miles in length, ranging in a 
N. and S. line, at an average distance of fifteen miles from 
the coast. The islets on it are all low, as I have been in- 
formed by Captain B. Allen ; the water deepens suddenly on 
the outside of the reef, but not more abruptly than off many 
of the sedimentary banks : within its southern extremity 
(off Honduras) the depth is 25 fathoms ; but in the more 
northern parts, the depth soon decreases to 10 fathoms, and 
within the northernmost part, for a space of 20 miles, the 
depth is only from one or two fathoms. In most of these 
respects we have the characteristics of a barrier-reef ; never- 
theless, from observing, first, tbat the channel within the 
reef is a continuation of a great irregular bay, which pene- 
trates the mainland to the depth of 50 miles ; and secondly, 
that considerable spaces of this barrier-like reef (for instance, 
in lat. 16° 45' and 16° 12') are described in the charts as 
formed of pure sand ; and thirdly, from knowing that sedi- 
ment is accumulating in many parts of the West Indies ia 
banks parallel to the shore ; I have not ventured to colour 
this reef as a barrier. To add to my doubts, close outside 
this barrier-like reef, Turneffe, Lightliousc, and Glover xcaii 
are situated, and these have so completely the form of atolls, 
that if they had occurred in the Pacific, I should not have 
hesitated to colour them blue. Turneffe Beef seems almost 
entirely filled iip with low mud islets ; and the depth within 


tl'G other two reefs is only from one to three fathoms. From 
this circumstance, and from their similarity in form, struc- 
tm-e, and relative position, both to the bank called Northern 
Triangles, on which there is an islet between 70 and 80 
feet in height, and to Cozumel Island, the level surface of 
which is likewise between 70 and 80 feet high, it is probable 
that tho three foregoing banks are the worn-down bases of 
upheaved shoals, fringed with corals ; left uncoloured. 

In front of the eastern Mosqiiito coast there are, between 
lat. 12° and 16°, some extensive banks (already mentioned), 
with high islands rising from their centres, and others wholly 
submerged, both kinds being bordered, near their windward 
margins, by crescent-shaped coral-reefs. But it can hardly 
be doubted that these banks owe their origin, like the great 
bank extending from the Mosquito promontory, almost en- 
tirely to the accumulation of sediment, and not to the growth 
of corals ; hence I have not coloured them. 

Cayman Island : this island appears in the charts to be 
fringed ; and Captain B. Allen informs me that reefs extend 
about a mile from the shore, and have only from 5 to 12 feet 
of water within them ; coloured red. — Jamaica : judging 
from the charts, about 15 miles of the S.E. extremity, and 
about twice that length at the S.W. extremity, and some 
portions on the S. side near Kingston and Port Eoyal, are 
regularly fringed, and are therefore coloured red. From the 
plans of some harbours on the N. side, parts of the coast ap- 
pear to be there fringed ; but I have not coloured them. — 
St. Domingo : I have not been able to obtain sufficient 
information, either from plans of the harbours, or from 
general charts, to enable me to colour any part of the coast, 
except 60 miles from Port de Plata westward, which seems 
regularly (ringed : many other parts, however, of the coast 
are probably fringed, especially towards the eastern end of 
the island. — Fticrto Bico : considerable portions of tho 
southern, western, and eastern coasts, and some parts of tha 


northern coast, appear in the charts to be fringed ; coloured 
red. Some miles in length of the southern side of the Island 
of St. Thomas is fringed ; most of the Virgin Gorda Islands, 
as I am informed by Sir R. Schomburgk, are fringed ; the 
shores of Anegada, as well as the bank on which it stands, 
are likewise fringed ; these islands have been coloured red. 
The greater part of the southern side of Santa Crtiz ap- 
pears in the Danish survey to be fringed (see also Professor 
Hovey's account of this island, in Silliman's Journal, 
vol. XXXV. p. 74) ; the reefs extend along shore for a consider- 
able space, and project rather more than a mile ; the depth 
within the reef is three fathoms ; coloured red. — The An- 
tilles, as remarked by Voa Buch (Descrip. lies Canaries, 
p. 494), maybe divided into two linear groups, the western 
row being volcanic, and the eastern of modern calcareoiis 
origin ; my information is very defective on the whole group. 
Of the eastern islands, Barbzcda and the western coasts of 
Antigua and Mariagalante appear to be fringed ; this is also 
the case with Barbadoes, as I have been informed by a resi- 
dent ; these islands are coloured red. On the shores of the 
western Antilles, of volcanic origin, very few coral-reefs 
appear to exist. The island of Martinique, of which there 
are beautifully executed French charts on a very large scale, 
alone presents any appearance worthy of special notice. 
The south-western, southern, and eastern coasts, together 
forming about half the circumference of the island, are 
skirted by very irregular banks, projecting generally rather 
less than a mile from the shore, and lying from two to fivo 
fathoms submerged. In front of almost every valley, they 
are breached by narrow, crooked, steep-sided passages. The 
French engineers ascertained by boring, that these sub- 
merged banks consisted of madreporitic rocks, covered in 
many parts by thin layers of mud or sand. From this fact, 
and especially from the structure of the narrow breaches, 
these banks were probably formed by living reefs, which 


fringed the shores of the island, and once reached the sur- 
face. From some of these submerged banks reefs of Hving 
coral still rise abruptly, either in small detached patches, 
or in lines parallel to, but some way within, the margin. 
Besides the above banks which skirt the shores of the island, 
there is on the eastern side a range of linear banks, similarly 
constituted, 20 miles in length, extending parallel to the 
coast-line, and separated from it by a space between two and 
four miles in width, and from 5 to 15 fathoms in depth. 
From this range of detached banks, some linear reefs of 
living coral likewise rise abruptly ; and if they had been of 
greater length (for they do not front more than a sixth part 
of the circumference of the island) they would necessarily 
from their position have been coloured as barrier-reefs ; as 
the case stands, they are left uncoloured. 

Florida. — An account of the reefs on this coast, toge- 
ther with references to various authorities, will be found in 
Professor Dana's work on Corals and Coral Islands, 1872, 
p. 204.1 

The Bermuda Islands have been carefully described 
by Lieut. Nelson, in an excellent memoir in the Geol. 
Transactions (vol. v. part i. p. 103).^ In the form of the 
bank or reef, on one side of which the islands stand, there 
is a close general resemblance to an atoll ; but in the fol- 
lowing respects there is a considerable difference, — first, in 
the margin of the reef not forming (as I have been informed 
by Mr. Chaffers, R.N.) a flat, solid surface, which is laid 
bare at low water ; secondly, in the water gradually shoal- 
ing for nearly a mile and a half in width round the entire 
reef, as may be seen in Captain Kurd's chart ; and thirdly, 
in the size, height, and extraordinary form of the islands, 
which present little resemblance to the long, narrow, simple 

' [See Appendix II.] 

^ [An iuteresting account will also be found in Sir Wyville 
Thomson, Voyage of the Challenger, vol. i. chap, iv.] 


islets, seldom exceeding half a mile in breadth, which sur- 
mount the annular reefs of almost all the atolls in tliQ 
Indian and Pacific Oceans. Moreover, there are evident 
proofs (Nelson, ibid. p. 118) that islands similar to the ex- 
isting ones formerly extended over other parts of the reef. 
It would, I believe, be difficult to find a true atoll with 
land exceeding 30 feet in height ; Avhereas, Mr. Nelson es- 
timates the highest point of the Bermuda Islands at 2G0 
feet ; if, however, Mr. Nelson's view, that the whole land 
consists of sand drifted by the winds and agglutinated to- 
gether, is correct, this difference would be immaterial ; but, 
from his own account (p. 118), there occur in one place 
five or six layers of red earth, interstratified with the ordi- 
nary calcareous rock, and including stones too heavy for 
the wind to have moved, without having at the same time 
utterly dispel sed every grain of the accompanying drifted 
matter. Mr. Nelson attributes the origin of these several 
layers, with their embedded stones, to violent catastrophes; 
but farther investigation has generally succeeded in ex- 
plaining such phenomena by simpler means. Finally, I 
may remark that these islands bear a considerable resem- 
blance in shape to Barbuda in the "West Indies, and to 
Pemba on the eastern coast of Africa, which latter island 
is about 200 feet in height, and consists of coral-rock. 
I beheve that the Bermuda Islands, from being fringed by 
living reefs, ought to have been coloured red ; but I have 
left them uncoloured, on account of their general resem- 
blance in external form to a lagoon-island or atoll. Pro- 
fessor Dana (Corals and Coral Islands, pp. 218, 2G9) ranks 
them in this class. ^ 

' [The following particulars relating to Bennuda, taken from the 
Report of tlie Challenger Voyage, Narrative, p. 138, are of interest:— 

An excavation made to form a bed for the floating dock went 
dovm to 50 feet below low-water mark. It cut through calcareous 
mud, loose beds (coral-sands mixed with mollusks, smaller corala 
and other organisms), passins into a loosely coherent freestone 


Supplement on a remarkable Bar of Sandstone off Per- 
nambuco, on the Coast of Brazil. (Originally published 
iu the Philosophical Magazine, October 1841, p. 257.) 

In entering the harbour of Pernambuco, a vessel passes 
close round the point of a long reef, which, viewed at high 
water when the waves break heavily over it, would natu- 
rally be thought to be of coral formation, but when beheld 
at low water might be mistaken for an artificial breakwater, 
erected by cyclopean workmen. At low tide it shows itself 
as a smooth level-topped ridge, from 80 to 60 yards in width, 
with even sides, and extending in a perfectly straight line, 
for several miles parallel to the shore. Off the town it in- 
cludes a shallow lagoon or channel about half a mile in 
width, which farther south decreases to scarcely more than 
a hundred yards. Close within the northern point, ships 
lie moored to old guns let into the reef. Here, on the 
inner side, at low water spring-tides, a section of about 
seven feet in height is exhibited. This consists of hard 
pale-coloured sandstone breaking with a smooth fracture, 
and formed of siliceous grains, cemented by calcareous 
matter. Well-rounded quartz pebbles, from the size of a 
bean, rarely to that of an apple, are embedded in it, 
together with a very few fragments of shells. Traces of 

formed of the same material cemented ; and then, at a depth of 45 feet, 
through an old peat with land vegetation, shells of Helix bermitdensis, 
and bones of birds, beneath which was the ordinary hard ' base rock.' 
Kerpulaj are very abundant on the Bermuda reefs, and form evi- 
dently, by their mode of growth, miniature atolls from 2 to 20 feet in 
diameter, with little interior lagoons. It was found by soundings 
that on the S.E. edge of the bank the 100-fathom hne was about 1^ 
mile from the rocks awash. Then a slope of about 20° led down to 
350 or 400 fathoms, after which it varied, from 7° to 15°, to 1,000 
fathoms. The 100-fathom line on the N.E. edge was about 3 miles 
away; on the S.W. still further, and the submarine slopes were more 
gentle. The rock of the island appears to be of seolian origin, but it 
is not said whether this also forms the highest ground, which is 256 
feet above the sea.."l 


stratification are obscure, but in one spot there was an in- 
cluded layer of stalactitic limestone, an eighth of an inch in 
thickness. In another place some false strata, dipping 
landwards at an angle of 45°, were capped by a horizontal 
mass. On each side of the ridge quadrangular fragments 
have subsided; and the whole mass is in some places 
fissured, apparently from the washing out of some soft 
underlying bed. One day, at low water, I walked a full 
mile along this singular, smooth, and narrow causeway, 
with water on both sides of me, and could see that for 
nearly a mile further south its form remained unaltered. 
In Baron Eoussin's beautiful chart of Pernambuco {Le 
Pilote clu Brdsil) it is represented as stretching on, in an 
absolutely straight hne, for several leagues ; how far its 
composition remains the same, I know not ; but from the 
accounts I received from intelhgent native pilots, it seems 
to be replaced on some parts of the coast by true coral- 

The upper surface, though it must on a large scale be 
called smooth, yet presents, from unequal disintegration, 
numerous small irregularities. The larger embedded peb- 
bles stand out supported on short pedestals of sandstone. 
There are, also, many sinuous cavities, two or three inches 
in width and depth, and from six inches to two feet in 
length. The upper edges of these furrows sometimes 
slightly overhang their sides ; and they end abruptly with 
a rounded outline. A furrow occasionally branches into 
two arms, but generally they rim nearly parallel to each 
other, in a line transverse to the sandstone ridge. I know 
not how to account for their origin except through the 
washing to and fro of pebbles in originally shght depres- 
sions, by the waves which break daily over the bar. Op- 
posed to this notion is the fact that some of these furrows 
were lined with numerous small living Actinia. The 
exterior surface of the bar is coated with a thin layer of 


calcareous matter; this, on the outer subsided masses, 
which can be reached only at low water, between the 
successively breaking waves, is so thick that I could 
seldom expose the sandstone by the aid of a heavy hammer. 
I procured, however, some fragments, which were between 
three and four inches in thickness, and consisted chiefly of 
small Serpulce, including some Balani, with a few thin 
paper-like layers of &NuUipora. The surface alone is alive, 
and all within consists of the above organic bodies, filled up 
with dirty white calcareous matter. The layer, though not 
hard, is tough, and from its rounded surface resists the 
breakers. Along the vv^hole external margin of the bar, I 
only saw one very small point of sandstone which was ex- 
posed to the surf. In the Pacific and Indian Oceans the 
outer and upper margin of the coral-reefs are, as we have 
seen, protected by a similar coating ; but formed almost 
exclusively of several species of NulliporcB. Lieut. Nelson, 
in his excellent memoir on the Bermudas (Geol. Trans, 
vol. v. part 1, p. 117), says that the reefs there are formed 
of similar masses of SerpulcB ; but I suspect that they are 
only tlms coated. 

I enquired from some old pilots at Pernambuco whether 
there was any tradition of the bar having madergone any 
change during the lapse of time ; but they were unanimous 
in answering me in the negative. It is astonishing to re- 
flect, that although waves of turbid water, charged with 
sediment, are driven night and day by the ceaseless trade 
wind against the abrupt edges of this natural breakwater, 
yet that it has lasted in its present perfect state for cen- 
turies, or perhaps for thousands of years. Seeing that the 
surface on the inner side does gradually wear away, as 
shown by the pebbles on the little sandstone pedestals, this 
durability must be entirely due to the protection afibrded 
by the thin coating of Serjoulce, and other organic bodies. 


This is a fine example of what apparently ineCGcientmeana 
may be effectual.' 

I believe that similar bars of rock occur in front of some 
of the other bays and rivers on the coast of Brazil : Baron 
Roussin states that at Porto Seguro there is a • quay ' simi- 
lar to that of Pernambuco. Spaces of several hundred 
miles in length on the shores of the Gulf of Mexico, the 
United States, and of Southern Brazil are formed by long 
narrow islands and spits of sand, including extensive shal- 
low lagoons, some of which are several leagues in width. 
The origin of these linear islets is rather obscure : Pro- 
fessor Rogers (Pieport to British Association, vol. iii. p. 13) 
gives reasons for suspecting that they have been formed by 
the upheaval of sand-banks, deposited where currents 
formerly met. The bar of sandstone at Pernambuco has 
probably been formed in an analogous manner. The town 
stands partly on a low narrow islet and partly on a long 
spit of sand, in front of a low shore, bounded in the distance 
by a semicircle of hills. By digging at low water near the 
town, the sand is found consolidated into sandstone, similar 
to that of the bar, but containing many more shells. If, 
then, the nucleus of a spit of sand, extending in front of 
the bay, had formerly become consolidated, a small change, 
probably of level, but perhaps merely in the currents, might 
have given rise, by washing away the loose sand, to a struc- 
ture like that of the bar in front of Pernambuco and along 
the coast southward of it ; but without the protection af- 
forded by the successive growth of the above-named organic 
beings, its duration would have been short. 

' [There is an interesting acconnt of this reef, containing particulars 
of some borings undertaken in 1874, by Mr. J. C. Hawkshaw, in tlie 
QuarUjilj- Journal of the Geological Society for 1879 (vol. xxxv. p. 23'J).] 



By Peofessor T. G. Bonney, D.Sc, LL.D., F.R.a 

Since the publication of the last edition of Mr. Darwin'3 
work several important researches have been undertaken, 
which have added largely to the stock of knov/ledge con- 
cerning marine physical geography in general and coral 
reefs in particular. Of the valuable material thus obtained 
Mr. Darwin would, no doubt, have availed himself had hi3 
life been spared and his health allowed. Probably addi- 
tions would have been made to the text of this work, and 
not a few pages have been rewritten ; the older and less 
precise information being replaced by the more ample and 
exact results of recent explorations. The criticism to 
which Mr. Darwin's theory has been subjected during the 
last few years, and the hypotheses which have been ad- 
vanced by other workers, would, no doubt, have been dis- 
cussed by him in that candid and philosophic spirit which is 
so evident in all his writings. This revision may be comited 
as one of the heavy losses which science has suffered by his 
death. It became, then, a question, when a new edition 
of this book was called for, what should be done to it. 
Simply to reprint the last edition, without any notice 
of the important contributions which have been made to 
the knowledge of the subject during the last few years, 
seemed undesirable ; but any attempt to reconstruct the 
work seemed yet more undesirable. Mr. F. Darwin, after 


consultation with several of bis friends (including myself, 
whom he had asked to aid him in preparing the new 
edition for the press), accordingly decided to reprint ' The 
(Structure and Distribution of Coral Keefs ' from the 
edition of 1874, subject only to a few press corrections, and 
to give any important emendations or additions in the 
form of notes, so arranged as to be easily distinguished 
from those written by the late author. As regards the 
extent and amount of the additional matter, we thought 
that, as the volume was never intended as a text-book 
for examination purposes, it was needless to endeavour to 
concentrate within its pages every result of recent work, 
and it would suffice to call attention to the more important 
points, which would almost certainly have been noticed by 
the author in any new edition. 

Therefore, from a few papers left by Mr. Darwin, from 
information kindly supplied by Capt. Wharton and other 
friends, and from my own reading, I have added a few foot- 
notes to the text, and have given in this appendix a summary 
of the papers which appeared to me of special importance 
among those which have been pubhshed smce 1874. No 
attempt has been made to compile a bibliography of the 
literature of ' Coral Reefs.' This was a task, as I told Mr. 
F. Darwin candidly at the outset, which my previous studies 
and present occupations would not permit me to undertake, 
and it was also one which, for the reason above given, seemed 
to me needless. I believe, however, that I have looked 
through most of the recent literature, and I have selected 
therefrom certain papers, in which, as it seemed to me, the 
arguments for and against Mr. Darwin's theory were 
stated with considerable fulness. The remainder have been 
passed over, either because they did not contain original 
information, or because they would have supplied additional 
details, on the one side or the other, rather than fresli 
arguments. In making this selection I have been influenced 


to some extent by the adventitious prominence which, 
during the last two or three years, has been given to certain 
valuable and interesting communications on this subject. 
Of the papers selected I have given a fairly full abstract, 
which represents, to the best of my ability, the views of 
their authors, whichever side they may have espoused, so 
that I trust the reader will be enabled to understand the 
present state of this difficult question, and to appreciate 
the reasons which have led some very competent authoritiea 
to maintain, and others to reject, the theory advanced by the 
late Mr. Darwin. 

It is true that, as I have stated in the conclusion, the 
close study of the question has not materially altered the 
view which I entertained when I began the task, but I have 
done my best to make my abstract a fair statement of each 
writer's case. If, then, it should appear to any one that I 
ought to have given more prominence to this point and 
less to that, I may fairly plead that this has resulted from 
deficient apprehension, and not from conscious bias. I 
have placed first, arranging them chronologically, the 
papers which are more or less unfavourable to the dis- 
tinctive feature of Mr. Darwin's theory ; iLen those which 
in the main support it. 

The following is a summary of Mr. Murray's * views : 
Very nearly all oceanic islands, other than coral atolls, are 
now known to he of volcanic origin. Hence it is probable 
that the foundations of the latter are volcanic rocks and not 
those indicative of an ancient and pre-existent land. As 
shown by the soundings of the ' Tuscarora' and 'Challenger,' 
numerous submarine elevations exist which rise from depths 
of 2,500 to 3,000 fathoms to within a few hundred fathoms 
of the surface. The upper water of the ocean (to a deptli, 

' On the Structure and Origin of Coral Reefs and Islands. Bj 
John Murray. Proc. R. S. Edin. (1880), vol. x. p. 505. 


probably, of about 100 fathoms) tcoms witli organism?, 
calcareous and siliceous; such as alga3, protozoa, hydrozoa, 
mollusca and other mGmbors of the animal kingdom : these 
are drifted by the currents from place to place ; by these 
the reef-building corals are supplied with food. It has been 
estimated, as the result of experiment, that a mass of 
ocean water one mile square and 100 fathoms deep con- 
tains more than sixteen tons of carbonate of lime.' After 
death the ' skeletons ' of these organisms are showered 
down upon the bed of the ocean. In water wliich exceeds 
some 800 or 900 fathoms in depth their remains are more 
or less affected by the solvent power of the carbonic acid 
present in the water, but at less depths they accumulate. 
Thus any submarine bank which rose within the above- 
named depth would be brought nearer to the surface, and 
its upper part, as the water above it shallowed, would be 
colonised by larger pelagic organisms ; these, after death, 
would augment by their remains the increasing pile of 
material, which at last would arrive within the bathy- 
metrical zone in which reef-building corals can hve and the 
formation of an atoll would commence. 

As already pointed out by Mr. Darwin,'' the corals on 
the outer margin of a bank grow vigorously, while the 
diminution of food and the increase of sediment tend to 
check the development of those in the inner part. Thus, 

' I estimate that this amount of carbonate of lime is equivalent 
to a solid layer of the same area which is approximately -00009 of an 
inch thick. We may arrive at it thus: taking 2-7 as the specilio 
gravity of carbonate of lime, we shall find the volume of sixteen tona 
to be about 212-4 cubic feet, or 7-8 cubic yards. This has to be spread 
out over 3,097,G00 square yards (the number of square yards in a 
mile), giving the above result. Even if we make a large allowance 
for the fact that the carbonate of lime is not solid, but in the foriu 
of an aggregate of hollow shells, I believe -O'J of au inch is in excc.-,a 
rai'icr than in defect of the truth. 

' rage 87 


while the reef is still several fathoms below the surface, 
the corals in the central part are placed at a disadvantage, 
•which becomes greater as they are left behind in the up- 
ward race by their neighbours. In a small reef, the peri- 
phery for the supply of food to the interior is relatively 
large ; thus the lagoons in small atolls are also small and 
are soon filled up, while long and narrow banks have no 
lagoons. As the reef becomes larger the conditions 
become more favourable to the formation of lagoons, for 
(as is shown by experiment) the lagoon of such an atoll 
is less rich in pelagic life than the exterior water. Thus 
growth is checked ; many species of coral die, and their 
calcareous ' skeletons ' are exposed to the solvent action 
of sea-water. When the water outside becomes too deep 
for reef-building corals to live, the debris from the exist- 
ing reef, aided by the accumulation of organisms, forms a 
talus at the foot of its submarine cliffs, and thus the reef 
spreads slowly outward, ' like a fairy ring,' on foundations 
to which its own materials have contributed. Coral-reefs 
which have been elevated for some distance above sea- 
level are frequently found to rest upon a deposit thus con- 
stituted.' The lagoon channels have in many cases been 
subsequently formed by the solvent action of sea-water, and 
the islets in the lagoon channel are parts of the original 
reef still left standing. Where the reefs rise quite up to 
the surface and are nearly continuous, there is little coral 
growth in the lagoon or its channels ; where the outer reefs 
are much broken up the growth is relatively abundani. 

' At the Admiralty Islands, on the lagoon side of the 
islets of the barrier-reefs, the trees were found overhanging 
the water, and in some cases the soil was washed away from 
their roots. It is a common observation in atolls that the 
islets on the reefs are situated close to the lagoon shore. 

' The case of Tahiti is here described ; see p. 314, where it is aia- 
cussed by Prof. Dana. 


286 APPENDIX n. 

Tliese facts point out tlie removal of matter wliicli is going 
on in the lagoons and lagoon channels.' 

Elevation, not subsidence, is to be expected in a volcanic 
region, as there is an a ^jriori reason for attributint,' the 
phenomena of coral reefs — as resting on volcanic foun- 
dations — to elevation rather than to subsidence. The 
former hypothesis appears to Mr. Murray to accord with all 
the facts indicated by the published charts of coral-reefs, 
and thus is considered by him preferable to the latter. 

Mr. Murray's general conclusions may be briefly enun- 
ciated as follows : — 

1. That foundations have been prepared for barrier-reefa 
and atolls by the disintegration of volcanic islands, and by 
tlie building up of submarine volcanoes, and by the depo- 
sition on their summits of oi'ganic and otlier sediments. 

2. That the chief food of the corals consists of the abun- 
dant pelagic life of the tropical regions, and the extensive 
solvent action of sea-water is shown by the removal of the 
carbonate of lime shells of these surface organisms from 
the greater depths of the ocean. 

3. That when coral plantations build up from sub- 
marine banks they assume an atoll form, owing to the 
more abundant supply of food to the outer margins and 
the removal of dead coral-rock from the interior portions 
by currents and the action of the carbonic acid dissolved 
in the water. 

4. That barrier-reefs have built out from the shore on a 
foundation of volcanic debris or on a talus of coral blocks, 
coral sediment and pelagic shells, and the lagoon channel 
is formed in the same way as a lagoon. 

5. That it is not necessary to call in subsidence to 
explain any of the characteristic features of barrier-reefg 
or atolls, and that all their features would exist alike in 
areas of slow elevation, of rest, or of slow subsidence. 


Professor A. Agassiz ' accepts the views of L. Agassiz, 
Le Conte, aiid E. B. Hunt, that the Florida reefs can- 
not be explained by subsidence, but that the southern 
extremity of Florida is of comparatively recent growth, 
consisting of concentric barrier-reefs which have been 
gradually converted into land by the accumulation of in- 
tervening mud-Hats, and thus explains the details of the 
process and the manner in which the foundations of the 
reefs are formed. 

He rejects Le Conte's explanation that the substructure 
of the reefs was formed by the mass of material brought 
by the Gulf Stream, pointing out that more recent investi- 
gations have shown that it ran across, not parallel with, the 
peninsula, the curve of the eastern shore of the latter being 
due to a counter-current along the reef running westwards. 
The Gulf Stream, however, has an indirect influence by 
reason of the abundant food which it supplies to animals 
living on the Bank of Florida. Across the reefs, and 
through the channels betweeH the Keys, the tides set 
strongly, bearing the mud derived from coral and other 
organisms ; this gradually accumulates to form the inter- 
vetiing mud-flats, and when swept westwards enlarges the 
submarine plateau in that direction. The Tortugas, the 
most recent cluster of Florida reefs, are at the very ex- 
tremity of the slope upon which the line of these reefs 
has been built up. Nothing among them corresponds 

' Agassiz, Alexander. The Tortugas and Florida Eeefa. Mem. 
Amer. Acad. Arts and Sci., vol. xi. p. 107, 1885. In Three Cruises of 
the Blake, vol i. (Bulletin of the Museum of Comparative Zoology, 
Harvard College, vol. xiv., 1888), a chapter is devoted to the Florida 
Keefs, As, however, the line of argument and the principal facts are 
identical with those given above, I have not thought it necessary to 
give a separate analysis. A convenient and clear summary of the 
views of Semper, Eein, Murray, and Agassiz is given by Prof. A. Geikie 
in hif, presidential address to the Royal Physical Society of Edinburgh 
in 188;> (Proc. vol. viii. p. 1). 


with the extensive mud-flat wlaich extends at a deptli of a 
few feet below the surface northward of the Keys. Where 
there is a larger accumulation of material than usual on 
the submarine plateau, so as to bring its surface within the 
depth at which corals can flourish, a reef begins to form ; 
that is the history of the Tortugas. West of it, an in- 
cipient reef may be found now in process of formation, 
east of it all the reefs in their turn have had a like origin. 
Then the deposition of silt produces mud-flats, and material 
accumulates, till at last the channels are closed and the 
whole is added to the land. From Everglades to Cape 
Sable the work may be seen completed ; on the eastern 
coast, and beyond the latter place to Marquesas Key, it 
occurs in its various stages, until at last it is shown in its 
beginning. The backbone, however, of the Florida penin- 
sula is ascribed to a fold in the earth's crust in an earlier 
geological period. As a secondary result of this a great 
submarine plateau was formed directly in the track of the 
Gulf Stream, which has since been gradually built up by 
the accumulation of marine organisms of various kinds. 
The area within the 100-fathom hne on the west coast of 
the great Florida plateau is extraordinarily rich in organic 
life ; large fragments of the modern limestone were often 
brought up in trawl or dredge, consisting of the solid parts 
of the very species which now live on the top of this 
recent limestone. West of the western shore-line Florida 
now stretches out as an immense submarine plateau, 
forming a huge tongue, coated or veneered only by 
coral limestone over its very top. The eastern and 
western edges of Florida consist of recent limestones, 
predecessors of that now forming on the western and 
southern slopes of the Florida plateau. Very probably 
the part of the peninsula north of the Everglades has 
bad in the past a like formation. Pourtales plateau ia 
built of the same species of corals and shells as now live 


upon it. Of like origin are the great bank east of the 
Mosquito coast and the reefs on the south coast of Cuba ; 
the Basse-Terre of Guadeloupe is the same, now slightly 
elevated, and the barrier-reefs on the windward side of the 
West India Islands rest on plateaux of similar origin. At 
Barbados the nucleus is a trachytic mass round which 
are terraces formed of mollusca and radiata, still living in 
the sea, which have been successively lifted.' The author 
considers that in the West India Islands many volcanic 
masses, which probably have never reached the surface, 
form the foundation of these banks of organisms. 

It would seem probable that reef-building corals had 
little to do with building the peninsula north of Cape 
Florida. The author explains the Alacran reef (atoll- 
Bhaped) by a growth of corals upwards from a submarine 
bank, and shows that the slope is steep down to a depth of 
thirty fathoms, then more gradual. 

He lays much stress on the importance of currents 
bringing food, and points out that, on the lee side of a 
reef, corals may be killed by the drift of sediment. ' When 
Darwin wrote, and when we knew little of the limestone 
deposits formed by the accumulation of the debris of mol- 
luscs, echinoderms, polyps, and the like, upon folds of the 
earth's crust, the basal parts of barrier-reefs were difficult 
of explanation. The evidence gathered by Murray, Semper, 
and myself, partly in districts which Darwin had already 
examined, and partly in regions where his theory of reef- 
formation never seemed to find its proper application, has 
in part removed this difficulty. It tends to show that we 
must look to many other causes than those of elevation 

' In Three Cruises of the Blake, voL i. p. 79, Prof. A. Agassiz says : 
' In some instances coral reefs have unquestionably been uplifted. 
1 have seen the elevated reefs of Cuba, of San Domingo, and other 
West Indian Islands, and of Barbados, which are perhaps the most 
striking examples of elevated reefs.' 


and sulisidence for a satisfactory explanation of coral-roof 
formation. All-important among these causes are tlio 
prevailing winds and currents, the latter charged "wi'.h 
sediment which helps to build extensive plateaux from 
lower depths to levels at whicli corals can prosper. This 
explanation, tested as it has been by penetrating into tho 
thickness of the beds underlying the coral reefs, seems a 
more natural one, for many of the phenomena at least, 
than that of the subsidence of the foundation to which tho 
great vertical thickness of barrier-reefs has been hitherto 
referred. It is, however, difficult to account for the great 
depth of some of the lagoons — forty fathoms— on any other 
theory tl;an that of subsidence ' (p. 121). 

The author also describes the distribution of material, 
living and dead, on the Tortugas, the action of the waves 
in pounding up dead coral, molluscs, and other organisms. 
Thus a great quantity of calcareous ooze is formed (aided 
by the material which passes through the digestive cavity of 
holothurians, echinoderms, &c.). This silt, by i.s accumu- 
lation, kills the corals, which accordingly can only flourish 
where well ' scoured.' The water is often chalk colour for 
a considerable distance from tlie reefs ; it is sometimes, 
after a heavy wind, discoloured for six to ten miles from 
the outer re'efs. This process accounts for the scarcity of 
fossils. He also expresses the opinion that in this region 
the corals do not flourish at depths over six or seven 
fathoms, being probably choked by the ooze.^ 

' In Three Cruises of the Blake, vol. i. p. 74, Prof. A. Agassiz 
states that ' all the evidence accumulated by Dana, Darwin, Ehrenberg, 
Quoy, and Gaimard tends to show that the limit of reef-building 
corals is to be found at about twenty fathoms.' 

Prof. Agassiz's views in regard to Florida do not appear to have 
met with universal acceptance among American men of science ; for 
instance, Mr. W. H. Dall (Geology of Florida, Amer. Journ. Sci., 
3rd ser. xxxiv. p. IGl, 1877) says that in the southern part of Florida 
he saw no coral-rock or coral-reef formation : ' The coral formation 


Mr. Guppy ' describes the Solomon Archipelago, which 
includes seven or eight large islands, some being from 
seventy to eighty miles in length, and the highest rising 
from 8,000 to 10,000 feet above the sea, with a great 
number of smaller islands and islets, some of volcanic and 
others of recent calcareous formations. The author found 
exploration to be difficult and dangerous, but believes that 
he saw enough to give him a fair idea of the leading types 
of structure among these islands. The observations re- 
corded in the paper may be summed up in his own 
words : — 

The islands examined indicate upheaval, in some cases 
to at least 1,200 feet. 'There are, in the first place, 
numerous small islands and islets, less than a hundred 
feet in height, which are composed entirely of coral lime- 
stone. Then there are islands of larger size, which are 
composed in bulk of partially consolidated volcanic muds, 
such as are at present forming around oceanic volcanic 
islands. Coral limestones encrust the lower slopes of 
these islands, and do not attain a greater thickness than 
150 feet. In the next place we have islands of similar 
structure, but possessing in their centre some ancient 
volcanic peak that was once submerged. Then there are 

observed by Agassiz in the region in the Keys must be of very 
limited scope, as it has not been identified from the mainland of 
Florida by any modern geologist.' Further, Prof. A. Pleilprin in a 
paper on Explorations in Florida (Transactions Wagner Inst. Sci. 
Philadelphia, May 1887), noticed in the above-named volume (p. 230), 
Bays : ' No observed facts sustain the coral theory of formation pro- 
pounded by Agassiz. They prove, on the contrary, that the coral 
tract of Florida is confined to a border region on the south and 
south-east, and there are no tertiary reefs whatever.' But he admits 
that the southern area is one of shallow sea formation, so that there 
has been a gradual uniform progressive elevation over the whole. 

' H. B. Guppy. Observations on the Eecent Calcareous Forma- 
tion-j of the Solomon Group made during 1882-4. Tr. R. S. Edin, 
xxxii. p, 545 (1881-5). 


islands in which tlio volcanic peak has become an eccentiic 
nucleus, from which line after line of barricr-rcef has been 
advanced, overlying the volcanic muds ; ' islands in whicli 
he did not find the coral limestone of a thickness of 100 
feet. Then we have the upraised atoll, such as Santa 
Anna, which within the small compass of a height of 
470 feet displavs the several stages of its growth : ' first, 
the originally submerged volcanic peak, then the investing 
soft deposit, and over all the ring of coral limestone, that 
cannot far exceed 150 feet in thickness ; lastly, we come to 
the mountainous islands formed of old volcanic rocks, such 
as St. Christoval, which, although over 4,000 feet in height, 
showed to me no calcareous envelopes at a greater height 
than 500 feet above the sea, the coral limestone crust 
being even thinner than at the smaller and more recent 
islands.' From these considerations the author concludes 
' (1) that these upraised reef masses, whether atoll, barrier- 
reef, or fringing-reef, were formed in a region of elevation ; 
(2) that such upraised reefs are of moderate thickness, 
their virtual measurement not exceeding the limit of the 
depth of the reef-coral zone, i.e. not more than about 150 
feet ; (3) that these upraised reef masses in the majority 
of islands rest on a partially consolidated deposit which 
possesses the characters of the "volcanic muds" which 
were found during the ' Challenger ' expedition to be at 
present forming aromid volcanic islands ; (4) that this 
deposit envelopes anciently submerged volcanic peaks.' 

The earlier part of the next paper ' is occupied by a de- 
scription of a reef of the Solomon group and the distribution 
upon it of coral life. According to jNIr. Guppy's observationa 
the large masses of corals usually fiourish below the wash 
of the breakers, and in these regions corals generally do not 

" H. B. Guppy. Notes on the Character and Mode of Formation 
of the Coral lieeis of the Solomon Islands. Proc. R. S. Edin. riii, 
p. 857 (Session 1885-6). 


tlirive in the break of the trade-swell. * They are only to be 
found m luxuriance on the slopes of the declivity that is 
situated in depths between five and fifteen fathoms, a decK- 
vity which may be truly termed the growing edge of the reef.' 
At exceptionally low tides, when there is a heavy sea, large 
branches are apt to be torn oft' from corals growing beyond 
the usual reach of the breakers, and these are thrown up on 
the upper flat of the reef. But in cases where the reefs are 
protected from the heavier rollers, the corals living in the 
wash of the breakers are more numerous and in greater 
variety. The same rule holds good on the lee sides of small 
coral islands. Here the corals are often grouped in irre- 
gular patches or masses, which sometimes rise with wall-like 
sides from depths of twelve or fifteen feet of water. A large 
part of the interior both of lagoons and of their channels, 
is occupied by sandy and chalky mud ; but in the shallower 
portions, and especially in those situations which are near 
the breaks in the reef, corals, especially of foliaceous and 
branching habit, thrive in great profusion. As a rule 
corals are unable to sustain exposure to the air for long ; 
from one to two hours continuously appears the maximum 
of endurance, and that is reached only by a few species. 

Mr. Guppy considers that in this group the numerous 
detached submerged reefs or shoals, which he at depths of 
from 4 to 10 fathoms (that is, at depths which vary with 
the amounts of disturbance produced by the breakers), 
represent the earliest conditions of coral reefs. Numerous 
instances of such reefs are given in this memoir : one, 
Lark Shoal, covered by water having a minimum depth of 
7 fathoms, rises from a depth of 200 fathoms. The shoal 
within the 20 fathom line measures 1^ miles in one direc- 
tion and 1 mile in the other. There is no sign of a central 
depression, the summit being comparatively level and 
covered by from 7 to 10 fathoms of water. This general 
flatness of the upper surface is not peculiar to Lark Shoal, 


but was observed at similar depths in the case of others. 
Between such submerged reefs and those marked on the 
surface by a reef flat, with its accompanying islet, or by a 
sand key, intermediate conditions were not found, and Mr. 
Guppy is of opinion that reefs on arriving within from 4 to 
8 fathoms from the surface have reached the limit of their 
upward growth, and afterwards have to extend laterally. 
Hence he infers that detached submerged reefs are unable 
to raise themselves within the limit of constructive breaker- 
action without the assistance of a movement of elevation. 
Of such a movement, in this region, there are certainly 
proofs, and the same is the case hi the Low Archipelago, 
the Fiji and Pelew groups. 

Among the reefs which have reached the surface in 
the Solomon Archipelago, fringing and barrier reefs are 
much commoner than atolls. ' A line of barrier reef pro- 
bably not much under GO miles in length, and having in- 
numerable islets on its surface, fronts the eastern coasts of 
the islands of Now Georgia at a distance of from 1 to 3 
miles fi-om the shore. At St. Christoval the fringing reefs 
occasionally reach a mile in breadth, but usually do not 
exceed a quarter of a mile. The 100 fathom line lies 
generally about 1,200 yards from the edge of the reef Hat 
which would give an average slope of 10°.' Upheaval is indi- 
cated by a recently elevated fiat of coral rock, which is in 
some places 15 feet or so above high water level. North 
of St. Christoval are three small islands (named the Three 
Sisters). They commenced their growth, according to 
Mr. Guppy, as submerged flat-topped reefs, like those 
already mentioned. They were then elevated to about 
70 feet above the sea, and have since assumed an atoll 
structure. He regards them as ' based on three submerged 
peaks which lie at some unknown distance below the sur- 
face.' They are enclosed within the same hundred fathom 
line ; the submarine slope at first is gradual and then 


descends more rapidly ; on the weather side at an angle of 
rather more than 20°, on the lee side usually at a smaller 
angle. The highest points (of coral limestone) on the 
southernmost island rise to about 70 feet above the sea. 
It has two lagoons; the deeper is about 9 fathoms, and 
over the bar at its mouth the water is less than a fathom 
deep at low water neaps. 

Mr. Guppy is of opinion that this island commenced its 
history as two flat-topped submerged reefs, and that the 
atoll form has been assumed since these have been up- 
heaved. Santa Anna is an upheaved atoll with shore-reefs 
ranging from 150 to 600 yards wide according to the steep- 
ness of ihe land. On the westward the shore-reefs enclose a 
circular lagoon 700 to 800 yards wide and 16 to 17 fathoms 
deep. Utji island has shore-reefs of varying width, and on 
the east coast the shore-reef encloses a narrow lagoon a 
mile long and 10 fathoms or less in depth. There is also 
in that on the south coast a circular lagoon about 100 yards 
wide and 6 fathoms deep, approached by a narrow channel. 
Blu is a patch of coral reef which has been raised about 
100 feet above the sea, and is still encircled by living 
reefs. Bua Sura is an atoll of elongated form about 3 miles 
in length. Except for three wooded islets on the south 
side, its circumference is either just awash at low tide or 
is within a fathom of the surface ; but soundings in the 
lagoon to a depth of 37 fathoms failed to reach the bottom. 
The islets are at highest only 15 or 20 feet above the sea, 
' cliffs of coral rock 4 or 5 feet in height ' in one ' betokens 
recent elevation of a small amount.' Eddystone Island 
consists of ' two distinct islands (volcanic) which have 
become united by elevation of an intervening coral reef. 
On the east coast the submarine slope down to the 100 
fathom line is from 30° to 35°. There is a hole in the 
middle of the reef on the west side, about 150 yards across 
and 18 fathoms deep which Mr. Guppy thinks may nuxk 


ail old crater cavity, and a smaller one in a reef on the east 
side, where also there is an elevated barrier reef. Within 
a mile or two to the south are a couple of submerged coral 
patches with level summits covered by from 5 to 10 fathoms 
of water, and in each case measuring within the latter 
contour line about half a mile in length. They both rise 
on all sides from water in which casts of 100 fathoms did 
not reach the bottom. 

East oi BoiKjainville Island, m the strait between it and 
Choiseul Island, is a submarine plateau, about 16 miles in 
width, extending from the former, and covered by from 30 
to 50 fathoms of water. This, at its outer edge, termin- 
ates abruptly in a steep slope of from 15° to 25°, • which is 
sharply delineated on the charts, by the 100 fathom line, 
and descends to considerable depths.' There is a hole 80 
fathoms deep in its generally level surface, towards the 
middle of the strait, and another, not bottomed by 100 
fathoms of line off Cyprian Bridge Island. A narrow neck 
rather over 2 miles wide links this plateau to a smaller 
one prolonging Choiseul Island. ' Broken lines of barrier 
reef (sometimes elevated) and elongated coral shoals, 
covered by 4 to 10 fathoms of water, which may be re- 
garded as incipient barrier reefs, mark the edge of the Bou- 
gainville plateau, within a few hundred yards of the 100 
fathom line.' The western extremity of Choiseul is skirted 
by a broken line of barrier reef, which encloses a lagoon- 
like channel, and supports islets on which coral rock indicates 
an elevation of at least a few feet ; and there is an island 
in the lagoon which bears similar testimony, while ' the 
hills on the coast, composed as ihey are of foraminiferous 
and pteiopod mud encrusted by coral hmestones, have been 
antecedently upheaved.' Oivia atoll, about 2 miles long, 
has been built up above a group of islets composed of 
hornblende-andesite, each probably indicating four separate 
volcanic necks. This atoll rises from depths of 40 to 50 


fathoms, with a submarine slope, varying from 12° to 26°. 
The erosion line on the volcanic rocks indicates an up- 
heaval of some four or five feet, prior to the coral growth. 
The land is bordered by extensive flats, covered by less 
than a fathom of water, on which the coral appears to be 
dead, and two basins or lagoons occur within them, about 
20 fathoms deep. 

Sections are given of some barrier reefs. As the result 
of his investigations of these, Mr. Guppy concludes that, 
from the edge of the reef flat for the first 70 or 80 yards, 
there is usually a gradual slope to a depth of from 4 to 5 
fathoms. On this but little living coral is found. Beyond 
it, there is generally a rapid descent to a depth ranging 
between 12 and 18 fathoms, on the face of which the corals 
flourish: 'this is in fact the growing edge of the reef.' 
Below this descent sand and gravel, produced by the action 
of the breakers at the margin of the reef, collect at a depth 
generally of from 15 to 20 fathoms, though sometimes this 
occurs at greater depths. One section, that of Santa Anna 
Island, exhibits two submarine clifi's, — the one, after a 
rapid slope, occurs between the depths of 16 and 32 
fathoms ; the other, after reaching a depth of about 42 
fathoms, gives a drop of 25 fathoms, after which, a slope at 
an angle of 18° or 19° descends to considerable depths ; 
corals thrive in this case at a depth of 30 fathoms. 
Another section exhibits a second but less strongly 
marked drop at about 25 fathoms. 

In an explanation of the formation of barrier reefs, 
which Mr. Guppy regards as produced successively while 
the ground is uprising, it is admitted that on this hypo- 
thesis, lagoon channels should never be deeper than the 
limit at which reef-building corals can grow. ' But as a 
matter of fact the depths inside barrier reefs as well as 
atolls, not unfrequently exceed ' 25 fathoms — in corrobora- 
tion of which statement several cases of soundings in these 


positions of 40, 50 and even GO fatlioms are mentioned. 
This difficulty Mr. Guppy overcomes by the hypothesis that 
the limit for the development of reef-building coral is really 
determined, not so much by actual depth as by the con- 
dition of the water, especially in regard to the presence or 
absence of suspended mud (p. 888). 

Another consideration confirms Mr. Guppy in his 
opinion that reefs are often begun at a much greater depth 
than 25 fathoms. The usual foundation, so far as his 
observations go, is composed of partially consolidated 
volcanic mud or ooze, more or less foraminiferous, and 
generally abounding in recent shells, and is not a layer of 
detrital sand and gravel. But in all the soimdings about 
the reef, which often extended down to 50 fathoms, the 
armings never brought up any indication of the nature of 
the bottom other than sand and gravel. Hence it may 
be presumed that such reefs — as, for example, those in 
the Shortland Islands — began at depths greater than 50 
fathoms. But if it be urged that in this case the reefs 
should be more than 100 feet thick — and this amount is 
rarely exceeded in the Solomon Islands — he replies that, 
as a rule, reef corals will be confined to depths of 25 or 
80 fathoms, and the beginning of a reef in deeper waters 
will be an exceptional thing. It must also be remembered 
that the rapid subaerial denudation which occurs in these 
regions may, in some cases, have reduced the vertical 
tliickness of the reef. 

In support of Mr. Murray's view that reefs spread by 
an outward growth, Mr, Guppy states that he found the 
corals inside the lagoons to be much larger than those 
which occur near the outward border of the reef, and in 
the barrier reefs the corals were largest near the inucr 
edge of the flat, and diminished m size as the outer edge of 
the reef was approached. ' These facts are of importance, 
Bince, according to the theory of subsidence, the central 


portion of tlie lagoon of an atoll, and the inner portions of 
the lagoon channel of a barrier reef, are more recently pro- 
duced than any other portion of the area of such reefs.' 
The smallness of the outer corals is ' only to be explained 
on the hypothesis that the reef has gradually grown out- 
wards as from a centre, and quite independently of any 
movement of subsidence.' Further, the low coral lime- 
stone cliffs, which not seldom back the present reef flats, 
are probably lines of erosion, indicative of an epoch an- 
terior to the formation of the shore reefs, when these cliffs 
were washed by the sea. The disposition both of the 
vegetation and of the humus on the wooded islets shows 
that the lee side of such an islet is its oldest portion, and 
its weather side is its growing margm. 

Illustrative of the question of the removal of dead coral 
by solution, cases are mentioiied of masses of ^nadrei^ora 
and porites several feet across, the centres of which were 
dead and were depressed a few inches below the living ex- 
terior. During the bright sunlight the increased tempera- 
ture 1 of the sea water covering the reef flats probably assists 
in the solution of the dead coral ; moreover its destruction 
by organisms, to which other authors have called attention, 
must not be overlooked. Holothurians and echinids play 
an important part in this respect. 

The author concludes by stating that the calcareous 
sand and gravel which strew the outer slopes below the 
zone of livmg coral are largely composed of reef debris, of 
the tests of Orbitolites complanata and 0. heterostegina, 
of the joints of the calcareous alga Halimeda opuntia and 
of nullipores. Of these the foraminifera were found living 
between '■1 feet and 75 fathoms ; the alga does not appear 
to live below 10 fathoms. At greater depths than 100 
fathoms the sea bottom consists generally of volcanic mud 

' About 16° F. higher than the open sea where the water was only 
8 or 4 inches deep, and 8 higher where it was a foot deep. 


(with or without organisms), wliicli however forms also at 
all depths, from a few feet beneath the low tide level, in 
the case of large islands, the coasts of wliich, owing to 
the sediment brought down by streams, are bare of reefs. 

In a paper printed in the * Proceedings of the Royal 
Society ' (vol. xliii. p. 440, 1888), Mr. G. C. Bourne gives a 
minute description of the atoll of Diego Garcia, and discusses 
the theories of coral-reef formation in connection with the 
Chagos group. In the Laccadive, Maldive and Chagos 
group, ' there is no instance of a fringing or a barrier reef; 
nothing but coral structure rises above the waves ; all the 
islands are atolls.' The three groups are beHeved to stand 
on a submarine bank lying 1,000 fathoms below the surface, 
in an ocean of an average depth of 2,000 fathoms. At 
Diego Garcia, the shores externally ' slope away very 
rapidly to considerable depths, the sounding line giving 
depths of 2;j0 fathoms and upwards at a distance of a few 
hui'dred yards from the edge of the reef,' except in one 
case. The depths inside the lagoon vary up to 19 fathoms. 
Mr. Bourne describes the different kinds of coral rock, and 
gives reasons for supposing that there has been a recent 
elevation of a few feet. He calls attention to the changes 
produced by the action of waves and currents, and to the 
effect of the latter upon the growth of coral : showing 
how the living coral may be killed by a change in a current 
which, formerly clear, now brings sand. This material 
proceeds to entomb the dead coral, and then, on a return 
to the former conditions, a new growth of coral may take 
place. upon the stratum of sand. He is of opinion that the 
subsidence theory cannot be applied to explain the Great 
Chagos Bank (see p. 53), because its rim is ' on an avei-age 
not more than fathoms below the surface, and therefore 
Biiuated in a depth eminently favourable for coral growth, 
and there are actually six islets on the northern and western 


edges, rising above tlie water and some of them inhabited.' 
He indicates further difficulties in applying the theory 
of subsidence to the Chagos Bank, especially pointing out 
that the Six Islands atoll, within a few miles distance, has 
not been affected ; still he admits that the ' Saya de Malha 
Bank appears to have the characters of a submerged atoll, 
having a central depression of 65 fathoms surrounded by a 
rim which has only 8 to 16 fathoms on its eastern side, 
but 22 fathoms on the western.' On the whole, however, 
he considers that ' most of the coral formations of the 
Indian Ocean mark p.reas of elevation rather than of rest ; 
certainly they are not evidence of subsidence.' 

In regard to the explanation of the formation of 
lagoons by solution of the interior parts of the reef, and 
by the more rapid growth of the corals on its periphery, as 
being more directly in the track of food -bearing currents, 
Mr. Bourne observes : — ' Neither of these explanations has 
completely satisfied me. That sea-water exercises a solvent 
action upon carbonate of lime does not admit of doubt, and 
that the scour of tides, combined with this solvent action 
of the water, does affect the extent and depth of a lagoon is 
obvious. But I challenge the statement that the destructive 
agencies within an atoll or a submerged bank are in excess 
of the constructive. It would be nearer the mark to say 
that they nearly balance one another. In the first place the 
carbonate of lime held in solution by sea-water is deposited 
as crystalline limestone in the interstices of dead corals or 
coral ddbris. Anyone who is acquainted with the struc- 
ture of coralline rock knows how such a porous mass as a 
viczandrina head becomes perfectly solid by the deposition 
of lime within its mass. This deposition can only be 
effected by the infiltration of sea-water. In reckoning the 
solvent action of sea-water, therefore, account must be 
taken of the fact that a not inconsiderable proportion of 
the carbonate of lime held in solution is re-deposited in tho 


form of crystalline limestono. Of this, ifc seems, Mr. 
Murray has not taken sufficient account, and lias, there- 
fore, overstated the destructive agency of the sea. Secondly, 
the growth of corals, and the consequent formation of coral 
rock within the lagoon, is generally overlooked. 

' Whilst diving for corals at Diego Garcia I had ahundant 
opportunities of studying the formation of coral rock within 
the lagoon, in depths under 2 fathoms. The layers of 
tolerably compact rock thus formed are of no mean extent 
or thickness; they soon become covered with sand, and 
are thus protected from the solvent action of the water. I 
have found it impossible to reconcile Mr. Murray's views 
with what I saw of coral growth within a lagoon. Not only 
do the more delicate branching species of the madreporaria 
flourish in considerable numbers, but true reef-building 
species, porites, mceandrina, pocillopora, and various stout 
species of madrcpora, are found there. It is a mistake 
to suppose that certain species of corals are restricted to 
the external shores, others to the lagoon. My collections 
proved that many of the species growing in the lagoon at 
distances of five miles and upwards from its outlet are 
identical with those growing on the outer reef. In addi- 
tion to them -are numerous species, such as Seriatopora 
stricta, Mussa corymbosa, Favia lohata, Fungia dentata, 
and many others that are not found on the outside. The 
reason is that the last-named are either free forms such as 
fungia, or are attached by such slender and fragile stems 
to their supports that they could not possibly obtain a foot- 
hold and maintain themselves among the powerful currents 
and waves of the open ocean. 

' These various species, numbers of which grow clo?o 
together, form knolls and patches within the lagoon, and it 
cannot be doubted that their tendency is to fill it up. 
Again, in reefs which do not rise above the surface, or are 
awash for the greater part of their extent at low tides, 


preat quantities of debris, torn from the outer slopes, are 
constantly carried over the rim of the reef and tend to fill 
it up. Hence it follows that in a lagoon entirely surrounded 
by dry land, or nearly so, as is the case at Diego Garcia, 
the tendency to accumulation of material within the lagoon 
would be less than in submerged or incomplete atolls, for 
debris cannot be swept over into the lagoon, and the only 
constructive agency is the growth of coral. If the power of 
solution of sea- water is so great, it must be supposed that 
in complete or nearly complete atolls the lagoon would be 
deepening rather than shallowing ; yet at Diego Garcia the 
lagoon is obviously shallovang in many places, and has 
nowhere increased in depth since Captain Moresby's survey 
in 1837. Indeed, the southern part seems to have shoaled 
a fathom since that time, and this is the more remarkable, 
since the S.E. trade winds are by far the most constant 
and strongest winds there and tend to accumulate material 
at the northern rather than at the southern end. The 
fact is, that these winds sweep the sand out of the 
southern part, and thus leave an area particularly favour- 
ably situated for the growth of corals. Mr. Murray 
points out that larger atolls generally have deeper lagoons 
than small atolls, and urges this fact in support of his 
theory ; but here again the facts in the Chagos group are 
against him. Victory Bank is a submerged atoll, the 
Solomons is an atoll with a large extent of dry land ; in 
each the lagoon attains a depth of 17-18 fathoms, and in 
Diego Garcia the lagoon, although far larger, does not 
attain a greater depth. Peros Banhos is far smaller than 
the Great Chagos Bank, yet in both the lagoons attain 
nearly the same maximum depth, viz., 41 fathoms for 
Peros Banhos, 44 fathoms for the Great Chagos Bank. 
Speaker's Bank is very little larger than Peros Banhos ; 
its lagoon is far shallower, having a maximmn depth of 
24 fathoms.' 


Mr. Bourne passes on to consider the opinion expressed 
by certain authors that the favourable conditions for coral 
growth in the external slopes of a reef consist in the in- 
creased food supply brought by the superficial currents of the 
ocean. This explanation, for reasons given, he regards aa 
incomplete, being of opinion that the direction and velocity 
of currents are the most important circumstances. His 
observations, he states, are confirmed in every particular by 
those made by Dr. Hichson in Celebes, and communicated 
by him to the British Association in 1887.^ Mr. Bourne 
expresses the result of his observations in the following 
words : — 

' Corals grow best in places where a moderate current 
flows constantly over them. They are killed in still water 
by the deposition of sediment, and they will not grow in 
places where a strong current sets directly against them. 
I noticed at Diego Garcia in many places, but particularly 
at the east end of East Islet, that a strong and direct 
ocean current is most unfavourable to coral growth, and 
that the reef is barren and suffering rapid erosion at such 
exposed spots as allow the whole force of the current to 
fall directly upon them. As the current parts and flows 
round the obstacle, one meets with a reef covered with 
debris, but barren of live coral ; further on, as the current 
moderates in force, one finds a few growing heads of coral ; 
and, finally, at the further end of the reef, where the 
current has abated its force considerably, there is a luxu- 
riant bed of living corals and Alcyonaria. This can be 
seen in perfection on the southern reef of East Islet. Dr. 
nickson tells me that he has observed the same facts at 
Celebes, that direct and strong currents are unfavourable 
to coral growth, that moderate tangential cuiTents aro 
extremely favourable, and sluggish or stQl water again 
unfavourable. This view, which both of us can support 
' The paper is not printed in the volume for that year. 


by many observations, is much at variance with the old 
accepted saying that corals grow best where the breakers 
are the heaviest. It appeared to me that heavy breakers 
are not favourable to coral growth, because of the quantity 
of shingle which they dash against the soft-bodied polyps. 
Some massive forms might withstand the force of breakers 
and violent currents if the polyps could be sufficiently 
protected from the shingle, but the branching madrepores 
are soon broken off and swept away, and even the more 
massive mceandrina soon follows, for whilst the surface of 
the colony grows the base is dead, is soon riddled by boring 
sponges, serpulie, &c., and is no longer able to bear the 
strain put upon it. The great mass then breaks off and is 
rolled along the reef, pounding other corals in its course.' 

Still, as a rule, the outward portions of the reef are 
the most favourable for coral growth. Hence, if a bank of 
coral be established below water, there is a tendency in the 
coral at its margin to grow both outwards and upwards, so 
that at last an atoll form is developed. As the rim 
approaches the surface, it is raised by the piling up of 
debris, broken off by the waves, and may, in some cases, 
also be upheaved. But the waves, tides, currents, &c., 
tend also to destroy parts of the island, so that there is a 
constant struggle going on between the constructive and 
destructive agencies. The author then proceeds to apply 
his theories to the formation of the Maladive atolls, remark- 
ing in conclusion : — ' However one looks at the subject one 
must realise that the laws governing the formation of coral 
reefs are exceedingly complex, and that many circum- 
stances have to be taken into account before any perfect 
explanation of their structure can be obtained.' 

While these sheets were passing through the press a 
letter written by Mr. H. B. Guppy to Mr. J. Murray ap- 
peared in Nature (vol. xxxix. p. 23G), giving some account 


of the results of a visit to Keeling Atoll ' (known also as the 
Cocos Island). As the letter is only a preliminary note, it 
is difficult to analyse or appreciate the Avritcr's arguments, 
so it may suffice to say that he is convinced that ' several 
important characters of these islands escaped the attention 
of Mr. Darwin,' that ' these features throw considerable 
light on the mode of origin of lagoon islands, and give 
no support to the theory of subsidence.' According to 
Mr. Guppy's description ' Keeling Atoll consists essentially 
of a ring of horseshoe or crescentic islands, including a 
lagoon, and presenting their convexities seawards. The 
crescentic form is possessed in varying degrees by different 
islands ; some of the smaller ones are perfect horseshoe 
atollons and inclose a shallow lagoonlet, others again ex- 
hibit only a semi-crescentic form, while the larger islands 
have been produced by the union of several islands of tliis 
shape,' He states that from the effects of gales, &c., tlie 
islands are constantly altering in shape, and expresses his 
decided opinion that the ' small atolls and horseshoe 
islands only assume their horseshoe form after the island 
has been thrown up by the waves.' This is due to the 
sand and debris, which are swept along by a current, 
accumulating under the lee of the ends of a shoal on the 
face of which the current impinges, so that the island 
tends to extend, both laterally and to leeward, and thus 
gradually to assume, more or less, the shape of a crescent 
or horseshoe. Some estimates are given of the amount of 
material transported by the currents. 

Outside the seaward edge of the present reef, Mr. Guppy 
has observed a series of submerged lines of growing corals 
separated from each other by sandy intervals. Thus the 
outward extension of a reef is effected ' not so much 
by the seaward growth of the present edge of the reef, aa 
by the formation outside of it of a line of growing corals 
' See oh. i. sect. i. of the present work. 


wliicli, when it reaches the surface, reclaims, so to speak, 
the space inside it, which is soon filled up with sand and 
Teei-debris. The evidence, in fact, goes to show that a 
reef grows seaward rather by jumps than by a gradual 
outward growth. This inference is of considerable import- 
ance since it connects all classes of reefs together in the 
matter of their seaward growth, the degree of inclination 
of the submarine slope being the chief determining factor.' 

Proceeding now to the papers favourable to Mr. Darwin's 
views we may quote first a passage in Professor Bayley 
Balfour's description of the physical features of Rodriguez ' 
which has an important bearing on one point in recent 
controversies. After stating that the island is substantially 
a hilly mass of volcanic rock, the highest point being 
1,800 feet above sea level ; that the western slopes of this 
terminate in a wide coralline limestone plain, diversified 
with elevations ; and that a fringing reef of coral, studded 
with islets, skirts the island on every side, extending on 
the west about three miles from land, but with its edge at 
the eastern end within about one hundred yards of the 
beach ; he proceeds : — ' On the south-west the central vol- 
canic ridge gradually descends, the ravines become less 
deep, and the ground spreads out into a large coralline 
limestone plain. The demarcation betwixt the limestone 
and the volcanic rock is very sharp, but isolated patches of 
limestone are met with on the surface of the volcanic 
region in the vicinity of the main mass. . . . The lime- 
stone is not found along the northern or southern shores, 
until we near the eastern extremity, where patches occur 
at the mouths of valleys, and even at some distance from 
the shore. . . . On the southern shore between Eiviere 
Palmiste and Eiviere Poursuite, indications of raised 
beaches are seen, reaching about 20 feet above the sea 
» Philosoph. Transact. R. S. vol. clxviii. (1879), p. 289. 

308 ArPENDix II. 

level. The existence of these masses of coralline lime- 
stone indicates clearly a lower level of the island, and the 
evidence of raised beaches confirms this. But a considera- 
tion of the coral reefs points as clearly to a time when 
the island stood at a higher level. The present coral reef 
fringes the coast, extending, as I have mentioned, about 
three miles on the south-west side, but coming close in 
shore on the east. An older reef, however, exists, now 
quite submerged in some places to a depth of over 90 
fathoms. Upon it the present reef rests, and it extends 
westwards nearly fifteen miles from the present coast, 
while on the east it stretches about six miles. We have 
thus proof of great and intermittent oscillations of the 
level of the island. Of the islets scattered over the reef 
some are volcanic and the others are composed of coralline 
limestone and sand. They are all within the compass of 
the present reef, and only occur on its wider parts.' Eight 
islets are of volcanic origin ; the coralline limestone and 
small islets are more numerous, and are confined to the 
southern and western reefs ; none occur on the north. 

Mr. W. 0. Crosby ^ states that level tenaces with 
vertical walls, resting against the rugged mountains of the 
interior, and forming the shore of the island, are conspi- 
cuous features in the scenery of Cuba. They may be ob- 
served at various levels up to nearly 2,000 feet. The first 
preserves a sensibly uniform altitude of about 30 feet for 
hundreds of miles. It is breached by the rivers which flow 
into the sea, and is seen to be composed of coral : in short, 
it is an elevated fringing reef, similar to that which is now 
forming on the adjacent sea bottom. It varies in width 
from a few rods to a mile or more. Sand and gravel are 
occasionally interstratified, especially near the rivers, 
showing that they are older than the reef. 

> On the elevated Coral Eeefs of Cuba, by W 0. Crosby. Troo. 
Doston Nat. Hist. See. vol. xxii. (1882-3), p. 124. 

Crosby's description of cuba. 309 

The second reef rises steeply, often vertically from the 
inner edge of the first, and along the north coast varies 
from 200 to 250 feet, being more affected by atmospheric 
denudation. It is older than the other, and the organic 
structures therein are in part obliterated by crystallisation, 
but of their identity of origin there can be no doubt. The 
third has an altitude of about 500 feet, and is yet older, 
more solid and more crystalline. A fourth reef has an ele- 
vation probably of not less than 800 feet. These ancient 
coast-reefs, with slight interruption, extend round the whole 
coast of Cuba. Moreover, the limestone plateau of El 
Yunque is considered by Mr. Crosby to be an old coral reef. 
Its top is about 1,800 feet above the sea ; its sides for the 
upper 500 or 600 feet are an almost continuous wall of 

Now, these terraces, of which the lovrest is the most 
recent, obviously prove that Cuba has been elevated, and 
they mark stages in the uprising. But there is also evi- 
dence that, at the present day, the coast is sinking. This 
is indicated by the condition of the lower part of the valleys, 
which are invaded by the sea and are filled to a consider- 
able depth with land detritus. Moreover, if El Yunque 
be an ancient reef, it is even now, after undergoing con- 
siderable denudation, more than 1,000 feet thick, and in any 
case, the third reef, mentioned above, consists of not less 
than 400 feet (in vertical thickness) of coral rock. But the 
reef-building corals do not flourish, generally speaking, in 
water deeper than about 25 fathoms. Hence, the maximum 
thickness of a coral reef would be about 150 feet, and to 
obtain even this we must assume that, from the time when 
its growth became possible, till it reached the surface, the 
bed of the sea remained at rest. Thus the conclusion seema 
inevitable that the elevation of the island of Cuba was inter- 
rupted and diversified by periods of movement in the 
opposite direction. 


Professor J. D. Dana's paper ' ' On the Origin of Coral 
Reefs and Islands,' tliougli it deals with facts already 
puhhshed more than it adduces those which are novel, is 
so important, as the work of a naturalist whose personal 
knowledge of coral reefs is perhaps unequalled, that it calls 
for a rather full abstract. Professor Dana obtained the ex- 
perience, upon which his mdependent testimony is founded, 
in the course of three years spent in travelling among 
coral reefs and islands in the Pacific, during which the 
reefs of Tahiti, the Samoan (or Na\agator) Islands, the 
Hawaian Islands, and the Feejees were examined with 
care, and fifteen other coral islands visited, ' seven of 
these in the Paumotu Archipelago, one, Tongatabu, in the 
Friendly Group, two, Taputeuea and Apia, in the Gilbert 
Group, and five others near the equator east of the Gilbert 
Group— Swain's, Fakaafo, Oatafu (Duke of York's), Hull 
and Enderbury Island.' 

Professor Dana calls special attention to the eastern 
half of the Feejee Archipelago, where several of the great 
barrier reefs, from ten to twenty miles long, have but one 
or two emergent peaks of land. Nanuku, for instance, has 
one httle point near its south-eastern angle, 'a mile of 
peak within a barrier island 200 square miles in area. 
Bacon's Isles are the last two little peaks of a si ill larger 
lagoon ... a dozen of the easternmost islands are actual 
atolls — the last peak gone.' But in case it should be 
answered that these are the emergent portions of sub- 
marine volcanos, in which case the ring-shaped barriers 
become difficult of explanation, while they are easy on the 
theory of subsidence, Professor Dana adds, that move- 
ment in this direction is proved by the existence of deep 
fiord-like indentations in the rocky coasts of islands, both 
of those inside of barriers, and those not bordered by reefs. 
As examples of this structure, generally admitted to be 
' Anier. Jour. Sci. (1885), Ser. III. vol. xxx. pp. 89, 109. 


one of the strongest evidences of subsidence all the world 
over, he quotes the Marquesas Islands with the Gambler 
and Hogolen Islands, Eaiatea and Bolabola of the Tahiti 
Group and the Exploring Isles of the Feejees. Professor 
Dana also calls attention to the general parallelism between 
the average trends of coral islands and the courses, not 
only of the groups of which they form part, but also of the 
groups of high islands not far distant,^ and refers to the 
arguments drawn by Mr. Darwin from the fact that the 
larger coral islands have the same diversity of form as 
is found in the barrier-reefs of high islands and exhibit 
groupings such as would result from the sinking of a large 
island of ridges and peaks with encircling reefs. The 
depth of the lagoon, and of the channels inside of barrier 
reefs — in many cases two or three times greater than 
twenty fathoms — is very difficult to explain if there has 
been no subsidence ; so is that of the ocean near to atolls. 
Professor Dana, after noticing one or two considerations 
of a general nature, points out that ' if an atoll reef is not 
undergoing subsidence, the coral and shell material pro- 
duced which is not swept away and distributed by currents 
serves: (1) to widen the reef; (2) to steepen, as a conse- 
quence of the widening, the upper parts of the submarine 
slopes; (8) to accumulate, on the reef, material for beaches 
and dry land ; and (4) to fill the lagoon. But if, while 
subsidence is in progress, the contributions from corals 
and shells barely compensate for the loss by subsidence 
and current waste, the atoll-reef, unable to supply suffi- 
cient debris to raise the reef above tide level by making 
beaches and dry land accumulations, would (1) remain 
mostly a bare tide-washed reef; (2) lose in diameter or 
size because the debris that is not used to keep the reef at 

' This, however, I conceive, would not offer a difQculty to those 
who advocate submarine volcanic masses as a foundation for the 


tide-level is carried over the narrow reef to the lagoon by 
the waves whose tlirow on all sides is shoreward ; (3) lose 
in irregularity of outline and thus approximate towards an 
annular form ; (4) lose the channels through the reef into 
the lagoon by the growth of corals and by consolidating 
ddbris ; and (5) become at last a small bank of reef-rock 
with a half- obliterated lagoon basin. 

' The Pacific contains reefs of the three kinds : (1) atolls 
with much of the reef overgrown by trees and shrubbery ; 
(2) others, of large and small size, with the reefs mostly or 
wholly tide-washed ; (3) others, only two or three square 
miles in area without lagoons. Further, the different kuids 
are generally grouped separately and gradationally : (1) 
the islands of the Paumotu and Gilbert Archipelagos have 
usually half or more of the reef dry and green ; (2) the 
northern Carolines and the northern Marshall Islands and 
the eastern Feejees, although in fact^of large size, are 
mostly bare reefs ; while (3) the islands of the Phoenix 
Group, of the equatorial Pacific east of the hne of 180°, 
are, with one exception (Canton or Mary), not over four 
miles long. The three more southern of the Phoenix 
Islands, Gardner's, Hull's and Sydney, between 4° 25' S. 
and 4° 85' S., are two tofoiu: miles long and have lagoons; 
five (islands), including Phoenix, Birnie's and Kean's be- 
tween 3° 10' S. and 3° 30' S., and Howland and Baker's, 
north of the equator, are a mile and a half, and less, ra 
length, and have depressions at the centre but no lagoons. 
The depressions contain guano, and one of them, Kean's, 
has much gj-psum mixed with the guano ; Kean's and 
Phoenix have a foot or two of water at high tide, the tide 
rising six feet. Another of the number, Enderbury's, is 
three miles long, and has a half-dried lagoon which is 
very shallow and has no growing corals. To the north 
of these islands for fifteen degrees of latitude, the sea is 
an open one, and in the next ten degrees, to the line of 


tho Hawaian chain, the only islets not marked doubtful 
are " Coral Eeef, Awash " and Johnston Island. A similar 
gradation in size takes place in the Ellice, Eatack, and 
many other groups of the ocean.' Smallness of size and 
dried lagoon basins, ^yith occasionally a deposit of gypsum 
from evaporated sea water, are just the results which should 
be expected if the cause which had regulated the coral 
growth had been subsidence; and gradation in it would 
result from gradation in the amount of subsidence. 

Professor Dana states that he also came to the conclu- 
sion (and this appeared to be sustained by the ' Tuscarora ' 
and ' Challenger ' soundings) that the belt of maximum 
subsidence in the Pacific ran from the south of Japan in 
a south-easterly direction, passing south of the Marquesas 
Group towards Easter Island. The 'Tuscarora' sound- 
ings indicated that transverse to the trend of the Phoenix 
Islands {i.e. transverse to the belt of maximum sub- 
sidence), the mean submarine slopes appear to be 1 to 1*5 
or 1 to 1-7 (the former being nearly the maximum slopes 
of Cotopaxi, Mount Shasta, and several other volcanic 
summits of Western America), while the slopes along the 
trend are much less. This fact is more in accordance 
with a theory of extensive subsidence than of extensive 

Subsidence also is indicated by the deeply indented 
shores of the Marquesas Islands, but here, probably owing 
to the boldness of the coast line, reefs are few. Tahiti on 
the contrary affords no direct proof of subsidence, and 
none of elevation, beyond that of one or two feet. But its 
broad reefs are favourable, in Professor Dana's opinion, to 
the idea of subsidence, and he suggests that it has amounted 
to about 45 or 50 fathoms. In one island of the Navigator 
Group the indented shores seem to favour a local sub- 
sidence, but in the others there is no direct proof of 
movement in either direction. Subsidence, however, is 


indicated by the broad reefs, barrier islands, and atolls 
of the Feejee groups. 

Elevation undoubtedly has occurred in several localities, 
e.g. in the Austral, Hervey, Friendly, and even in the 
Sandwich Islands, but in all the amount is small — not, so 
far as he knows, anywhere exceeding 300 feet. These Pro- 
fessor Dana considers to be merely local phenomena, and he 
passes in review several facts showing the uncertainty of 
evidence as to little or no subsidence, or as to recent ele- 
vation from narrow reefs and the volcanic character of 
islands. Further, these local elevations in coral seas, 
where they do occur, are spread over very large areas. For 
instance, the Paumotu Archipelago, consisting of more than 
eighty atolls and two barrier islands, contains only three 
or four atolls that are over 12 feet high. Of these Mctia 
is 250 feet high. Dean's, probably at its highest 15 or 20 
feet, is 60 miles to the N.N.E.; far to the S.E. of that, 
nearly 1,450 miles from Metia, isEhzabeth Island, 80 feet. 
• Locate these points on a continent, and Pacific distances 
and the length of Pacific chains of atolls will be appre- 

Professor Dana next reviews the arguments in favour 
of hypotheses other than that of subsidence, and deals with 
the soundings of the ' Challenger ' off Northern Tahiti, upon 
which' great stress has been laid by Mr. ]\Iurray and others. 
Here, from the edge of the barrier reef, the sea bottom, 
covered partly Avith growing corals, deepened gradually in 
about 250 yards to 40 fathoms {i.e. to considerably below 
the depth at which reef-building, corals usually can grow), 
then from this Hmit the bottom dropped down in about 
100 yards to 100 fathoms ; at first precipitately at an angle 
of 75°, then more gradually, but above 45° ; and for 
another 150 yards the sea bed still shelved down at 30°, 
but beyond this the slope diminished in the course of a 
inile to 6°, where at last the depth was 590 fathoms. 



In other words, we may regard this part of Taliiti as a 
submerged momitain 8,G00 feet high. Up to a contour 
line of 3,000 feet the ground shelves upwards, at first 
gently, then more steeply, till it attains a slope of 30° ; 
from the top of this rises a hne of cliffs about 350 feet high 
crowned by a slope of which the angle is 18°. The craggy 
zone is strewn, we are told, with large masses of coral — 
like a talus beneath a line of cliff— mingled with fine debris ; 
in about 100 yards there is only sand, which continues to 
the lower part of the mountain, where it gives place to mud, 
composed of volcanic and coral sand and various organisms, 
generally minute. 

Great stress is laid on the occm-rence of this area of 
coral crags and ' screes ' as indicative of the mode in which 
a reef is enabled to grow outwards on a foundation, built 
from its own ruins. Such a mode of enlargement (as Pro- 
fessor Dana points out) had, however, been obviously ad- 
mitted as possible in particular cases, ^ and so cannot be 
regarded as contrary to the general hypothesis put forward 
by Mr. Darwin. But he calls attention to the fact that the 
above observations prove : (1) that the currents romid Tahiti 
are evidently weak because they carry little coral debris so 
far as a mile from the edge of the reef ; (2) that very large 
masses of coral are lying about below the submarine cliffs 
at depths of from 240 to 600 feet, i.e. far below the depth 
at which the waves could exert any serious rending force. 
The position of these blocks, always below 240 feet— too 
far from the edge of the reef to have been borne from it 
and washed at last over the brow of the steep declivity — ■ 
seems only to be explicable when it is regarded as indica- 
ting a stage in the past history of the reef, and is a me- 
morial of a time when this declivity was the edge of a 
growing reef, and its brow was beaten by the waves.^ 

' See pp. 22, 67, &c. of the present volume. 

' Professor Dana considers that waves do little rending below the 

316 APPENDIX 11. 

The case at Tahiti appears to ho confirmed hy other 
instances ; such as Captain Fitzroy's sounding (in the 
* Beagle ' Voyage) at Keehng Atoll, 2,200 yards from the 
breakers, when no bottom was found at a depth of 1,200 
fathoms ; and the soundhig by Captain Wilkes off Clermont 
Tonnerre (Paumotu Archipelago), where 'the lead brought 
up an instant at 350 fathoms, then dropped off and de- 
scended to 600 fathoms, toming up bruised with small 
pieces of red and white coral attached ' ; as well as that by 
the same 'a cable's length from Ahii,' where the lead 
struck a ledge of rock at 150 fathoms, and brought up 
finally at 300 fathoms. Still, it would be well that the 
older soundings should be repeated, and the subject be 
more fully investigated.' 

In regard to Professor Agassiz's argument that the 
Florida reefs are the result of drifting material (see p. 287), 
and in no way require or indicate subsidence, Professor 
Dana points out that there is little in the great barrier reef 
of Eastern Australia, which has some correspondence in 
position with the sand reefs off eastern North America, to 
suggest a similarity of origin. Full of irregularities of 
direction and of interruptions, it follows in no part an even 
line. In the northern part, the barrier, while varying 
much in its course, is barely 30 miles from the land ; in the 
southern half it extends out 150 miles from the coast, and 
includes a large atoll-formed reef. Further, in the Pacific 
Ocean, the trends, whether of coral island groups, or of the 
single islands, frequently do not correspond with the direc- 
tion of the oceanic currents, or indeed of any cijrrent 

depth to which they can bare the bottom, so as to obtain an effective 
broadside stroke, which he thinks rarely exceeds in the most extreme 
cases 20 feet vertical. At 240 feet he believes the disi^lacement of 
the water would be at most only a few inches, and thus the battering 
power would obviously be 7iil. 

• See below, p. 319, for an account of Masamarhu Island in tha 
Bed Sea. 


wliich is not determined by tlieir existence. Moreover, 
to prove formation from drifting does not suffice to dis- 
prove subsidence. The length of Sandy Hook. varies in 
consequence of the action of currents, yet this does not 
disprove Professor Cook's conclusion that the New Jersey 
Coast is subsiding. 

Further, it must be borne in mind that subsidence has 
undoubtedly taken place in the region around Florida, 
though at present we can only prove this to have lasted into 
the earlier Pleistocene, the difficulty of dealing with this 
being augmented by the occurrences of elevated coral reefs, 
in Jamaica at 2,000 feet above the sea, and in Cuba perhaps 
even at the same, and certainly at 1,000 feet. The former, 
while obviously proving elevation, are considered by Mr. 
Crosby to prove also the occurrence of epochs when move- 
ment was in the opposite direction (see p. 309). 

The next section calls attention to the vague character 
of the evidence adduced for the building up from deep 
waters of important banks, composed of organic debris, to 
serve as the foundations of a coral reef, though doubtless 
such a thing may occasionally occur. But in regard to the 
existence of submerged mountain masses, which have been 
indicated as suitable for the like purpose, it may be well 
to give Professor Dana's quotation of the argument, and 
his comment thereon. Mr. Murray ' observes that the 
" soundings of the Tuscarora and Challenger have made 
known numerous submarine elevations : mountains rising 
from the general level of the ocean's bed, at a depth of 
2,500 or 3,000 fathoms up to within a few hundred fathoms 
of the surface." But " a few hundred fathoms," if we make 
" a few " equal 2, means 1,200 feet or more, which leaves a 
long interval yet unfilled.' 

Lastly, Professor Dana reviews the proposed explana- 
tion of the ring-like shape of atolls, and of the channels 
which exist in the enclosure of atolls or between atolls ; 


and points out that the larger the atoll the purer the sea- 
water of the lagoon, so that the latter would speedily reach 
a limit to its expansion in consequence of the non-growth 
of coral; nideed, as a matter of fact, these larger lagoona 
contain plenty of living coral. The second (and ultimately 
more important) factor in the enlargement of a lagoon, 
viz. the corrosion of the dead coral by the solvent action 
of water, he regards as a hypothesis which has little direct 
evidence in its favour and much indirect against it. The 
connection of channels with prevailing currents cannot be 
shown, and the former in many cases are sufficiently deep 
to be well below the limit of abrasion. Indeed, as a rule, 
so far from these channels being enlarged by solution and 
abrasion, they tend to be closed by the growth-of living 
coral, and many of the lagoons in the smaller islands are 
without channels. Hence in them, as there can be no 
appreciable transference of water, the action of solution 
must be reduced to a minimum. Yet these closed atolls 
are by no means exceptional. For instance, in the case of 
about sixty coral islands mapped by the Wilkes expedition, 
of those which range from 1^ to 3 miles in the longer 
diameter of the reef, nine have no lagoon, but only a small 
depression in its place, which is dry, except in the case of 
two where water gets in at high tide. Of those under 6 
miles in length, having lagoons, 17 in number : 16 have 
no entrance to the lagoon at low tide ; the other has an 
entrance of large size. Of those 6 miles or over in length, 
29 in number : 17 have channels and 12 have none ; those 
having channels are generally over 10 miles long. It 
must also be understood that the opponents of the subsid- 
ence theory are compelled to admit it in order to explain 
the depth of certain lagoons. Hence Professor Dana con- 
siders that the hypothesis of elevation or lateral spreading 
during a period of rest is inadequate as a general expla- 
nation of the problem. 

masXmarhu island. 319 

To tlie arguments advanced by the author ah'eady 
quoted we add the diagrams annexed, for the use of which 
we are indebted to the courtesy of Messrs. Macmillau & Co. 
They appeared in Nature (voL xxxvi. p. 413), illustrating a 
communication from Captain Wharton, and represent two 
sections on a true scale, made by Captain Maciear (H.M.S. 
'Flying Fish,' of the slope of the coral reef surrounding 
the small island of Masamarhu, situated in the Eed Sea 
hi lat. 18° 49' N., long. 88° 45' E. The dotted lines show 
where the soundings were obtained, and the words indicate 
the nature of the bottom. 

It will be observed that there is a remarkable and 
significant correspondence between these two sections, 
which, as the plan indicates, are taken nearly half a mile 
apart. In each the surface of the fringing-reef, after 
shelving very gently downwards to a depth of about three 
or four fathoms, is bounded by a submarine cliff. This in 
one section (No. I.) continues almost unbroken to a depth 
of about 500 feet, except that a kind of ledge or terrace is 
clearly indicated at a depth of rather less than 100 feet. 
In the other section (No. II.) the foot of a great submarine 
cliff is found at about 500 feet, but in this case the cliff" is 
distinctly divided into two precipices by a shelving bank oi 
coral and sand, which begins at a depth of about 140 feet 
and reaches the brow of the lower precipice at about 260 
feet. This bank is covered by ' sand and coral.' At this 
depth in each section the island is, as it were, defended 
by a deep and narrow ditch, the edge of its steep glacis 
being formed by a sharp arete of coral which in one case 
rises into soundings of about 250 feet. From this the 
former section shows a second rapid fall down to another 
ditch, the bottom of which lies more than 1,200 feet below 
the sea level. This in section resembles the other one, 
and the height of its counterscarp is more than 300 feet. 
From the edge of this the glacis for a short distance is nearly 

No. I. 

No. IT. 

Scu/e of feet 

Sketch of MasAmarhu I. 

showini; .Tvptoximate 

position nf .Sections 



level, and tlien descends at an angle of some tliirty degrees. 
In the lower diagram we find no indication of this second 
ditch, but a long slope begins at the foot of the submarine 
cliff at a depth of about 850 feet, which is very nearly 
identical with that of the flat part of the glacis in the 
former section. 

It will be observed that the upper ditch (that common 
to both sections) has its bottom at a depth of full 500 feet, 
or about 85 fathoms — that is, at more than three times the 
average depth at which reef-building corals cease to live, 
while the least depth of the final submarine slope is 850 
feet, or more than 140 fathoms. These ditches seem irre- 
concilable with any idea of an outward-spreading growth 
of the reef, and must, I think, be indicative of a subsidence 
which isolated the outward and more flourishing edge of 
a shore reef, and progressed rather too rapidly to allow 
its corals to extend across the trench thus formed and effect 
a union with the main mass. Of course if a fissure-like 
hollow were once established between two masses of growing 
coral in a subsiding area, it would not be readily filled up, 
unless the edge of its outer wall were sufficiently near the 
surface to sufl'er much from the violence of the waves. 

The former section seems to me inexplicable under the 
conditions ordinarily admitted for coral growth, unless we 
suppose that the bottom of the lower ditch, now at a 
depth of over 1,200 feet (200 fathoms), was formerly situ- 
ated within about 25 fathoms of the surface ; so that a sub- 
sidence of more than 1,000 feet may fairly be claimed for 
the coral reef of Masamarhu. 

I'rofessor Dana,' in an article which appeared while 
this sheet was in the press, adduces some new and very 

' Points in the Geological History of the islands Maui and Oahu. 
By J. D. Dana, Anaer. Jour. Sci. vol. xxxvii. p. 81 (February 1889). 
I am indebted to my friend Professor Judd for calling my attention 
to this article. 



important evidence in regard to the Sandwich Islands. 
Oahu, the island on which is the town of Honolulu, gives 
indications of a recent upward change of level, amounting 
to 60 feet at least on its northern, and about 20 feet les3 
on its southern side. But this is not all. Several deep 
borings have been made in different parts of the island, 
the particulars of which are recorded. The following may 
be taken as examples. 

I. James CamphelVs Well, at xvestfoot of Diamond Head, 
not far from sea-level. 

Gravel and beach sand . 

Tufa like that of Diamond Head 

Hard coral rock, like marble 

Dark brown clay . 

Washed gravel 

Deep red clay 

Soft white coral . 

Soapstone-like rock 

Brown clay and hroken coral 

Hard blue lava 

Black and red clay 

Brown lava . 





. 50 . 

. 270 . 


. 505 . 


. 75 . 


. 25 . 


. 05 . 


. 28 . 


. 20 . 


. 110 . 


. 45 . 


. 28 . 


. 249 . 


II. King's Well No. 2, about half a mile west of Diamond 
Hill, and 350 yards from the sea-shore. 

Sand and coral . 
White coral rock . 
Yellow sand . . 
Hard lava . . 
WJdte coral rock 
Blue clay 

Tough clay and coral 
Blue clay 
Hard coral rock 
Soft coral , 



33 . 



43 . 


47 . 


110 . 


25 . 


05 . 


30 . 


40 . 


SO . 


masXmarhu island. 


King's Well, etc. — continued. 



Tough clay 

5 . 


White coral rock 

. 40 . 


Tough clay 

. 30 . 


WJiite coral rock 

. 100 . 


Tough clay 

5 . 


Coral and clay 

. 70 . 


Tough clay 

. 2S . 


Black sand 




. 120 . 


III. Wellin 

Thomas Squai'e, Honolulu 






Soil 6 feet, with 6 feet of black sand, 


clay 4 feet . 


. 16 . 

White coral rock 

. . «. 

. 20O . 


Brown clay 

a . . 

. 44 . 


Coral rock 

a • • a 

. 10 . 


Brown clay 

• • « « 

« GO . 


White coral rock 

. o a 

. 50 . 


Brown clay 

. 80 . 


Bed rock or lava, penetrated . 

. 4'J . 


The evidence of these borings, which is corroborated 
by others quoted in the paper, points to a very consider- 
able subsidence in this region, to the amount of at least 
800 feet, and in all probability of considerably more than 
1,000 feet. Moreover, the 'hard coral rock, like marble' 
(No. I.) can hardly be anything but a ' fossil reef ; the base 
of this, it will be observed, after some upheaval, is even now 
at a depth of 825 feet (137^ fathoms), and the reef has a 
continuous thickness of 505 feet (full 84 fathoms). 

The above abstracts may suffice, I hope, to give a fair 
representation.of the arguments for and against Mr. Dar- 
win's theory, which have been advanced during the last 
fourteen years. That theory is regarded by some enthu- 
Biastic opponents, as already on the threshold of the limbo 


appointed for exploded hypotheses. To this opinion T 
cannot declare myself a convert, for reasons which, in 
conclusion, I shall endeavour to indicate. 

First, however, I may remark that certain of Mr. 
Darwin's critics occasionally appear to have perused his 
book with overmuch haste, and to have overlooked the 
fact that he admits such possibilities as local upheavals, 
the lateral growth of reefs, and modes of formation similar 
to those asserted for the Florida reefs ; ' that in short, most 
of the causes on which stress has been laid by his critics 
have been already noticed by him, so that he differs from 
them, not in overlooking such causes, but in assigning 
to them a subordinate value. Moreover, it may not be 
unfair to call attention to the want of unanimity among 
his opponents : some advocating solution as a primary cause 
in the shaping of atolls, while others rely chiefly on the 
mode of growth of the polyps. Such a divergence obviously 
does not prove Mr. Darwin right, but it does indicate that 
as yet no other hypothesis has been able to secure a general 
acceptance, and that the problem still demands the exercise 
of cautious induction, which was iiis method of procedure, 
and does not justify the over-confident boldness of assertion 
which has characterised at least one critic of his work. 

The chief arguments which have been advanced against 
Mr. Darwin's theory, as it appears to me, may be tlius 
summarised : — 1. That such evidence as can be obtained 
in regions where extensive coral reefs exist is favourable 
to upheaval rather than to subsidence. 2. That lateral 
growth is a most important factor in the formation of a reef, 
the polyps, as they aavance, being supported on a founda- 
tion composed partly of the broken fragments of the reef, 
partly of other marme organisms, and that by means of the 
latter deeply submerged banks are sometimes augmented 
vertically until they are brought within the zone of reef-coral 

' See pp. 22, 23, TJ, 120, 121, 171, etc. 


life. 8. That lagoons and lagoon channels are materially 
enlarged by the destruction of dead coral through the 
solvent effects of sea-water. 4. That in the past history 
of the earth we find no evidence in favour of the formation 
of coral reefs in areas of subsidence, or in other words 
that fossil coral reefs are less than some 25 fathoms thick. 

1. Much stress is evidently laid upon the fact that 
many coral islands afford evidences of a certain amount of 
upheaval. This amount, in most cases, is but slight, and 
its significance appears to me to have been exaggerated. 
Undoubtedly, it proves that the record which is the most 
obvious indicates an upward and not a downward motion, 
but in so doing it introduces a difficulty which will 
presently be noticed. These indications, however, do not 
of themselves prove a general upheaval, but only oscilla- 
tion. Every geologist is aware that movements in any 
given direction are frequently neither uniform nor contin- 
uous. For instance, no one doubts that the western coast 
of Scandinavia, and, in a less degree, that of Great Britain, 
have very considerably subsided since the sculpture of their 
leading physical features, and yet from the Land's End to 
the North Cape we constantly find proofs that the latest 
movements have been in an upward direction. Even in 
the case of the more important, but much rarer, upheaval 
of reefs, as at the island of Cuba, the coral masses are so 
thick that we must assume the practical arrest of all up- 
ward movement during tlie growth of the reef. In this 
case also, if the coral reef be only a sort of cap concealing 
a hill of pre-existent rock, we may reasonably be surprised 
that the ' ashlar-work ' of coral limestone has in no case 
BO far yielded to the action of the atmospheric agencies as 
to lay bare its inner support. 

Doubtless there are many reefs to which either explan- 
ation might be applied, but there are some which, unlesa 
coral polyps can build at depths much greater than 25 


fathoms, can only be explained by subsidence. It is sought 
to elude this difficulty by supposing that the reef builders, 
under specially favourable circumstances, may commence 
operations at depths considerably greater than the usual 
limit. It is indeed true that reef corals are sometimes 
dredged alive from depths much exceeding 25 fathoms, 
but the result of all recent researches has certainly been 
to confirm the general correctness of this bathymetrical 
limit, and the proposed evasion of the difficulty is at 
present a mere hypothesis, which bears a suspicious resem- 
blance to the epicycles devised to prop up the Ptolemaic 
system of astronomy. 

While the existence of ' continental rocks,' as they may 
be called, in oceanic islands would have almost proved a 
general subsidence, I do not see that the frequent occur- 
rence of volcanic rocks is seriously opposed to it. The 
arrangement of the majority of coral islands, whether 
wholly composed of organic material, or incrusting a 
nucleus of volcanic rocks, is indicative of lines of weakness 
in the earth's crust, which would give rise to movements 
in either direction, and in each case the islands would be 
connected with extruded masses of volcanic rocks, ejected 
at various points along these lines.* Thus, we have to 
consider which of two hypotheses is the more probable : 
(a) that mounds thus formed have, in the majority of 
cases, failed to reach the surface, but have nevertheless 
generally arrived within a comparatively short distance of 
that goal; or (6), that they, after having in many cases 
overtopped the surface, have again subsided. The latter, 

' It must not be forgotten that though the peaks of mountain 
ranges are frequently composed of ' continental rocks,' instances are 
by no means wanting, as in the Andes, Caucasus, &c., where the 
higher portions are volcanic. In more insulated mountain masses, 
as those of Etna, Kilimanjaro, Ararat, of some of the islands of the 
Malay Aixhipelago and on the western coast of North America, wo 
have instances of volcanoes forming the highest part of the land. 


I must confess, seems to me tlie more probable, especially 
when we remember that subsidence very commonly occurs 
in a district when it has recently ceased to be the scene of 
volcanic disturbances on a large scale. 

2. In regard to the lateral spreading of reefs, like a ' fairy 
ring,' as it has been happily expressed, there is no doubt 
that, as has been admitted by Mr. Darwin,' some augmen- 
tation may occur in this way ; but to regard this as a factor 
of prime importance in the development of a reef seems tome 
to import new and serious difficulties. Let us assume that 
the submarine mound or shoal on which the reef is founded 
remains at rest during the whole period of the growth of the 
latter, and that this commences on the area (regarded, for 
simplicity, as a plain) included within the bathymetrical 
contour line of 25 fathoms. For a considerable period, 
until the edge of the reef arrives within a few fathoms, pro- 
bably less than ten (see p. 315), of the surface of the sea — • 
that is, for full three-fifths of its w^hole vertical growth — 
the exterior slopes will only be augmented by the accu- 
mulation of marine organisms, a process which cannot be 
rapid. Hence, for a considerable time, until the reef 
itself has completed the greater part of its growth, and 
begins to augment the talus with its own ruins, the process 
of laying the foundation for a new coral growth, and thus 
the lateral spreading of the reef, will be slow. 

Consider, then, the case of a reef where this process haa 
begun, and for simplicity regard it as a cylinder cap- 
ping a flat-topped cone. Obviously, if the reef begin to 
spread laterally, the volume of the foundation required to 
support the new growth increases far more rapidly than 
the area from which material can be supplied. Hence, 
as the reef advances outwards, the rate of increase will 
rapidly diminish, unless we suppose either an extraordi- 
nary annual destruction of growing coral, or an increased 
' See pp, 22, 67, 70 of this work. 


accumulation of otlier organisras. Moreover, unless wa 
rely on solution for enlarging the lagoon, this will remain 
of its original size, and thus will be small in comparison 
with the ultimate area of the atoll. No doubt, for a time, 
as the reef is approaching the surface of the sea, the more 
rapid growth of the coral at its outer margin will cause 
it to be saucer-like in section, and thus somewhat enlarge 
the lagoon, but as soon as the upward growth ceases this 
process is arrested and the atoll can only spread laterally 
and thus must increase in breadth, while the lagoon, if 
there be no solution, tends rather to diminish in size. 

It is, however, stated on good authority ^ that coral 
growth, as a rule, is by no means entirely arrested in 
a lagoon, and we cannot suppose that so long as there is 
free passage for a considerable stratum of water above the 
reef — i.e. so long as there are soundings of 8 or 10 fathoms 
over it — the polyps on its inner part will suffer materially 
from want of food or properly aerated water. Hence the 
lagoon will not be formed at all mitil the reef has made 
some progress upwards, so that it should always be com- 
paratively shallow, not exceedmg a few fathoms in maxi- 
mum depth. From the above considerations it appears to 
me that the ' fairy-ring ' hypothesis is inadequate unless 
it be inseparably linked with that of ' solution.' 

At this period we may not unfitly notice another con- 
sideration which has been urged, viz. that many shoals, 
chiefly of volcanic origin, which lie at too great a depth to 
be colonised by reef-building polyps, may be raised up to 
the proper level by the accumulation of marine organisms. 
That this may sometimes occur cannot be denied, but it 
must be remembered that, unless the shoal lie at a very 
moderate depth below the required level, the process 
of accumulation will be extremely slow. Mr. Murray's 
estimate of the quantity of carbonate of hme present in the 
' See pp. 302, 318 of this work. 


minute organisms wbicli inhabit the upper stratum of the 
ocean water seems at first considerable, but when we 
estimate its thickness in a given area, this proves to be 
extremely small. Hence, unless we assign a very brief 
existence to each individual, and thus suppose a heavy rain 
of sliells on the ocean floor, the foundation for the future reef 
will rise but slowly, and its initiation, in the case of those 
which now exist, must be carried back to a rather remote 
epoch. Here, again, we may inquire whether a cause, which 
must not be wholly overlooked, has not, through an error 
in mental perspective, been brought into undue prominence. 
3. The solution theory, which indeed by no means 
meets with universal acceptance among Mr. Darwin's 
critics, appears to me beset with considerable difficulties. 
The solubility of carbonate of lime in ocean-water cannot 
of course be denied ; but is there satisfactory evidence that 
this is a factor of primary importance to the case of a 
coral reef? The apparently rapid solution of calcareous 
organisms at great depths has but httle bearing on what 
occurs at small depths, and the good preservation of the 
' globigerina ooze ' down to depths of some 2,000 fathoms, in 
itself indicates that solution to any important amount takes 
place under very exceptional conditions. The rottenness 
frequently noted in dead coral is mainly due to the decom- 
position of the animal tissues with which the mineral 
constituent is incorporated : thus the process is one of 
disintegration more than of solution. The dead coral is 
no doubt to some extent dissolved, but it mainly forms a 
sand or mud. This of course, in some cases, will be swept 
out by currents mto the open ocean, and thus the coral 
will be removed from its place of growth, but it may well 
be doubted whether this substitute for a true solution will 
be for long a factor of prime importance in the genesis 
of a lagoon. There is moreover some evidence directly 
opposed to the theory of solution at a moderate depth, as, for 


example, the blocks of recent limestone which were dredged 
by Professor A. Agassiz off the Florida reef.' Under what 
circumstances, then, will the sea- water act as a solvent on 
the dead coral ? I think we must reply. When the fluid 
is rather rapidly altering its position in regard to the sub- 
stance attacked. Thus rain and streams are important 
solvents, and so might be breakmg waves or tidal ebb and 
flow, but when the water is at rest or is only spreading 
with a slow, diffusive movement, its solvent action is ex- 
tremely slight. For instance, chalk often is, and must 
often have been, saturated with water, yet numbers of the 
minute organisms which enter into its composition are still 
perfectly distinguishable. The same is true of many 
other Hmestones ; indeed the effect of water often seems 
conservative rather than destructive. It smks down into 
the body of the rock, carrying with it the carbonate of Hme 
which has been obtained h-orn the exposed superficial part 
of the mass, but on reaching the level of saturation, when 
it only percolates by diffusion, it commonly deposits its 
burden, fiUmg up with mineral calcite the interstices of the 
organic materials. Hence the comparatively quiet waters 
of a lagoon would be favourable to the consolidation rather 
than to the destruction of the dead coral, save only within 
a very limited distance fi'om the surface. Moreover, the 
remains of organisms, when once the interstitial animal 
tissues have been replaced, appear to be less soluble than 
the other parts of a rock, as is indicated by the familiar 
' weathering out ' of fossils. Eeef rock also appears very 
apt to assume a solid and semi-crystalline condition 
(p. 17), and in regard to this we must not overlook n 
peculiarity of coral which, as it seems to me, has an 
important bearing on the subject. Dead coral is very 
readily converted into dolomite, which is a much less 
BoluMe salt than calcite. Further, the conditions which 
' See p. 288 of this work. 


would prevail in a lagoon, v/lien its waters had become 
unsuitable for coral life, would be those which would be 
exceptionally favourable to the formation of dolomite. It 
seems, then, from the above considerations that we cannot 
regard the corrosive effect of sea-water as an agent of more 
than very secondary importance in modifying the structure 
of an atoll. 

4. In regard to the negative geological evidence. Here 
we must not overlook two considerations — one that the 
structure of a coral reef is very commonly more or less 
composite ; broken coral, shells, &c., forming a part, and 
sometimes predominating when from one cause or another 
the growth of the polyps is temporarily checked (p. 155) ; 
hence in some cases, what is really a continuous reef may 
be supposed, if only an occasional section be visible, to be 
a series of thin reefs — the other (the more important and 
general) that the characteristic structure of dead coral 
becomes rapidly inconspicuous and may be only discover- 
able in thin sections under the microscope. Where dolo- 
mitisation has occurred it may be actually obliterated, for 
the niolecular changes involved in the process are often 
sufficient to destroy every trace of an organism. We may 
thus be prevented from recognising many ancient coral 
reefs. Moreover, the aporosa and madreporaria, which 
are now the chief reef- builders, have only become common 
since the conclusion of Palaeozoic ages, so that the largest 
volume of the geological history of the earth is excluded 
from consideration, because in the times which it covers 
the habits of the reef-builders may have been different. 
Reefs also, it must be remembered, are restricted at the 
present day to almost tropical regions, so that, notwithstand- 
ing any variation of climate, they must always have been 
less frequent and less luxuriant m northern latitudes — that 
is to say, in those regions with which geologists are best 
acquainted. StiU, instances of thick reefs of comparatively 


late date are on record,' and if those geologists are right who 
consider the Schlern dolomites as heing to a great extent 
due to reef-building corals, we have in the Triassic deposits 
of the Itahan Tyrol reefs thick enough to satisfy the most 
exacting requirements. 

It is then, I think, premature to regard the theory 
■which was advanced by Mr. Darwin, and has received the 
approval of an observer of such an exceptional experience 
as Professor Dana, as conclusively disproved by the results 
of the more recent investigations. That this theory may 
have been expressed in terms a little too comprehensive, 
that there may be a larger number of exceptional cases 
than was at first supposed, is quite possible. This, however, 
is the almost inevitable lot of every gi-eat generalisation. Its 
author concentrates, and rightly concentrates, his atten- 
tion on the salient features, as one v>^ho gazes first at a 
mountain group fixes his eyes upon the principal peaks 
and for a time pays little attention to, perhaps even 
under-estimates the importance of, the subordmate ranges ; 
nevertheless his conception of the physical structure of the 
region, though modified, is not overthrown by the work of 
subsequent travellers. This may prove to be the case in 
regard to the present controversy. It may very possibly 
be found that, as remarked by Mr. Bourne, the history cf 
coral reefs is more varied and complicated than was at first 
supposed, but it seems to me that, as the evidence at 
present stands, it is insufficient to justify a decision adverse 
tf> l\rr. Darwin's theory as a general explanation. 

I Scg pp, 309, 322 of this work. 

ABOLA ^^lW'"':0(k- 





Fig. 6. 
Out hull to 60 miles 

1:*° '° %( mUdou Matte 

L j WUa, doit Madou. 



k® f 

inckto 60 miles 


J l 






O - 


The Dames In italics are a'l names of places, and refer exclusively to the Aiipen'lii; la 
we;l-defiiied archipelagoes, or groups of i-^lauds, tlie name o/ each separate island u> 
not given. References in square brackets refer to the new appendiy. 

Abrolhos, Brazil, coated by corals, 

Abrolhos {Australia), 235 
Absence of coral-reefs from cer- 
tain coasts, 81 
Acaba, (julf of, 2G6 
ActcEon group, 200 
Admiralty group, 224 
Admiralty islands, [285] 
Africa, east coast, fringing-reef of, 
76. Madreporitic rocks of, 
Africa, east coast, 254 
Agassiz, Prof. A. 

on Tortuga and Florida 

reefs, [287] 
effect of Gulf stream, [287] 
growth of Florida reefs, [288] 
effect of currents on reefs, 

depth of lagoons, [290] 
formation of silt [291] 
Age of individual corals, 96 
Aiou, 231 
Aitutaki, 204 
Aklabra, 251 
Alert reef, 222 
Alexander, Grand Duke, island, 

Allan, Dr. 

on Holuthurias feeding on 
corals, 20 

Allan, Dr. 

on quick growth of corals at 

Madagascar, 104 
on reefs affected by currents, 
Alloufaco:i, 214 
Alpltoiise, 250 
Amargura, 214 
Amboina, 232 
America, west coast, 199 
Amirantes, 249 
Anachorites, 225 
Anambas, 240 

Anamouka, description of, 177 
Anamouka, 213 
Andaman islands, 239 
Antilles, 274 
Aj^jMO reef, 241 
Arabia Felix, 2G0 

great extent of, interspersed 

with low islands, 122 
of subsidence and of eleva- 
tion, 191 
of subsidence appear to ba 

elongated, 191 
of subsidence alternating 
with areas of elevation, 
Ai'nt group, 231 
Arzobispo, 230 
Ascension, no reef at, 83 



Ascidia, dci^th at which found, 

Assoniption, 250 
Astova, 250 

Atlantic islands, 83, 217 

breaches in their reefs, 39, 

dimensions of, 27 
dimensions of groups of, 123 
not based on craters, or on 
banks of sediment, or on 
rock, 119, 124, 125, 126, 194 
of irregular forms, 28, 146 
steepness of their Hanks, 31, 

164, 229 
width of their reef and islets, 

their lowness, 122 

lagoons, 35, [285], 

general range, 107 
with part of their reef sub- 
merged, and theory of, 37, 
140, 147 
whole reef submerged, and 
theory of, 38,146, 147, [285] 
Augiistin, St., 217 
Aurora island, an upraised atoll, 

Atirora, 201 
Austral islands, recently elevated, 

167, 177, 186 
Austral islands, 204 
Australia, N.W. coast, 234 
Australian barrier-reef, 63, 106 
Australian hamer, 222 

Balmy an group, 243 
Bahama banks, 208, 270 
Balnbac, 240 

Balfour, Prof. Bayley, descrip- 
tion of Eodriguez, [307] 
Bally, 236 
Bampton shoal, 222 
Banks' islatids, 220 
Banks in the West Indies, 267 
Barbes, St., 240 
Baring, 218 


of Australia, 63, 166 
of New Caledonia, 63, 67 

breaches through, 135 
not based on worn -down 
margin of rock, 06 

on banks of sediment, 

on submarine cratera, 
steepness of their flanks, 57, 

their probable vertical thick- 
ness, 04, 134 
theory of their formation, 
133, 137, [297] 
Bashee islands, 243 
Bass island, 206 
Bitoa, 215 
Beaupri reef, 221 
Beechey, Capt., 

obligations of the author to, 

on submerged reefs, 37 
account of Matilda island, 
Belcher, Sir E. 

on boring through coral- 
reefs, 99 
on changes in Chain atoll, 

on Clipperton rock, 199, 
Bellinghausen, 203 
Bengal, gulf of, elevation of east- 
ern shores of, 181 
Bermuda islaiuls, 275 
Beveridge reef, 212 
Bligh, 220 
Bolabola, view of, 3 
Bombay shoal, 245 
Bo7iin bay, 238 
Bonin group, 230 
Bonney, Prof. T. G., 

discussion of arguments 
against Mr. Darwin's 
theory, [325] 
movements of upheaval and 
subsidence, [o2()j 



Bonney, Prof. T. ii., 

on lateral spreading of reef, 

on the solution theory, [329] 
on the geological evidence, 
Borings through coral reefs, 99, 

Borneo, W. coast, recently ele- 
vated, 180 

E. coast, 237 
S.W. and W. coast, 2i0 
N. coast, 240 
western bank, 2-10 
Boscawen, 214 
Boston, 218 
Bouka, 223 
Bourbon, 249 
Bourne, Mr. G. C, 

Chagos group, [300] 

on solution of dead coral, 

on lateral spreading of reefs, 

coral gro\vth in lagoons, [302] 
imi^ortauceof currents, [304] 
conclusions, [305] 
Bourou, 232 
Bouton, 238 
Brazil, friuging-reefs on coast of, 

Breaches through barrier-reefs, 

Brook, 207 
Bunker, 207 
Biinoa, 240 
Bijron, 217 

Cagayanes, 241 
Candelaria, 224 
Carbonate of lime in ocean 

water, [284], [328] 
Cargados Carajos, 248 
Caroline archipelago, 225 
Caroline island, 206 
Carteret shoal, 231 
Caryophyllia, depths at which it 

lives, 117 

Cavilli, 241 

Cayman island, 273 

Celebes, 233 

Ceram, 232 

Ceylon, recently elevated, 171 

Ceylon, 247 

Chagos Great bank, [300] 

Chagos, Great bank, description 

and theory of, 53, 149, 

Chagos group, 149 
Chagos group, 247, [300] 
Chama shells embedded in coral 

rock, 100 
Chamisso on corals preferring the 

surf, 85 

in the state of Keeling atoll, 

of atolls, 130, 170 

leading into the lagoons, 59 

of atolls, 39, 148, [312] 

the Maldiva atolls, 

through barrier-reefs, 135 
Chase, 217 
China sea, 244 
Christmas atoll, 100 
Christmas aioll, 208 
Christmas island {Indian Ocean), 

Clarence, 207 
Clarke, W. B., on recent elevation 

of the Loyalty islands, 178 
Clipperton rock, 109 
Cochin China, 246 
Cocos, or Keeling atoll, 7, [305] 
Cocos (or Keeling), 240 
Cocos island [Pacific), 199, 214 
Coetivi, 250 
Comoro group, 251 
Composition of coral formations, 


coral rock on Keeling atoll, 

coral rock on other atolls, 

coral rock, 156 



Cook islands, recently elevated, 

Cook islands, 204 
Corallian sea, lOfi 
CoraUian sea, '2"22 

their distribution and ab- 
sence from certain areas, 
destroyed by loose sediment, 

at Keeling atoll, 16 
JIanritius, 74 
JMetia, 98 
organic remains of, 15G 
at Oahu, [822] 
Coral-rocks bored by vermiform 

animals, 20, 156 

dead but upright in Keeling 

lagoon, 21 
depths at which they live, 

108, [293], [298], [326] 
off Keeling atoll, 11 
killed by a short exposure, 

living in the lagoon of Keel- 
ing atoll, 18 
quick growth of, in Keeling 

lagoon, 17 
merely coating the bottom 

of the sea, 79 
standing exposed iu the Low 
archii3elago, 170 
Cornwallis, 218 
Cosmoledo, 250 
Couthouy, Mr. 

alleged proofs of recent eleva- 
tion of the Low archipe- 
lago, 170 
on external ledges round 
coral islands, 140 
Crescent-formed reefs, 146 
Crosby, Mr. W.O., on the raised 
reefs of Cuba, [308] 
proofs of ancient subsi- 
dence, [309] 
Cuba, 270, [308] 
Cuming, Mr., on the recent ele- 

vation of the Philippines, 

Dana, Prof., on the reef of 

Hawaii, 82, 175 
distribution of coral-reefs as 

affected by the temjiera- 

ture of the sea, 85 
upraised coral-rock of Metia, 

boring through coral-rock, 

99, [322] 
depth at which corals live, 

subsidence of the Mendana 

island, 105, 204 
subsidence in the Caroline 

archipelago, 109 
slight recent subsidence of 

the Paumotu archipelago, 

extension of the Hawaii ar- 
chipelago, 192, 211 
Feejee islands, 215 
outline of some islands in- 
dicative of subsidence, 

distribu tion of reef -materiala, 

different kinds of reef in 

Pacific, [312] 
submarine slopes, [313] 
local elevations, [314] 
on soundings at Tahiti, 

on lateral spreading of reefs, 

on Florida reefs, [316] 
on Mr. Murray's explanation 

of the foundation of reef3 

and of ring-shaped atolls, 

Danqer islaiids, 207 
Dangerous or Low archipelago^ 


at which reef-building corala 

Uve, 108, [293], [29Sj, 




Depths at Mauritius, the Red 

Sea, and in the Maldiva 

archipelago, 113 
at which other corals and 

corallines can live, IIU 
Dhalac group, 2G1 
Diego G-arcia, slow growth of 

reefs, 92 
Dimensions of the larger groups 

of atolls, 12;j 
Discovery shoal, 2J:5 
Disseverment of the Maldiva 

atolls, and theory of, 50, 

Distribution of coral-reefs, 80 
Dolomitisation of coral, 18, [330] 
Domingo, St., 273 
Dory Port, recently <3levated, 179 
Dori/ Port, 231 
Duchassaing on rapid growth of 

corals, 107 
Duff's islands, 221 
Durour, 225 

Eap, 227 

at Keeling atoll, 25 

in groups of atolls, 131 

in Navigator archipelago,178 
East Indian archiiDclago recently 

elevated, 179 
Easter, 200 
Echequicr, 225 

on the banks of the Red 
Sea, 78, 260 

on depth at which corals 
live in the Red Sea, 113 

on corals preferring the surf, 

on the antiquity of certain 
corals, 96 
Einieo, 201 
Elevated Reef of Mauritius, 74 

of Rodriguez, [307] 

of Cuba, [308] 
Elevations, recent, 

proofs of, 175, [291] , [311] 

immense areas of, 190 

EUvi, 227 
Elizabeth island, 98 

recently elevated, 176, 186 
Elizabeth island, 200 
Ellice group, 216 
Encircled islands, 
their height, 62 
their geological composition, 

62, 68 
Eoua, description of, 177 
Eoua, 213 
Erupted matter, probably not 

associated with thick 

masses of coral rock, 


Fais recently elevated, 179, 191 

Pais, 227 

Fanning, 209 

Farallon da Medinilla, 229 

Farsan Group, 262 

Fataka, 220 

Fii-lji archipelago, 214, [310] 


feeding on corals', 20 

killed in Keeling lagoon by 

heavy rain, 27 
Fissures across coral islands 

132, 262 
FitzRoy, Capt., 

on a submerged shed at 

Keeling atoll, 25 
on an inundation in the Low 

archipelago, 130 
Flint, 206 
Flores, 236, 259 
Florida, 270, 275, [287] 
Folger, 230 
Formosa, 243 
Forster, theory of coral forma- 

tions, 127 
Frederic reef, 222 
Freeivill, 231 
Friendly group recently elevated, 

i77, 186 
Friendly archipelago, 212 

absent where coast preci- 
pitous, 69 




breached in front of streams, 

described by MM. Qnoy and 
Gairaard, 175 

not closely attached to shel- 
ving coasts, 72 

of east coast of Africa, 76 

of Cuba, 75 

of Mauritius, G9, 71 

on woru-down banks of rock, 

on banks of sediment, 78 

their appearance when ele- 
vated, 74 

their growth influenced by 
currents, 79, [b04] 

by shallowness of sea, 

Galapagos archipelarjo, 109 

Galcga, 250 

Gambler islands, section of, v'53 

Gambler islands, 201 ' 

Gardner, 208 

Gaspar Rico, 218 

Geological composition of coral 
formations, 156, [;!-31] 

Gilbert archipelago, 217 

Gilolo, 233 

Glorioso, 250 

Gloucester island, 130 

Glover reef, 272 

Gomez, 200 

Gouap, 227 

Goulou, 227 

Grampus, 230 

Grand Cocal, 216 

Graves, on tlie recent elevation 
in the Bonin archipelago, 

Great Chagos bank, description 
and theory of, 53,150, [300] 

Grey, Capt., on sand-bars, 72 

Giiedes, 231 

Gu^jpy, Mr., on Solomon archi- 
pelago, [291] 
proofs of upheava.1, [291] 
summary of opinion, [292] 

Gupjiy, Mr., 

growth of corals, [292] 

development of reefs, '293] 

different kinds of reefs, 

description of variouB 
islands, [294] 

barrier reefs and their for- 
mation, [297] 

depth at which reefs begin, 

lateral spreading of reefs, 

removal of dead coral, [299] 

nature of sea-bed near reefs, 

on Keeling atoll, [306] 

Hales, Mr., on subsidence in the 

Caroline archipelago, KiO 
fl.all, Capt. B., on Loo Choo, 181 
Halstead, Capt., elevation of 

eastern shore of Gulf of 

Bengal, 181 
Harvey islands recently elevated, 

185, 18G 
Harvey or Cook islands, 204 
Height of encircled islands, G2 
Hcrmites, 225 
Hogoleu, 226 
Hoiuthurise feeding on corals, 

Honduras, reef off, 272 
Honoluhi, boring at, [323] 
Horn, 214 

Houlman's Abrollios, 235 
Hnahcine, 203 
Hull island, 208 
Humplirey, 207 
Hunter, 214 
Hurricanes, effects of on coial 

islands, 129 

Immauvi, 260 
Independence, 216 
India, east coast recently ele- 
vated, 181 
India, 260 



Irregular reefs in shallow seas, 

Islets of coral-rock, their forma- 
tion, 15 
their destruction in the Mal- 
diva atolls, 50 

Jamaica, 273 
Jarvis, 207 

Java recently elevated, 180 
Java, 23G 

Johnston island, 209 
Jua7i da Nova, 2o0 
Juan da Nova (Madagascar), 253 
Jukes, P of., the barrier reef of 
Australia, 64 

Kalatoa, 237 

Kamtschatka, proofs of its recent 

elevation, 189 
Karkalang, 233 
Keeling atoll, section of reef, 7, 8, 


iwrth atoll, 246 
south atoll, 216 
Keffing, 232 
Kemin, 206, 208 
Kennedy, 221 
Kcppel island, 214 
Keppell, Capt., on the reefs and 

elevation of Celebes, 233 
King, Capt., on distribution of the 

different classes of reefs, 

Kumi, 214 

Laccadive groitp, 247 

Ladrones or Marianas, recently 

elevated, 172 
Ladrones archipelago, 229 
Lagoon of Keeling atoll, 18 

bordered by inclined ledges 

and walls, and theory of 

their formation, 41, 139, 

[285], [301] 
of small atolls filled up with 

Ecdiment, 42 


Lagoon-channels within barrier 

reefs, 59 
Lagoon-reefs, all submerged in 

some, atolls and all rising 

to the surface in others, 90 
Lancaster reef, 206 
Larrack, 260 
Latte, 213 

Laztghlan islands, 222 
Ledges round certain lagoons, 41, 

Lctte, 234 

Lighthouse reef, 272 
Lloyd, Mr., on corals reflxing 

themselves, 105 
Loo Choo recently elevated, ISl 
Loo Choo, 244 
Loioisiade, 222 
Low archipelago, alleged proofs 

of its recent elevation, 170 
Loio archipelago, 200 
Lowness of coral islands, 122 
Loyalty group, 221 
Loyalty islands, 

recently elevated, 186, 221 
Lucepara, 240 
Lutk6, Adm., on fissures across 

coral islands, 132 
TjQzon recently elevated, 180 
Luzon, 241 
Lyell, Sir C, 

on channels into the lagoons 

of atolls, 39 
on the lowness of their lee- 
ward sides, 148 
on the antiquity of certain 

corals, 96 
on the a^jparent continuity 

of distinct coral-islands, 

on the recently elevated beds 

of the iled Sea, 184 

MacAslcill Islands, formed of up- 
raised coral rock, 191,227 

Ufacassar strait, 237 

Macclesfield bank, 245 

Maclear, Capt., on Masaraarhu 
Island, [319] 




quick growth of corals at, 

madreporitic rock of, 181 
Madagascar, 252 
Madjiko-sima, 243 
Madura (India), 247, 2C0 
Madura (Java), 23() 
Mahlos Mahdoo, theory of for- 
mation, 144 
Maitea, 202 

Malacca recently elevated, 180 
Malacca, 239 
Malcolmson, Dr., 

on recent elevation of E. 
coast of India, 181 

on recent elevation of Cama- 
ran island, 183 
Maiden, 207 
Maldiva atolls, 

and theory of their forma- 
tion, 44, 141, 142 

steepness of their flanks, 30 

growth of coral at, 103 
Maldiva arcliipclago, 247 
Mangaia island, 

recently elevated, 17G, 186 
Mangaia, 205 
Mangs, 230 
Manotiai island, 204 
Marianas, recently elevated, 179 
Mariana archipelago, 229 
Maricre, 228 

Marquesas, subsidence of, 1G5 
Marquesas, 203 
Marsliall archipelago, 218 
Marsliall island, 230 
Martinique, '.-74 
Martires, 228 

Mary's, St., in Madagascar, har- 
bour made in reefs, 88 
Mary island, 208 
MasdmarhiL island, sections of, 

Matilda atoll, 101 
Mauki island, 204 

fringing-reefs of, G9, 71 

depths at which corals tliero 
live. 110 



recently elevated, 181 
Mauritius, 248 
Maurua, section of, C5 
Maurtia, 202 
Menchicoff atoll, 28, 146 
Mcndana isles, 203, 221 
Mendana island, subsidence of, 

Metia, 98, 123, 176, 186 
Tilexico, gulf of, 2G9 
Millepora complanata at Keeling 

atoll, 10 
Mindoro, 241 
^fohilla, 252 
Mopeha, 203 
Moresby, Capt., on boring through 

coral reefs, 99 
Marty, 233 
Mosquito coast, 273 
Murray, Mr., on the structure 

and origin of coral reefs 

and islands, [283] 
quantity of carbonate of 

lime present in ocean 

water, [284] 
lateral spreading of coral 

reefs, [2-t.5] 
solvent action of sea-water, 

summary of conclusions, 

Musquillo atoll, 146 
Mysol, 232 

Namourrek group, 145 

Natunas, 240 

Navigator archipelago, elevation 

of, 178 
Navigator archipelago, 211 
Nederlandisch islands, 216 
Nelson, Lieut., 

on the consolidation of ooral- 

rock, under water, 98 
theory of coral formations, 

on the Bermuda islands, 
New Britain, 224 



New Caledonia, 

steepness of its reef, 57 
banier-reef of, 63, 07, 139, 
145, 1(36 

New Caledonia, 221 

Neio Guinea (E. end), 224 

Netv Guinea (W. end), 231 

New Hanover, 224 

New Hebrides recently elevated, 

New Hebrides, 218 

New Ireland recently elevated, 

New Ireland, 224 

Netv Nantucket, 208 

Nicobar Islands, 239 

Niouha, 214 


at Keeling atoll, 13 
on the reefs of atolls, 34 
on barrier reefs, 57 
their wide distribution and 
abundance, 117 

Oahu, borings at, [322] 
Objectionsto the theory of subsi- 
dence, 153 
Ocean islands, 211, 217 
Ono, 215 
Onouafu, 214 
Orrnus, 260 
Oscar group, 216 
Oscillations of level, 166, 184, 193 
Oualan or Ualan, 225 
Ouluthy atoll, 101 
Outong Java, 224 


S.W. coast, 240 
NW. coast, 241 
ivcstern bank, 245 

Pahnerston, 204 

Palmyra, 209 

Paracells, 245 

Paraquas, 246 

Patchow, 243 

Paumotu archipelago, 170 

Paumotu archiijelacjo, 200 


Peel island, 230 

Pelew islands, 227 

Pemba island, singular form of, 

Pemba, 255 
Penrhyn, 206 
Peregrino, 206 
Pernambuco, bar of sandstone at, 

73, 277 
Persian gulf, recently elevated, 

Persian gulf, 259 
Pescado, 207 
Pescadores, 244 
Peyster group, 216 
Philip, 227 
Philippine archipelago, recently 

elevated, 180 
Pliilinpine archipelago, 241 
Phanix, 208 
Pierre, Si., 250 
Piguiram, 227 
Pitcairn, 201 
Pii island, 217 
Pitt's bank, 152 
PZaWe, 250 
Pleasant, 217 
Porites, chief coral on margin of 

Keeling atoll, 9 
Postillions, 237 
Pouynip^te, 168 

its probable subsidence, 169 
Pouynipite, 226 
Pratas shoal, 244 
Proby, 214 
Providence, 250 
P«t'rto iJico, 273 
PztZo Anno, 228 
Pulo Leat, 240 
Pumice floated to coral islands, 

Pylstaart, 212 
Pyrard de Laval, astonishment 

at the atolls in the Indian 

Ocean, 2 

Quoy and Gaimard, 

depths at which corals live, 




Quoy and Gainiard, 

duscription of reefs appli- 
cable only to fringing- 
rocfs, 174 

Raivaivai, 206 
Eange of atolls, 167 
Rapa, 20() 
Bear son, 207 
Ked Sea, 

banks of rock coated by reefs, 

proofs of its recent elevation, 

supposed Eubsidenca of, 
184, [319] 
Red Sea, 260 

irregular in shallow seas, 

rising to the surface in some 
lagoons and all submerged 
in others, 91 
their distribution, 80 
their absence from some 

coasts, 81 
lateral spreading of , 22, [285], 

[298], [301], [315], [327] 
formation of, '291] 
Rcvilla-gujcdo, 199 
Eing-formed reefs of the Maldiva 
atolls, and theory of, 45, 
Rodriguez, 248, [307] 
Rosario, 230 
Rose island, 212 
Botches, 217 
Roug, 226 
Routoumah, 216 
Rowley slioals, 235 
liiippell. Dr., on the recent ds- 
posits of Ked Sea, 183 

Sable, lie de, 248 
Sahia de Malha, 248 
Salomon archipelago, 223 
Bamoan, or Navigator archipe- 
lago, elevation of, 178 

Samoa archij.clago, 211 
Sandahvood, 234 
Sandbars, parallel to coasts, 73 
Sandwich archipelago recently 
elevated, 175 
extension of, 191,211 
Sandwich archipelago, 209, [322] 
Sanserot, 228 
Santa Cruz, 220, 274 
Savage island recently elevated, 

Savage, 212 
Savu, 234 

Sajja or Sahia de Malha, 248 
Scarborougli slwal, 215 
Scarus feeding on corals, 20 
Schouton, 224 
Scilly, 208 
Scoriie floated to coral islands, 

Scott's reef, 235 

of islands encircled by bar- 
rier reefs, (50, 133 
of Bolabola, 134 

in Keeling lagoon, 19 
in other atolls, 36, 48 
injurious to corals, 87 
transported from coral- 
islands far scaMard, 157 
formation of, [291] 
Semper, Prof., 

on the Pelew islands, 228 
on the reef of the Philippine 
archipelago, 242 
Seniavhie, 227 
Serangani, 233 
Seychelles, 249 
Ship-bottom quicldy coated with 

coral, 100 
Smyth island, 209 
Society archipelago, 165 

stationary condition of, 169 
alleged proofs of recent ele 
vation, 185 
Society archipelago, 201 
Socotra, 260 

Solomon archipelago, [291] 
Solor, 236 



Solution of dead coral, [286], 

[299], [301], [329] 
Sooloo islands, recently elevated, 

Sooloo islands, 2-10 
Souvoroff, 207 

on growth of coral, 108 
Spanish, 228 
Starhuck, 20G 
Stones transported in roots of 

trees, 157 
Storms, effects of on coral-islands, 

Stutehbury, Mr., 

on the growth of an Agari- 

cia, lUG 
on upraised corals in Society 
archipelago, 185 

of Keeling atoll, 23-25 
extreme slowness of, 193 
areas of, apparently elon- 
gated, 191 
areas of, immense, 190 
great amount of, 193 
indicated by shape of coast, 

at Masamarhu island, [319] 
at Oahu, [322] 
Snez, gulf of, 2G6 
Sulphur islands, 230 
Sumatra, recently elevated, 180 
Sumatra, 238 
Sumbaioa, 23(3 
Surf favourable to the growth cS 

massive corals, 85 
Sioallow shoal, 245 
Sydney island, 208 

Tahiti, alleged proofs of its re- 
cent elevation, 185, [314J 

Tahiti, 201 

Tanasserim, 239 

Tapamanoa, 202 

Temperature of the sea at the 
Galapagos archipelago, 82 

Tenhnber island, 231 

Teturoa, 202 

Theories on coral formations, 
119, 127, [28GJ, [290], 
[298], [305] 

Theory of subsidence, and objec- 
tions to, 12G, 153, [283], 

Thickness, vertical, of barrier- 
reefs, GG, 135 

Thomas, St., 274 

Tikopia, 220 

Timor recently elevated, ISO 

Timor, 234 

Timor-laut, 231 

Tokan-Bessces, 237 

Tongatabou, description of, 177 

Tongatabou, 213 

Tonquin, 24G 

Toubai, 203 

Toufoa, 213 

Toupoua, 220 

Traditions of change in coral- 
islands, 129 


embedded in coral-rock, 156 
left exposed in the Low 
archipelago, 170 

Tnbularia, quick growth of, 106 

Tumbclan, 240 

Turneffe reef, 272 

Turtle, 215 

Ualan, 225 


section of, G5 

its state and changes in its 
reefs, 1G9 
Vanikoro, 220 
Vavao, 213 
Vine reef, 222 
Virgin Gorda, 274 
Viti archipelago, 214 

islands, with living corals on 

their shores, 81 
matter, rarely associated 
with thick masses of coral- 
rock, 157 





authorities for their position 
on the map, l(jO 

their presence determined 
by the movements in pro- 
gress, 189 

absent or extinct in the areas 
of subsidence, ISG 

Waigiou, 231 

Wallis island, 214 

Washington, 20'J 

Wells' reef, 222 

Wellstead, Lieut., account of a 

ship coated with corals, 

West Indies, 

banks of sediment, fringed 

by reefs, 78 
recently elevated, 183 
West Indies, 26(3 
Wharton, Capt., on Masdraarlm 

island, [329] 


Whitsunday island, 
view of, 2 

changes in its state, 130 
Williams, Ecv. J., 

on traditions of the nativea 
regarding coral-islands, 
on antiquity of certain corals, 
WolchonsJcy, 200 
Wostock, 2UG 

Xulla islands, 232 

York island, 208 
Yucutan, coast of, 272 

Zones of different kinds of corals, 
outside the same reefs, 
7i, 90, 100 

University of California 


405 Hilgard Avenue, Los Angeles, CA 90024-1388 

Return this material to the library 

from which it was borrowed. 

A 000 022 629 o