that Darwin set foot on was the Cocos (Keeling) Islands in
the eastern Indian Ocean. Here he keenly accepted observations of the undercutting of coconuts and the erosion of the
shoreline as “tolerably conclusive evidence” in support of
his theory.
It is important to discriminate that Darwin’s theory
applies to the structure of reefs, based on their long-term
evolution, at timescales of millions of years, whereas the
surface morphology of the atolls reflects late Holocene
formative processes that operate over much shorter timescales (Stoddart, 1973b). Stoddart (Stoddart, David Ross
(1937–)) has emphasized the difference in time scale; reef
structure (Reef Structure) being the outcome of millions of
years of geological evolution, whereas the surface morphology results from the most recent (Holocene) adjustments of
form to the processes that operate, including subtle changes
in sea level. At the time Darwin proposed his theory, the significance of sea-level fluctuations associated with the glaciations was unknown, but it is possible to incorporate our
latest understanding of the oscillations of sea level into the
gradual formation of the sequence of limestones that underlie a typical atoll.
Darwin’s subsidence theory of atoll evolution was tested
by deep drilling on Funafuti Atoll in 1896–1898 which,
although it failed to reach the underlying volcanic basement, recovered more than 300 m of shallow-water carbonates implying subsidence (Spencer et al., 2008). Subsidence
of volcanic basements upon which atolls are founded was
eventually substantiated by drilling on the atolls of Bikini
and Eniwetak in the Marshall Islands (see a synthesis
in Guilcher, 1988, and Bikini Atoll, Marshall Islands;
Enewetak Atoll, Marshall Islands). Daly recognized the significance of sea-level fluctuations (Daly, 1934) and
documented evidence from across the Pacific that recorded
a sea level above present level. However, the Glacial
Control Hypothesis that Daly advocated to explain reef
development, further developed by Wiens (1959, 1962),
presumed that reefs were totally planed off at low sea level
and that the entire structure of modern reefs was Holocene.
The antecedent karst hypothesis advocated by Purdy (1974)
corrected this mistaken view, and recognized the significance of Antecedent Platforms, often of late Pleistocene
age. It has now been widely shown that the reef rim on modern atolls is underlain by older Pleistocene reefs (McLean
and Woodroffe, 1994; Montaggioni, 2005).
Darwin’s theory was enthusiastically adopted by several other prominent scientists, most notably Dana (Dana,
James Dwight (1813–1895)), and became widely debated.
An alternative view proposed by Murray (1889) revolved
around a belief that atolls developed as a result of solution
of lagoons. Although considered by Gardiner (1931), this
view became discredited when it was realized that seawater
was supersaturated with calcium carbonate. Wood-Jones
(1912) proposed an alternative view based on his time in
the Cocos (Keeling) Islands. He thought that it was the production of sediment around the margin of an atoll and its
transport and deposition in the interior that prevented coral
growth in the center of reef platforms. However, this
sedimentation theory was not widely supported. W.M.
Davis (Davis, William, Rorris (1850–1934)), in his review
of the origin of reefs (Davis, 1928), considered that the only
real contender against Darwin’s view was the glacial control theory proposed by Daly, and in a subsequent review,
Cotton (1948) regarded the subsidence, glacial control
and antecedent topography theories as plausible.
In addition to his observations on Cocos, Darwin undertook a substantial compilation of information on reefs, and
in his book published in 1842, he included a map of the distribution of atolls, which provided further evidence of the
fact that most atolls occur in mid ocean (Darwin, 1842).
Further exploration was to extend knowledge about
atolls. The distinguished American geologist, James
Dana, extended Darwin’s ideas as a result of his visits to
Kiribati, Tuvalu, the Tuamotu Archipelago, the Society
Islands, Fiji, and the Phoenix and Hawaiian Islands. He
was a firm supporter of the subsidence hypothesis (Subsidence Hypothesis of Reef Development). Alexander Agassiz (Agassiz, Alexander (1835–1910)) undertook
extensive studies of reefs on extended voyages at the turn
of the twentieth century. His 9-month cruise on Albatross
in 1899–1900 enabled him to describe 30 atolls in the
Tuamotu Archipelago and 28 in Tuvalu, and the Marshall
and Caroline Islands. He followed this with further
descriptions of the Maldives Archipelago in 1901–
1902, where his observations built on those of Gardiner
who had mounted an expedition in 1899–1900. Agassiz
attempted to drill Wailangilala atoll in Fiji, but recovered
material only from the upper 26 m. The Chagos Archipelago was described during the Percy Sladen expedition,
and a fuller account of this period of reef exploration is
given by Spencer et al. (2008). The Cocos (Keeling)
Islands were examined by Guppy (1889); Wood-Jones
(1912); and Gibson-Hill (1947), making this one of the
best known atolls by mid-twentieth century.
Darwin had realized that drilling through an atoll was
the optimal way to test his theory and he wrote before
his death to Agassiz in an effort to encourage such drilling
(see Darwin, Charles (1809–1882)). Such drilling was
finally undertaken on Funafuti Atoll. The Royal Society
of London sponsored a program involving a series of expeditions in the 1890s to Funafuti. The objective was to drill
the perimeter of Funafuti to test Darwin’s theory of reef
development. The initial fieldwork was led by Professor
W. Sollas in 1896; further drilling was undertaken in
1897, together with field mapping by T. Edgeworth David
(David, Tannant Edgeworth (1858–1934)) and George
Sweet, and the final stage of drilling, although still in shallow-water carbonates, was overseen by Alfred Finckh in
1898. At the time the fact that the core did not reach volcanic basement at more than 300 m depth appeared inconclusive, although it was clear that shallow-water
carbonates persisted below depths at which they are now
forming. In retrospect we now know from the strontium
isotope stratigraphy that dolomite in the lower core was
formed through diagenesis between 1 and 2 million years
ago. The upper 26.4 m of the core has been radiocarbon
58
ATOLLS
the eastern Indian Ocean. Here he keenly accepted observations of the undercutting of coconuts and the erosion of the
shoreline as “tolerably conclusive evidence” in support of
his theory.
It is important to discriminate that Darwin’s theory
applies to the structure of reefs, based on their long-term
evolution, at timescales of millions of years, whereas the
surface morphology of the atolls reflects late Holocene
formative processes that operate over much shorter timescales (Stoddart, 1973b). Stoddart (Stoddart, David Ross
(1937–)) has emphasized the difference in time scale; reef
structure (Reef Structure) being the outcome of millions of
years of geological evolution, whereas the surface morphology results from the most recent (Holocene) adjustments of
form to the processes that operate, including subtle changes
in sea level. At the time Darwin proposed his theory, the significance of sea-level fluctuations associated with the glaciations was unknown, but it is possible to incorporate our
latest understanding of the oscillations of sea level into the
gradual formation of the sequence of limestones that underlie a typical atoll.
Darwin’s subsidence theory of atoll evolution was tested
by deep drilling on Funafuti Atoll in 1896–1898 which,
although it failed to reach the underlying volcanic basement, recovered more than 300 m of shallow-water carbonates implying subsidence (Spencer et al., 2008). Subsidence
of volcanic basements upon which atolls are founded was
eventually substantiated by drilling on the atolls of Bikini
and Eniwetak in the Marshall Islands (see a synthesis
in Guilcher, 1988, and Bikini Atoll, Marshall Islands;
Enewetak Atoll, Marshall Islands). Daly recognized the significance of sea-level fluctuations (Daly, 1934) and
documented evidence from across the Pacific that recorded
a sea level above present level. However, the Glacial
Control Hypothesis that Daly advocated to explain reef
development, further developed by Wiens (1959, 1962),
presumed that reefs were totally planed off at low sea level
and that the entire structure of modern reefs was Holocene.
The antecedent karst hypothesis advocated by Purdy (1974)
corrected this mistaken view, and recognized the significance of Antecedent Platforms, often of late Pleistocene
age. It has now been widely shown that the reef rim on modern atolls is underlain by older Pleistocene reefs (McLean
and Woodroffe, 1994; Montaggioni, 2005).
Darwin’s theory was enthusiastically adopted by several other prominent scientists, most notably Dana (Dana,
James Dwight (1813–1895)), and became widely debated.
An alternative view proposed by Murray (1889) revolved
around a belief that atolls developed as a result of solution
of lagoons. Although considered by Gardiner (1931), this
view became discredited when it was realized that seawater
was supersaturated with calcium carbonate. Wood-Jones
(1912) proposed an alternative view based on his time in
the Cocos (Keeling) Islands. He thought that it was the production of sediment around the margin of an atoll and its
transport and deposition in the interior that prevented coral
growth in the center of reef platforms. However, this
sedimentation theory was not widely supported. W.M.
Davis (Davis, William, Rorris (1850–1934)), in his review
of the origin of reefs (Davis, 1928), considered that the only
real contender against Darwin’s view was the glacial control theory proposed by Daly, and in a subsequent review,
Cotton (1948) regarded the subsidence, glacial control
and antecedent topography theories as plausible.
In addition to his observations on Cocos, Darwin undertook a substantial compilation of information on reefs, and
in his book published in 1842, he included a map of the distribution of atolls, which provided further evidence of the
fact that most atolls occur in mid ocean (Darwin, 1842).
Further exploration was to extend knowledge about
atolls. The distinguished American geologist, James
Dana, extended Darwin’s ideas as a result of his visits to
Kiribati, Tuvalu, the Tuamotu Archipelago, the Society
Islands, Fiji, and the Phoenix and Hawaiian Islands. He
was a firm supporter of the subsidence hypothesis (Subsidence Hypothesis of Reef Development). Alexander Agassiz (Agassiz, Alexander (1835–1910)) undertook
extensive studies of reefs on extended voyages at the turn
of the twentieth century. His 9-month cruise on Albatross
in 1899–1900 enabled him to describe 30 atolls in the
Tuamotu Archipelago and 28 in Tuvalu, and the Marshall
and Caroline Islands. He followed this with further
descriptions of the Maldives Archipelago in 1901–
1902, where his observations built on those of Gardiner
who had mounted an expedition in 1899–1900. Agassiz
attempted to drill Wailangilala atoll in Fiji, but recovered
material only from the upper 26 m. The Chagos Archipelago was described during the Percy Sladen expedition,
and a fuller account of this period of reef exploration is
given by Spencer et al. (2008). The Cocos (Keeling)
Islands were examined by Guppy (1889); Wood-Jones
(1912); and Gibson-Hill (1947), making this one of the
best known atolls by mid-twentieth century.
Darwin had realized that drilling through an atoll was
the optimal way to test his theory and he wrote before
his death to Agassiz in an effort to encourage such drilling
(see Darwin, Charles (1809–1882)). Such drilling was
finally undertaken on Funafuti Atoll. The Royal Society
of London sponsored a program involving a series of expeditions in the 1890s to Funafuti. The objective was to drill
the perimeter of Funafuti to test Darwin’s theory of reef
development. The initial fieldwork was led by Professor
W. Sollas in 1896; further drilling was undertaken in
1897, together with field mapping by T. Edgeworth David
(David, Tannant Edgeworth (1858–1934)) and George
Sweet, and the final stage of drilling, although still in shallow-water carbonates, was overseen by Alfred Finckh in
1898. At the time the fact that the core did not reach volcanic basement at more than 300 m depth appeared inconclusive, although it was clear that shallow-water
carbonates persisted below depths at which they are now
forming. In retrospect we now know from the strontium
isotope stratigraphy that dolomite in the lower core was
formed through diagenesis between 1 and 2 million years
ago. The upper 26.4 m of the core has been radiocarbon
58
ATOLLS