Destruction of Atlantis by a great earthquake and tsunami? A geological analysis of the Spartel Bank hypothesis

Abstract

Numerous geographical similarities exist between Plato’s descriptions of Atlantis and a
paleoisland (Spartel) in the western Straits of Gibraltar. The dialogues recount a cata- strophic event that submerged the island ca. 11.6 ka in a single day and night, due to violent earthquakes and floods. This sudden destruction is consistent with a great earth- quake (M 􏰂 8.5) and tsunami, as in the Gulf of Cadiz region in 1755 when tsunami run- up heights reached 10 m. Great earthquakes (M 8–9) and tsunamis occur in the Gulf of Cadiz with a repeat time of 1.5–2 k.y., according to the sedimentary record. An unusually thick turbidite dated as ca. 12 ka may coincide with the destructive event in Plato’s ac- count. The detailed morphology of Spartel paleoisland, as determined from recently ac- quired high-resolution bathymetric data, is reported here. The viability of human habi- tation on this paleoisland ca. 11.6 ka is discussed on the basis of a new bathymetric map.

Full text

q2005 Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or editing@geosociety.org.
Geology; August 2005; v. 33; no. 8; p. 685–688; doi: 10.1130/G21597.1; 5 figures. 685
Destruction of Atlantis by a great earthquake and tsunami? A
geological analysis of the Spartel Bank hypothesis
Marc-Andre´ Gutscher Centre National de la Recherche Scientifique, Institut Universitaire Europe´en de la Mer, Universite´de
Bretagne Occidentale, F-29280 Plouzane´ , France
ABSTRACT
Numerous geographical similarities exist between Plato’s descriptions of Atlantis and a
paleoisland (Spartel) in the western Straits of Gibraltar. The dialogues recount a cata-
strophic event that submerged the island ca. 11.6 ka in a single day and night, due to
violent earthquakes and floods. This sudden destruction is consistent with a great earth-
quake (M .8.5) and tsunami, as in the Gulf of Cadiz region in 1755 when tsunami run-
up heights reached 10 m. Great earthquakes (M 8–9) and tsunamis occur in the Gulf of
Cadiz with a repeat time of 1.5–2 k.y., according to the sedimentary record. Anunusually
thick turbidite dated as ca. 12 ka may coincide with the destructive event in Plato’s ac-
count. The detailed morphology of Spartel paleoisland, as determined from recently ac-
quired high-resolution bathymetric data, is reported here. The viability of human habi-
tation on this paleoisland ca. 11.6 ka is discussed on the basis of a new bathymetric map.
Keywords: earthquake, tsunami, Iberia, paleoseismology, geoarcheology.
Figure 1. A: Location map of south Iberian–Moroccan region with relief shaded (>200 m
light gray, >1000 m medium gray, >2000 m dark gray). Dimensions of coastal plain sur-
rounding Gulf of Cadiz are 450 3300 km, consistent with Plato’s description in
The
Critias
(3000 32000 stadia). Subduction fault plane (light shading) has been proposed
to be source of 1755 earthquake (Gutscher, 2004). B: Eustatic sea-level curve since 30
ka (after Labeyrie et al., 1987; Bard et al., 1996). C: Bathymetric map of western Straits
of Gibraltar showing paleoshoreline at 14.5 ka (2100 m contour). Modern shoals rising
to <100 m depth (shaded black) represent paleoislands, as pointed outby Collina-Girard
(2001, 2003, 2004), and form basis of Spartel Bank hypothesis.
INTRODUCTION
In recent years several studies have sought
to explain the origin of legends and myths
deeply rooted in ancient cultures in terms of
geological phenomena. Faulting and hydro-
carbon gas emissions were demonstrated to
have existed at the temple of Apollo in Del-
phi, Greece, and were reported in ancient doc-
uments to have influenced the oracle (Piccardi,
2000; de Boer et al., 2001). The paleogeog-
raphy of the ancient harbor of Illium (Troy)
was investigated using modern sedimentolog-
ical techniques (Kraft et al., 2003) and was
found to correspond closely to the Homeric
accounts. The recurrent deluge story (e.g., in
the Epic of Gilgamesh, Greek mythology, and
the Book of Genesis in the Old Testament) has
been interpreted in terms of the catastrophic
flooding of settlements along the Black Sea as
the Bosphorous spillway was breached ca.
5500 B.C. (Ryan and Pitman, 1998; Lerico-
lais, 2001). Several authors have attributed the
biblical accounts in Exodus (Old Testament)
to the catastrophic eruption of Santorini (The-
ra), Greece, ca. 1600 B.C., the ash falls of
which may have been the source of the
‘‘plague of darkness’’ in Egypt (Stanley and
Sheng, 1986; Bruins and van der Pflicht,
1996).
Archeological upheavals such as the decline
or disappearance of civilizations have been at-
tributed to severe natural disasters (e.g., vol-
canic eruptions and/or earthquakes). It has
been suggested that the abandoning and/or de-
struction of numerous cities in the eastern
Mediterranean ca. 1200 B.C. was partly due
to a sequence of destructive earthquakes (M
.6.5) along active plate boundaries (Nur and
Cline, 2000). The caldera and island collapse
of Thera-Santorini created tremendous ash
falls (Stanley and Sheng, 1986) and generated
a giant tsunami (McCoy and Heiken, 2000),
which has been blamed for the downfall of the
Minoan civilization. Some geologists and ar-
cheologists believe that this event may have
inspired the Atlantis legend (Galanopolous
and Bacon, 1969).
Geographical similarities between paleois-
lands in the western Straits of Gibraltar,which
existed during and shortly after the Last Gla-
cial Maximum (LGM), between 20 and 11 ka,
have been discussed (Collina-Girard, 2001). It
was proposed that their gradual inundation by
rising sea levels may have provided the basis
for the Atlantis legend (Collina-Girard, 2001),
and strong earthquakes (Lisbon, 1755) in the
region were also noted (Collina-Girard, 2003,
2004). The purpose of this paper is to examine
this Spartel Bank hypothesis in light of new
evidence on the tectonics and paleoseismolo-
gy of the Gulf of Cadiz–Straits of Gibraltar
region (Gutscher et al., 2002; Gutscher,2004).
New high-resolution bathymetric data from
Spartel paleoisland are presented, and the vi-
ability of human habitation on this paleoisland
between 14 and 9 ka is examined.
PLATO AND THE GEOGRAPHY AND
CHRONOLOGY OF ATLANTIS
The earliest surviving written records de-
scribing an ancient Atlantis culture are the di-

686 GEOLOGY, August 2005
Figure 2. Regional map showing effects of great Lisbon earthquake of
1 November 1755, with isoseismals shown in color (after Martinez-
Solares et al., 1979). Historically reported tsunami run-up heights are
shown (Baptista et al., 1998). Initial seafloor displacement for east-
dipping subduction fault plane is shown (Gutscher et al., 2005). Gorringe
Bank has also been proposed as source of 1755 earthquake (Johnston,
1996). Inset shows isoseismals in Europe and Africa (Johnston, 1996).
alogues of Plato, The Timaeus and The Critias
(Plato, 360 B.C.), wherein key details are
found concerning the geography and chronol-
ogy pertinent to the Spartel Bank hypothesis
and to the sudden destruction of Atlantis. Only
the most important elements are summed here;
an extensive discussion of the Spartel Bank
hypothesis can be found in Collina-Girard
(2001, 2003, 2004).
The chronology given by Plato indicates de-
struction of Atlantis ca. 11.6 ka, 9 k.y. before
Egyptian priests in Sais recounted the tale to
Solon. (Solon lived ca. 600 B.C.; Plato lived
from 420 to 340 B.C.). A small island is de-
scribed, located in the Atlantic beyond the
Straits of Gibraltar (Fig. 1A), ‘‘This power
came forth out of the Atlantic Ocean...and
there was an island situated in front of the
straits which are by you called the Pillars of
Heracles.’’ The distance to the center of the
island is given as 50 stadia, or ;7.5 km (1
stadium 5150 m). In The Timaeus, the sud-
den destruction is described: ‘‘there occurred
violent earthquakes and floods; and in a single
day and night of misfortune all your warlike
men in a body sank into the earth, and the
island of Atlantis in like manner disappeared
in the depths of the sea. For which reason the
sea in those parts is impassable and impene-
trable, because there is a shoal of mud in the
way; and this was caused by the subsidence
of the island.’’
SEA-LEVEL CHANGES AND THE
SPARTEL BANK HYPOTHESIS
During the LGM (20–15 ka), global eustat-
ic sea level was 130–100 m lower than present
and rose to 250 m by 11 ka (Labeyrie et al.,
1987; Bard et al., 1996) (Fig. 1B). Sea level
remained fairly stationary at ;265 m from
ca. 12.5 ka until 11.5 ka (Bard et al., 1996).
These eustatic sea-level changes were due to
melting of the ice sheets at the onset of the
most recent interglacial period.
In the western Straits of Gibraltar, several
shoals rise to within 50 m of sea level and
were islands prior to 11 ka (Collina-Girard,
2001). The largest of these paleoislands was
;5–6 km in size, and may be a candidate for
a formerly inhabited sunken island (Fig. 1C).
Both the effect of lower sea levels on the pa-
leocoastline and the presence of islands in the
western Straits of Gibraltar were discussed ex-
tensively as being the possible origin of the
Atlantis legend (Collina-Girard, 2001, 2003,
2004). The possibility of oral transmission of
this inundation event over a period of 6 k.y.
(until the advent of written records) was also
discussed at length (Collina-Girard 2001,
2003, 2004). The Spartel Bank hypothesis as
outlined in these studies emphasized gradual
destruction by inundation lasting several cen-
turies (due to a sea-level rise of 4 m per cen-
tury). However, the sudden destruction de-
scribed by Plato (in a single day and night)
requires a catastrophic event.
REGIONAL TECTONICS AND
PALEOSEISMOLOGY
On 1 November 1755, the great Lisbon
earthquake (estimated Mw 58.5–9) struck
southwest Iberia and northwest Morocco
(Johnston, 1996; Gutscher, 2004)(Fig. 2). Ob-
served intensities from Cadiz (Spain) and Tan-
giers (Morocco) were I 57, suggesting sim-
ilar intensities in the western Straits of
Gibraltar (Martinez-Solares et al., 1979; Lev-
ret, 1991). The associated tsunami devastated
the Gulf of Cadiz region, with reported run-
up heights exceeding 5 m for port cities in
southwest Iberia and northwest Morocco
(Baptista et al., 1998) (Fig. 2).
Recent evidence supports the existence of
an active subduction zone beneath the Gulf of
Cadiz and Straits of Gibraltar (Gutscher et al.,
2002; Gutscher, 2004), that poses a long-term
risk of great earthquakes (Fig. 3). The poten-
tial seismogenic zone, with mean dimensions
estimated as 180 3210 km, is capable of gen-
erating earthquakes of Mw 8.6–8.8 with a pe-
riodicity of 1–2 k.y. (Gutscher et al., 2005).
Tsunami modeling of a subduction source in-
dicates a strong focusing effect in the eastern
Gulf of Cadiz–Straits of Gibraltar area, which
amplifies wave heights (Gutscher et al., 2005).
This is in agreement with historical reports of
extreme wave heights (15 m in Cadiz, 17 m
in Tangiers) observed in nearby cities (Baptis-
ta et al., 1998).
Two different types of sedimentological
data from the Gulf of Cadiz area suggest that
great earthquakes and tsunamis occur with a
periodicity of 1.5–2 k.y. Coarse-grained
tsunami-induced deposits in the lagoon near
Cadiz correlate with the 1755 earthquake, and
indicate a tsunami height .6 m in order to
wash over the barrier of the sand bar (Luque
et al., 2001). An older coarse-grained deposit
is dated as 200 B.C., thus suggesting a period
of 2 k.y. (Luque et al., 2001). Sediment cores
from the Horseshoe abyssal plain (Lebreiro et
al., 1997) indicate 8 major turbidites since 12
ka, which may be markers of great earth-
quakes in the past (Fig. 3). The most recent
turbidite (H1) is 10–25 cm thick and has been
dated as being contemporaneous with the
1755 earthquake (Thomson and Weaver,
1994). If the turbidites record the history of
great earthquakes, then a repeat time of ;1.5
k.y. is indicated (Gutscher, 2004). Turbidite
H8 has a mean thickness of 50–120 cm and a
total estimated volume of 5.8 km
3
(Lebreiro
et al., 1997) (Fig. 3). It is the thickest of the
postglacial series and has been dated as 12.05
ka (Lebreiro et al., 1997). For comparison, the
turbidite associated with the great Lisbon
earthquake of 1755 has an estimated volume
of ;1km
3
.
NEW BATHYMETRIC DATA FROM
SPARTEL PALEOISLAND
In July 2003, high-resolution bathymetric
data from Spartel paleoisland were acquired
with R/V Le Suroit (Fig. 4). At the 130 m
depth contour (lowest sea-level stand during
the most recent glacial maximum ca. 20 ka),

GEOLOGY, August 2005 687
Figure 3. Left: Map of Gulf of Cadiz region; thickness of turbidite H8 in Horseshoeabyssal
plain is indicated (after Lebreiro et al., 1997). Core locations are shown by white circles and
major faults are shown as thick lines. Estimated depth to top of subducting plate is shown
(in kilometers). Right: Schematic stratigraphy based on cores from Horseshoe abyssal plain
(after Lebreiro et al., 1997).
Figure 5. Outline of Spartel paleoisland as
function of time and rising sea levels: left—
chronology based on eustatic sea-level var-
iations only; right—chronology assuming 40
m of tectonic subsidence since 12 ka.
Figure 4. High-resolu-
tion bathymetric map (5
m grid spacing) of
Spartel paleoisland.
Data were acquired by
R/V
Le Suroit
(using
Simrad EM300 multi-
beam system) in July
2003 during TV-GIB
cruise.
an island of 6.5 km length (ENE-WSW) and
4 km width was present (Fig. 5). This is much
smaller than the 14-km-long paleoisland sug-
gested in earlier studies, on the basis of less
accurate hydrographic and navigation maps of
the area (Collina-Girard, 2001, 2004).
The western and southern portions of the
paleoisland were flattest, and today present the
aspect of a paleoterrace at ;120 m water
depth that may record a prolonged sea-level
lowstand. The backbone of the paleoisland is
an ENE-trending ridge at ;60–90 m water
depth, with a second morphologic high situ-
ated to the SE. These highs are marked by
slightly curved parallel bands, likely outcrops
of strata, possibly folded sedimentary flysch
of the outer Betic and Rif allochthonous units.
The 120 m to 100 m depth contours outline a
reduced island of ;5 km length, with a shel-
tered bay, facing east toward the Mediterra-
nean (Figs. 4 and 5). The 90 m and 80 m
contours define a scattered archipelago, no
wider than a few hundred meters, consisting
only of the rocky ridges and not likely to be
hospitable to habitation (Fig. 5). Thus, assum-
ing only eustatic sea-level variations, Spartel
paleoisland would have been reduced to wave-
swept rocky islets by 13 ka at the latest
(Fig. 5).
DISCUSSION
One of the most remarkable coincidences is
that the type of destruction described by Plato
(in a single day and night, by violent earth-
quakes and floods) is a very accurate descrip-
tion of the sudden (catastrophic) destruction
associated with a great (M .8) earthquake.
In 1755, tsunami waves persisted for as long
as ;24 h (Baptista et al., 1998) and likewise
following the 26 December 2004, tsunami in
the Indian Ocean. The occurrence of this type
of earthquake and tsunami in the geographic
region chosen by Plato for his narrative ap-
pears to be more than just fortuitous.
The Gulf of Cadiz–Straits of Gibraltar re-
gion is above an east-dipping subduction zone
(Gutscher et al., 2002), apparently marked by
a wide locked seismogenic zone and a long
repeat interval (as much as 2 k.y.) between
great earthquakes (Gutscher, 2004). Subduc-
tion zones are environments of locally strong
uplift and strong subsidence. Coseismic sub-
sidence caused by great earthquakes in the
Cascadia forearc are reported to be 0.5–2 m
(Clague, 1997), and coseismic subsidence of
1–2 m was observed for the great Alaska
earthquake of 1964 (Holdahl and Sauber,
1994). During the great Sumatra earthquake of
December 2004, coastlines were significantly
changed through the combined effects of the
tsunami-induced erosion and local earthquake-
induced subsidence. Some low-lying islands
were partially submerged. Spartel paleoisland
is located in the foreland of the Betic-Rif
mountain belt. Land studies of Pliocene–
Quaternary marine terraces exposed along the
Spanish side of the Straits of Gibraltar suggest
continuing active tectonic uplift of the Internal
Betic-Rif units at a rate of ;1 mm/yr (Zazo
et al., 1999) and subsidence in the foreland
region.
Assuming only eustatic sea-level variation,
Spartel paleoisland would have been uninhab-

688 GEOLOGY, August 2005
itable at 11.6 ka (because it would have been
reduced to small rocky islets ,500 m in size;
see Fig. 5). In order for the island to have been
inhabitable at the time described in Plato’s di-
alogues, at least 40 m of total tectonic subsi-
dence must have affected the island since then
(which represents a mean subsidence rate of
3.5 mm/yr). In such a scenario, the 100 m
depth contour line would represent the paleo-
shoreline ca. 11.6 ka (Fig. 5). Subsidence of
40 m could possibly be explained by ;5m
of coseismic tectonic subsidence during each
of the 8 great earthquakes during the past 12
k.y. (the number of earthquakes indicated by
the turbidite record). This amount of subsi-
dence is greater than that known for compa-
rable tectonic settings (Cascadia, Alaska).
However, the presence of an additional crustal
fault may account for the remainder. Rapidly
subsiding and tectonically active regions like
the Gulf of Corinth, Greece (Armijo et al.,
1996), can exhibit modern-day subsidence at
rates .5 mm/yr.
If the earthquake ca. 11.6 ka was excep-
tionally large, as suggested by the great thick-
ness of turbidite H8 found in the Horseshoe
abyssal plain (Fig. 3), then perhaps as much
as 10 m of coseismic subsidence may have
occurred. If one then considers the added im-
pact of a 10-m-high tsunami wave (Fig. 2),
then everything within 20 m of the previous
sea level would have been obliterated. The re-
maining rocky islets, after the sudden and ir-
reversible 10 m immersion into the sea, would
bear little resemblance to the formerly 5-km-
long island with its east-facing bay (Fig. 5).
The combined effects of the continuing rise in
sea level and additional subsidence due to
earthquakes would completely submerge the
remnants within a few thousand years.
CONCLUSIONS
The high-resolution bathymetric data ac-
quired on Spartel Bank indicate a significantly
smaller island at the 130 m contour (6.5 km
34 km) than the 15-km-long island reported
in previous studies (Collina-Girard, 2001,
2003, 2004). Furthermore, assuming only eu-
static sea-level variation, Spartel paleoisland
would have been uninhabitable at 11.6 ka (two
small rocky islets ,500 m in size). Thus,
these new bathymetric data, taken alone, do
not confirm the Spartel Bank hypothesis; rath-
er, they render it highly unlikely. However,
taking into account strong tectonic subsidence
due to great earthquakes, and the sudden de-
struction by a great tsunami, the Spartel Bank
hypothesis may be viable. Although the cata-
strophic destruction described by Plato is con-
sistent with the geological and tectonic history
of the Straits of Gibraltar, this does not imply
that Atlantis ever existed. It simply means the
account is geologically plausible. The ques-
tion remains, was the paleoisland of Spartel
inhabited nearly 12 k.y. ago? In order to ob-
tain an answer, it will be necessary to conduct
a detailed survey of the seafloor in this area
and search for signs of construction or
artifacts.
ACKNOWLEDGMENTS
I thank the Captain and crew of the R/V Le Su-
roit for their fine work during the cruise TV-GIB
and Graeme Cairns, Stephane Dominguez, and oth-
ers for their inspiration and direct participation. I
also thank David Fastovsky and Renee Hethering-
ton for constructive suggestions that helped clarify
the focus of this paper. This is contribution 948 of
the Institut Universitaire Europe´en de la Mer.
REFERENCES CITED
Armijo, R., Meyer, B., King, G.C.P., Rigo, A., and
Papanastassiou, D., 1996, Quaternary evolu-
tion of the Corinth Rift and implications for
the late Cenozoic evolution of the Aegean:
Geophysical Journal International, v. 126,
p. 11–53.
Baptista, M.A., Heitor, S., Miranda, J.M., Miranda,
P.M.A., and Mendes Victor, L., 1998, The
1755 Lisbon earthquake; evaluation of the tsu-
nami parameters: Journal of Geodynamics,
v. 25, p. 143–157.
Bard, E., Hamelin, B., Arnold, M., Montaggioni, L.,
Cabioch, G., Faure, G., and Rougerie, F.,
1996, Deglacial sea-level record from Tahiti
corals and the timing of global meltwater dis-
charge: Nature, v. 382, p. 241–244.
Bruins, H.J., and van der Pflicht, J., 1996, The Ex-
odus enigma: Nature, v. 382, p. 213–214.
Clague, J.J., 1997, Evidence for large earthquakes
at the Cascadia subduction zone: Reviews in
Geophysics, v. 35, p. 439–460.
Collina-Girard, J., 2001, Atlantis off the Gibraltar
Strait? Myth and geology: Paris, Acade´mie des
Sciences Comptes Rendus, v. 333, p. 233–240.
Collina-Girard, J., 2003, La transgression finiglaci-
aire, l’arche´ologie et les texts (exemple de la
grotte de Cosquer et du mythe de l’Atlantide):
CIESM Workshop Monographs, v. 24,
p. 63–70, www.ciesm.org/publications/
Santorini04.pdf.
Collina-Girard, J., 2004, Du vestige ge´ologique au
vestige litteraire, Gibraltar et l’Atlantide:
LUKHNOS Connaissances Helenniques,
v. 100, p. 9–21.
de Boer, J.Z., Hale, J.R., and Chanton, J., 2001,
New evidence for the geological origins of the
ancient Delphic oracle (Greece): Geology,
v. 29, p. 707–710.
Galanopolous, A.G., and Bacon, E., 1969, Atlantis:
The truth behind the legend: New York,
Bobbs-Merill, 216 p.
Gutscher, M.-A., 2004, What caused the Great
Lisbon earthquake?: Science, v. 305,
p. 1247–1248.
Gutscher, M.-A., Malod, J., Rehault, J.-P., Contruc-
ci, I., Klingelhoefer, F., Mendes-Victor, L., and
Spakman, W., 2002, Evidence for active sub-
duction beneath Gibraltar: Geology, v. 30,
p. 1071–1074.
Gutscher, M.-A., Baptista, M.A., and Miranda, J.M.,
2005, The Gibraltar Arc seismogenic zone
(part 2): Constraints on a shallow east-dipping
fault plane source for the 1755 Lisbon earth-
quake provided by tsunami modeling and seis-
mic intensity: Tectonophysics (in press).
Holdahl, S.R., and Sauber, J., 1994, Co-seismic slip
in the 1964 Prince William Sound earthquake:
A new geodetic inversion: Pure and Applied
Geophysics, v. 142, p. 55–81.
Johnston, A., 1996, Seismic moment assessment of
earthquakes in stable continental regions—III.
New Madrid, 1811–1812, Charleston 1886
and Lisbon 1755: Geophysical Journal Inter-
national, v. 126, p. 314–344.
Kraft, J.C., Rapp, G., Kayan, I., and Luce, J.V.,
2003, Harbor areas at ancient Troy: Sedimen-
tology and geomorphology complement Ho-
mer’s Illiad: Geology, v. 31, p. 163–166.
Labeyrie, L.D., Duplessey, J.C., and Blanc, P.L.,
1987, Variations in the mode of formation and
temperature of oceanic deep waters over
the past 125,000 years: Nature, v. 327,
p. 477–482.
Lebreiro, S.M., McCave, I.N., and Weaver, P., 1997,
Late Quaternary turbidite emplacement on the
Horseshoe abyssal plain (Iberian margin):
Journal of Sedimentary Research, v. 67,
p. 856–870.
Lericolais, G., 2001, La catastrophe du Bosphore:
Pour la Science, v. 284, p. 30–37.
Levret, A., 1991, The effects of the November 1,
1755 ‘‘Lisbon’’ earthquake in Morocco: Tec-
tonophysics, v. 193, p. 83–94.
Luque, L., Lario, J., Zazo, C., Goy, J.L., Dabrio,
C.J., and Silva, P.G., 2001, Tsunami deposits
as paleoseismic indicators: Examples from the
Spanish coast: Acta Geologica Hispanica,
v. 36, p. 197–211.
Martinez-Solares, J.M., Lopez, A., and Mezcua, J.,
1979, Isoseismal map of the 1755 Lisbon
earthquake obtained from Spanish data: Tec-
tonophysics, v. 53, p. 301–313.
McCoy, F.W., and Heiken, G., 2000, Tsunami gen-
erated by the late Bronze Age eruption of The-
ra (Santorini), Greece, in Keating, B.H., et al.,
eds., Landslides and tsunamis: Pure and Ap-
plied Geophysics, 157, p. 1227–1256.
Nur, A., and Cline, E.H., 2000, Poseidon’s horses:
Plate tectonics and earthquake storms in the
late Bronze Age Aegean and Eastern Mediter-
ranean: Journal of Archaeological Science,
v. 27, p. 43–63.
Piccardi, L., 2000, Active faulting at Delphi,
Greece: Seismotectonic remarks and a hypoth-
esis for the geologic environment of a myth:
Geology, v. 28, p. 651–654.
Ryan, W., and Pitman, W., 1998, Noah’s flood: The
new scientific discoveries about the event that
changed history: New York, Simon and
Schuster, 319 p.
Stanley, D.J., and Sheng, H., 1986, Volcanic shards
from Santorini (Upper Minoan ash) in the Nile
Delta, Egypt: Nature, v. 360, p. 733–734.
Thomson, J., and Weaver, P., 1994, An AMS radio-
carbon method to determine the emplacement
time of recent deep-sea turbidites: Sedimen-
tary Geology, v. 89, p. 1–7.
Zazo, C., Silva, P.G., Roy, J.L., Hillaire-Marcel, C.,
Ghaleb, B., Lario, J., Bardaji, T., and Gonza-
lez, A., 1999, Coastal uplift in continental col-
lision plate boundaries: Data from the last in-
terglacial marine terraces of the Gibraltar
Strait area (south Spain): Tectonophysics,
v. 301, p. 95–109.
Manuscript received 21 February 2005
Revised manuscript received 4 April 2005
Manuscript accepted 18 April 2005
Printed in USA