The International Journal of Nautical Archaeology (2006) 35.2: 187–194
doi: 10.1111/j.1095-9270.2006.00113.x
Florida’s Mystery Wreck
FLORIDA’S
J.
F. Mckinnon
MYSTERY
& D. A.
WRECK
Blackwell
Publishing
LtdScott-Ireton
Jennifer F. McKinnon
Flinders University, Department of Archaeology, GPO Box 2100, Adelaide 5001, Australia
Della A. Scott-Ireton
Florida Public Archaeology Network, 212 East Church Street, Pensacola, Florida 32591, USA
During the summer of 2005 the Florida Bureau of Archaeological Research and the Florida Keys National Marine Sanctuary
investigated an unusual shipwreck in 6 m (20 ft) of water off Marathon in the Florida Keys. Known locally as the ‘Mystery
Wreck’ it remained relatively undisturbed for over two decades until State of Florida and Sanctuary archaeologists conducted
a research project to learn more about this fascinating early shipwreck. This paper presents a summary of field investigations
and a preliminary interpretation of the shipwreck.
© 2006 The Authors
Key words: shipwreck, Florida, Iberian, ecosystem, ballast.
D
uring the summer of 2005 the Florida
Bureau of Archaeological Research
and the Florida Keys National Marine
Sanctuary (FKNMS) investigated an unusual
shipwreck in 6 m (20 ft) of water on the edge of
Hawk Channel off Marathon in the Florida Keys
(Fig. 1).
In May 1972 the Florida Division of Archives,
History, and Records Management (later the
Division of Historical Resources) entered into a
salvage contract with Brown, Knott, and Nipp
Enterprises to explore and salvage the remains
of an unidentified possible colonial-Spanish
shipwreck. The ballast-mound was known by
local divers but had never been investigated. The
salvage company believed the wreck to be one
of the remaining undiscovered wrecks of the
Spanish Plate Fleet lost in a hurricane in 1733
(Walton, 1994; Weller, 2001). Wrecks of the 1733
fleet were being discovered during the 1960s and
’70s and were producing small amounts of coins
and other artefacts which fuelled the search for the
remaining ships, two of which remain undiscovered.
The salvors, according to the terms of their
contract, performed limited excavations around
the edges of the ballast-mound using dredges,
blowers, and hand-tools. A section of ballast near
the middle of the mound was removed in a futile
search for non-existent ‘treasure’. Artefacts recovered
included ballast-stones, encrusted iron objects,
208 musket balls, 127 grape shot, 34 lead shot,
three cannonballs, firebricks, several fragments
of wood, a rudder gudgeon, a piece of melted
copper, one piece of bone, lead sheathing, a
broken mano (grinding stone), an anchor and
anchor ring, bags of olive-jar body sherds and
four olive-jar necks. These artefacts were conserved
and are curated at the Division of Historical
Resources in Tallahassee, Florida.
The lack of precious metals indicated that the
Mystery Wreck was not one of the 1733 fleet and
the salvors soon lost interest. Their contract lapsed
and no further organized investigations were
undertaken at the site. However, the shipwreck,
known locally as the ‘Mystery Wreck’, was impacted
over the years by occasional visitors who, in a
half-hearted search for treasure, disturbed sections
of the ballast pile. It was even featured in a video
called ‘Galleon Hunter’ in which a local diver
visited the 1733 sites and other shipwrecks,
promoting their histories and diving adventures.
But in terms of serious study it remained
undisturbed for over two decades until State of
Florida and Sanctuary archaeologists conducted
a research project to learn more about this
unusual early shipwreck.
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society.
Published by Blackwell Publishing Ltd. 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
NAUTICAL ARCHAEOLOGY, 35.2
Figure 1. Mystery Wreck location map. (courtesy of Jeff Moates)
In the summer of 2004 the underwater
archaeological team of the Florida Bureau of
Archaeological Research conducted fieldwork
as part of a National Oceanic and Atmospheric
Administration (NOAA) grant to research and
record the 1733 Spanish Plate Fleet wrecks
with the goal of producing public educational
materials. During this research project local
informant Dave McCampbell offered to show the
team additional unrecorded shipwrecks; one of
these was the Mystery Wreck. The site presented
as a compact and consolidated ballast-mound
composed of square-shaped cut igneous stones
(probably granite) atop a patch reef on the edge
of Hawk Channel in the Atlantic Ocean, 3.2 km
(2 miles) south of the city of Marathon on Vaca
Key. Salvage activities and the natural scouring
action of currents and waves had produced
cavities under the ballast that revealed massive
ship timbers. An undercut at the seaward end of
the ballast exposed stern timbers and as many as
eight framing stations extending under the ballast.
An odd crack or crevice bisected the otherwisecompact ballast-pile. The reef surrounding the
shipwreck was in pristine condition including an
array of hard and soft corals. The entire site was
home to a thriving ecosystem of tropical sealife
from the smallest of crustaceans to large predators
such as sharks and barracudas. The intriguing
site with its compact, nearly-undisturbed ballastmound and visible construction features presented
188
an opportunity to research early maritime cultural
resources in the Florida Keys. Although construction
details suggested the Mystery Wreck was a colonial
shipwreck, its square-cut ballast, completely
unlike the smooth, egg-shaped river-rock ballast
of the 1733 wrecks, indicated that this shipwreck
was not related to the 1733 fleet disaster.
The research potential of the shipwreck,
together with the Sanctuary’s continuing efforts
to record submerged cultural resources in Florida
waters, provided an opportunity for the State
and Sanctuary to work together. In 1998 a
Programmatic Agreement between NOAA and
the State of Florida for historical resource
management in the Florida Keys National Marine
Sanctuary was signed (National Oceanic and
Atmospheric Administration et al., 1998). The
agreement provided a framework for joint
management of submerged cultural resources
within the Sanctuary. This formal partnership
enabled a research project to be formed for the
investigation of the Mystery Wreck. A grant
from the NOAA National Marine Sanctuary
Program’s Maritime Heritage Program provided
funding for a 10-day project in June 2005 to
perform intensive investigations on the Mystery
Wreck. Operations were based from the NOAA
research vessel R/ V Odyssey with dock space and
support provided by the Marathon City Marina.
The project was a collaborative effort involving
professional scientists, volunteers, and the local
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society
J. F. MCKINNON & D. A. SCOTT-IRETON: FLORIDA’S MYSTERY WRECK
Figure 2. Crevice bisecting the ballast-mound. (courtesy of
the Florida Bureau of Archaeological Research)
community. News media were invited to cover the
project and diving reporters were given a tour of
the wreck-site. Aerial photographs of the site and
surrounding area were provided by the Florida
Fish and Wildlife Conservation Commission. In
addition to State of Florida and NOAA archaeologists, visiting scholars from the City of Key
West, the Mel Fisher Maritime Heritage Society
(MFMHS), and NOAA headquarters in Silver
Spring, Maryland, helped to investigate the wreck.
A mooring was placed at the wreck-site for
the research vessels, and diving operations were
conducted from the boats. A baseline was
established along the longitudinal axis of the
wreck, which lies in a north-east/south-west
direction with the vessel’s bow to the north-east.
Because vertical relief of the ballast mound exceeds
1.5 m, a secondary baseline was established along
the starboard side of the vessel off the mound.
This was used to record the undercut portions of
the ballast and protruding framing stations and
other ship features. The extent of the ballastmound was recorded together with large coral
features growing on top of it. Profile measurements
were obtained using a plumb-bob dropped from
a surface buoy to record the unusually high
vertical relief of the ballast and surrounding
patch reef. Approximately 6 m forward from the
aft end of the mound, the otherwise consolidated
ballast is bisected by a deep transverse crack that
extends the entire width of the mound (Fig. 2).
The fissure, up to 70 cm wide, adds to the mystery
of the wreck and provides shelter for small
tropical fish and invertebrates.
The ship’s length of keel is approximately
19 m; its beam is unknown. Ballast extends
beyond the exposed ship-remains and measures
22 × 15 m. Exposed sections of the hull include
stern timbers, portions of the bow assembly, and
timbers along the midship area. Stern timbers
comprise the eroded sternpost, three closelyspaced tail-frames, and the remains of both port
and starboard garboard strakes. Remains of the
bow include the forward end of the keel and two
small, curved disarticulated bow frames. Timbers
along the midships, exposed by past salvage
activities, represent the vessel’s keelson, two
floors, a rider, a fragment of ceiling plank, and
what appears to be a small section of the pumpbox (Fig. 3). Wood samples from all accessible
timbers were analysed by Dr Lee Newsom at
Pennsylvania State University (Table 1).
An unusual site-formation feature was
encountered at an area of what initially appeared
to be frames exposed from the stern to midship
on the starboard side (Fig. 4). Salvors, instead of
digging straight down into the wreck through the
ballast in a systematic manner, picked at the
loose stones along the lower edge of the mound.
Because the wreck rested on a coral reef rather
than on sand, stones that had fallen along the
bottom exterior of the hull could be accessed
and removed. This created a surprising and
interesting visual experience. The wreck appears
to be resting on its keel with lower hull planking
removed, allowing a diver to look into the
bottom portion of the shipwreck and see intact
frames exposed underneath the ballast mound
and slightly protruding from the edge of the
mound. Shining a light into the void between
frames under the ballast shows sections of the
keel, as well as groupers, large lobsters, and a
huge green moray eel living in the space.
Closer inspection of the ‘frames’ revealed they
were not made of wood, but seemed to be
composed of a concrete-like substance (Fig. 5).
Corey Malcom, of the MFMHS, was aware of a
sort of cement called granel consisting of lime,
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society
189
NAUTICAL ARCHAEOLOGY, 35.2
Figure 3. Drawing of midship timbers. (courtesy of Irina Franklin)
Table 1. Scantlings and wood species
Timber
Length
Moulded
Sided
Wood Species
Keel
∼19 m
Stern: 0.25 m
Stern: 0.20 –0.33 m
Bow: 0.31 m
0.25 m
0.15 –0.2 m
0.16 m
0.18 m
0.19 m
0.19 m
0.2 m
Stern: 0.18–0.2 m
Bow: 0.16–0.25 m
Quercus sp.
(white oak)
Quercus sp.
Quercus sp.
Quercus sp.
Keelson
Sternpost
Rider
Tail-frame
Tail-frame
Tail-frame
Tail-frame
Floors
0.15 m
0.4 m
1
2
3
4
0.33 m
Stern: 0.33 m
Bow: 0.5 m
Garboards
Ceiling
Hull planking
Logwood
1.6 m
sand, and small pebbles that sometimes was
poured between the frames of 17th-century
Iberian ships. This mixture was specified by
contract to be used in the building of Nuestra
Señora de Atocha, wrecked in 1622 (Alonso
Ferrera, contract of 1616 (Archivo General de
Indias (AGI) Contratación 4895) for four 550ton galleons, including Atocha). The granel acted
as permanent ballast, as well as protecting the
floors and lower frames from damage due to
shifting cargo. The same or a similar substance
190
#19:
Quercus sp.
Picea sp. (spruce)
probably Picea sp.
Pinus sp. (pine) including
P. halepensis (‘Aleppo pine’) and
P. canariensis (Canary Islands pine)
Guaiacum sp. (lignum vitae)
seems to have been used in the construction of
the Mystery Wreck. What first was thought to be
wooden frames actually is poured permanent
ballast that filled the spaces between frames,
providing a sort of negative image of the lower
hull timbers. The poured ballast extended to
approximately the turn of the bilge on the
starboard side and allowed the recording of
frame spacing (0.25 m at bow and stern; 0.45 m
midships). The sternmost existing poured ballast
feature appears to have an arched bottom which
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society
J. F. MCKINNON & D. A. SCOTT-IRETON: FLORIDA’S MYSTERY WRECK
Figure 4.
Figure 5.
Ballast-mound with exposed ship structure. (courtesy of Jeff Anderson)
Permanent ballast frame ‘negatives.’ (courtesy of the Florida Bureau of Archaeological Research)
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society
191
Figure 6.
Mystery Wreck site plan. (courtesy of the Florida Bureau of Archaeological Research)
NAUTICAL ARCHAEOLOGY, 35.2
192
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society
J. F. MCKINNON & D. A. SCOTT-IRETON: FLORIDA’S MYSTERY WRECK
would have allowed water to flow to the ship’s
pump. Additional ballast features, however, were
not accessible to verify arching of the bottom face.
The accompanying site-plan indicates the
complex cultural and natural features of the
Mystery Wreck site (Fig. 6). The central
undisturbed ballast of compact stones is denoted
by dark shading. Scattered ballast on the
outboard edges is less compact, the result of
natural site-formation processes as well as past
salvage activities. Reef structure including hard
and soft corals is indicated around the ballastmound. Few corals have colonized the central
mound, but a dense and vibrant forest of colourful
corals and sea-fans surrounds the wreck-site. The
ship is embedded in living reef, not all of which
could be depicted in this perspective. Extreme bow
and stern timbers are exposed at the forward and
aft ends of the ballast-mound. ‘Frames’ protruding
from the ballast are negatives formed by the poured
permanent ballast. More ship structure, indicated
by dotted lines, is visible than could be shown in
plan view. This structure extends at least 4 m into
the void under the ballast. The fissure can be seen
bisecting the stern section. A mountain star coral
cored to determine its age is visible on the ballast
near the starboard midship edge.
In addition to the recording of the cultural
features of the site, the natural features were
recorded by NOAA. The patch reef surrounding
the Mystery Wreck is seldom visited by sport
divers because of unpredictable visibility (caused
by its location on the edge of the murky water
of Hawk Channel). As a result, the site has not
suffered from excessive human impact, unlike
many other reef areas in the Keys. Together with
the lack of human impact, the nutrient-rich water
of Hawk Channel provides an ideal habitat for
coral formation. Several species were recorded,
including giant brain coral (Colpophyllia natans),
several species of mountain star coral (Montastrea
sp.), and staghorn coral (Acropora cervircornis).
Some of the brain corals exceeded 4 m in
diameter, and the largest mountain star coral
stood over 2 m off the bottom. One mountain
star coral, 1.5 m in diameter, growing atop the
ballast, was cored to determine its age, which
exceeded 100 years. A plethora of purple sea-fans
(Gorgonia ventalina) and other soft corals are
attached to ballast-stones and extend into the
reef structure around the wreck. This habitat
provides shelter and feeding grounds for marine
animals including grey angelfish, nurse sharks,
sea turtles, damselfish, barracudas, groupers,
Figure 7. Tropical sea-life inhabiting the Mystery Wreck.
(courtesy of the Florida Bureau of Archaeological Research)
snappers, lobsters, molluscs and crustaceans of
many varieties (Fig. 7).
Few artefacts were encountered during the
investigation— the result of indiscriminate collecting by the initial salvors and scavenging by
random sport divers. A piece of kaolin pipe-stem
(internal diameter 1/8 in) and a small sherd of
undecorated white tin-glazed coarse earthenware,
identified as majolica, were recovered near the
pump-box. The largest artefact recovered was a
piece of logwood wedged under the forwardmost
bow frame. Made of lignum vitae, measuring 1.6 m
(just over 5 ft) long and with one carved end, it
initially was thought possibly to be a gunner’s
wedge or similar tool, used to try to heave the
ship off the reef where it had grounded. Further
analysis indicates it probably was among the
cargo and became trapped under the hull as the
ship disintegrated. Small bits of olive jar of
indeterminate age were scattered around the
shipwreck, but not collected.
Based on analysis of artefacts recovered by the
salvors, in addition to construction features and
material collected during the project, the Mystery
Wreck appears to have been a small vessel built
in the Iberian tradition, approximately 19 m long
and at least 7 m in beam. It possibly was an aviso
or dispatch vessel travelling in company with
other ships. Some time during the first half of the
17th century, the ship was sailing along the
remote islands of the Keys in sight of land when
it left the deep water of Hawk Channel and ran
hard aground on an isolated patch reef. The
vessel was damaged and stuck fast by the bow,
and the crew was unable to get it off the reef, so
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society
193
NAUTICAL ARCHAEOLOGY, 35.2
it was stripped and abandoned. As it slowly
deteriorated, the ballast concreted together in a
solid mass. When the keel finally collapsed and
no longer supported the ballast, the concreted
mass broke in two, producing the fissure that
bisects the mound. Wooden timbers not covered
by sediments were consumed by marine organisms
and destroyed by waves and currents, leaving
behind ghost impressions in permanent ballast
poured into the lower hull. The surrounding reef
community engulfed the wreck and marine life
colonized the ballast, helping to protect the remains
from natural erosional forces. Today, the Mystery
Wreck is part of the maritime heritage resources
of the Florida Keys National Marine Sanctuary
and the State of Florida and presents an opportunity
for future research and interpretation of our
colonial maritime history.
Acknowledgements
The authors wish to thank the staff of the Florida Bureau of Archaeological Research, the Florida Keys National Marine
Sanctuary, and NOAA’s Maritime Heritage Program for making this project possible. Visiting and local archaeologists,
biologists, photographers, reporters, and volunteers, too numerous to list but individually appreciated, contributed to the
success of the Mystery Wreck Project. Special appreciation is extended to Roger Smith for his knowledge and expertise, and
to Jeff Moates for assistance with graphics.
References
National Oceanic and Atmospheric Administration, the Advisory Council on Historic Preservation, and the State of Florida,
1998, Programmatic Agreement among the NOAA, the ACHP, and the State of Florida for Historical Resource Management
in the Florida Keys National Marine Sanctuary. Tallahassee, FL.
Walton, T., 1994, The Spanish Treasure Fleets. Sarasota, FL.
Weller, B., 2001, Galleon Alley: The 1733 Spanish Treasure Fleet. Lake Worth, FL.
194
© 2006 The Authors. Journal Compilation © 2006 The Nautical Archaeology Society