|
THE FLORIDA ANTHROPOLOGIST
Published by the FLORIDA ANTHROPOLOGICAL SOCIETY, INC.
VOLUME 59, NUMBERS 3-4 SEPTEMBER-DECEMBER 2006
SPECIAL ISSUE: The Miami Circle: Fieldwork, Research and Analysis II
-j-
*- s.-.-
zZ'
r i
E
78
.F6
F58
THE FLORIDA ANTHROPOLOGIST is published by the Florida Anthropological Society, Inc., P.O. Box 6356, Tallahassee, Florida 32314.
Subscription is by membership in the Society. Membership is NOT restricted to residents of the State of Florida nor to the United States
of America. Membership may be initiated at any time during the year, and covers the ensuing twelve month period. Dues shall be payable
on the anniversary of the initial dues payment. Members shall receive copies of all publications distributed by the Society during the 12
months of their membership year. Annual dues are as follows: student $15, individual $30, family $35, institutional $30, sustaining $40
or more, patron $100 or more, life $500, and benefactor $2500. Foreign subscriptions are an additional $5 U.S. to cover added postage
and handling costs for individual, family, or institutional membership categories. Copies of the journal will only be sent to members with
current paid dues. Please contact the Editor for information on recent back issues.
Requests for information on the Society, membership application forms, and notifications of changes of address should be sent to the
Membership Secretary. Donations should be sent to the Treasurer or may be routed through the Editor to facilitate acknowledgment in
subsequent issues of the journal (unless anonymity is requested). Submissions of manuscripts should be sent to the Editor. Publications
for review should be submitted to the Book Review Editor. Authors please follow The Florida Anthropologist style guide (49[1], 1996,
pp. 37-43) in preparing manuscripts for submission to the journal and contact the Editor with specific questions. Submit four (4) copies
for use in peer review. Only one set of original graphics need be submitted. The journal is formatted using Word Perfect 12.0. All
manuscripts must be submitted in final form on DOS-formatted 3.5 inch diskettes or CDs in Word Perfect, Word, or ASCII format. Address
changes should be made AT LEAST 30 DAYS prior to the mailing of the next issue. The post office will not forward bulk mail nor retain
such mail when "temporary hold" orders exist. Such mail is returned to the Society postage due. The journal is published quarterly in
March, June, September, and December of each year.
OFFICERS OF THE SOCIETY
President: Sheila Stewart, 2130 Burlington Avenue North, St. Petersburg, FL 33713 (sjstew@gte.net)
First Vice President: David Burns, 15128 Springview St., Tampa, FL 33624 (daveburns@prodigy.net)
Second Vice President: Christine Newman, 504 17th St., St. Augustine, FL 32084 (acistaugustine@bellsouth.net)
Corresponding Secretary: Antoinette Wallace, (ab_wallace@bellsouth.net)
Membership Secretary: Beth Sarney, 5331 Lansdowne Way, Palmetto, Fl 34221 (excav8@comcast.com)
Treasurer and RegisteredAgent. Joanne Talley, P.O.Box 788, Hobe Sound, FL 33475 (jo@whiticar.com)
Directors at Large: Gloria Fike, 2815 SE St. Lucie Blvd., Stuart, FL 34997 (GloriaFike@prodigy.net); Robert J. Austin, P.O. Box
2818, Riverview, FL 33568-2818 (bob@searchinc.com); Tommy Abood, 3857 Indian Trail #403, Destin, FL
32541 (lost.horizon@earthlink.net)
Immediate Past President: David Burns, 15128 Springview St., Tampa, FL 33624 (daveburns@prodigy.net)
Newsletter Editor: David Burns, 15128 Springview St., Tampa, FL 33624 (davebuns@prodigy.net)
JOURNAL EDITORIAL STAFF
Outgoing Editor: Ryan J. Wheeler, P.O. Box 6356, Tallahassee, FL 32314-6356 (rwheeler@nettally.com)
Co-Editors: Deborah Mullins, P.O. Box 357605, Gainesville, FL 32635-7605 (dmullins.fl.anthropologist@gmail.com) and Andrea P.
White, P.O. Box 6356, Tallahassee, FL 32314-6356 (apw@archaeologist.com)
Book Review Editor: Dan Hughes, 2301 8th Ave. North, St. Petersburg, Florida 33713 (hughesarc@juno.com)
Technical Editor: vacant
EditorialAssistant: George M. Luer, 3222 Old Oak Drive, Sarasota, FL 34239-5019 (gluer@grove.ufl.edu)
Printer: Durra-Print, 717 South Woodward Ave., Tallahassee, FL 32304
Bulk Mail: Capital City Mailing, 4013 Woodville Hwy, Tallahassee, FL 32311
Back Issue Sales: Palm Beach Museum of Natural History (www.pbmnh.org)
EDITORIAL REVIEW BOARD
Albert C. Goodyear, Institute of Archaeology and Anthropology, University of South Carolina, Columbia, SC 29208 (goodyear@sc.edu)
Jerald T. Milanich, Florida Museum of Natural History, University of Florida, Gainesville, FL 32611 (jtm@flmnh.ufl.edu)
Jeffrey M. Mitchem, Arkansas Archeological Survey, P.O. Box 241, Parkin, AR 72373 (jeffmitchem@juno.com)
Nancy Marie White, Department of Anthropology, University of South Florida, Tampa, FL 33620-8100 (nwhite@chumal.cas.usf.edu)
Robert J. Austin, P.O. Box 2818, Riverview, FL 33568-2818 (bob@searchinc.com)
NOTE: In addition to the above Editorial Review Board members, the review comments of others knowledgeable in a manuscript's subject
matter are solicited as part of our peer review process.
VISIT FAS ON THE WEB: www.fasweb.org
THE FLORIDA i
ANTHROPOLOGIST
*
Volume 59 Numbers 3-4 'NcE 19A1
September-December 2006
SPECIAL ISSUE: The Miami Circle: Fieldwork, Research and Analysis HI
TABLE OF CONTENTS
Editor's Page 131
Analysis of Ceramics from Brickell Point, 8DA12. Robert S. Carr 133
Integrated Spatial Technologies: High Definition Documentation
of the Miami Circle and Royal Palm Circles. Lori Collins, Travis Doering and Robert S. Carr 161
The Animal Interments at the Miami Circle at Brickell Point Site (8DA12). Alison A. Elgart 179
Pumice Artifacts from the Miami Circle at Brickell Point (8DA12). Ryan J. Wheeler 191
Geochemical and Petrologic Characterization of Pumice Artifacts from the Miami Circle-Brickell
Point Archaeological Site plus Other Sites in Florida Potential Provenance Locations. Stephen A. Kish 209
An Analysis of the Prehistoric Human Remains Found at the Miami Circle
at Brickell Point site (8DA12). Alison A. Elgart and Robert S. Carr 241
FAS 2006 AWARD RECIPIENTS 251
OBrrUARY: Dan D. Laxson. George M. Luer 253
BOOK REVIEWS
Gibson and Carr: Signs of Power The Rise of Cultural Complexity in the Southeast. Jon C. Endonino 263
Nance, Whittington, and Borg: Archaeology and Ethnohistory oflximchd. E. Christian Wells 265
Kolianos and Weisman: The Florida Journals ofFrank Hamilton Gushing and The Lost Florida Manuscript of
Frank Hamilton Gushing. Uzi Baram 266
Authors 268
Cover: Data point cloud from the 3-D laser scanning of the Miami Circle; for more information see the article by Collins, Doering and Carr
beginning on page 161.
Published by the
FLORIDA ANTHROPOLOGICAL SOCIETY, INC.
ISSN 0015-3893
EDITOR'S PAGE
This is the third Special Issue of The FloridaAnthropologist
dedicated to research on the Miami Circle site. Volume 53(4),
published in 2000, presented some of the first scientific writing
on the Miami Circle site, which was discovered in 1998 in
downtown Miami and became the focus of a grass-roots effort
to save the carved limestone feature and surrounding midden
deposits from the onslaught of redevelopment in the City of
Miami. Volume 57(1-2), published in 2004, followed with
additional articles and specialized analyses about the site. With
this issue, the Miami Circle becomes one of the best studied
archaeological sites in southern Florida. Continuing the multi-
disciplinary approach begun in 2000, this issue presents an
analysis of ceramics from the site by Project Director Bob Carr,
which documents additional connections with Middle Wood-
land cultures elsewhere in Florida that were also observed in the
site configuration and studies of some of the other artifact types
from the site. Bob's article is followed by the fascinating
research ofLori Collins and Travis Doering who used 3-D laser
scanning technology to document the features comprising the
Miami Circle and the circular features recently discovered
across the Miami River at 8DA11. Alison Elgart writes about
the animal interments from the Miami Circle-a completely
articulated shark, a sea turtle carapace and a dolphin skull-and
places them within the context of the site, other animal burials
from Florida and the broader literature on the significance of
such burials. Ryan Wheeler and Steve Kish present their
research on the pumice artifacts discovered at the Miami Circle
and their attempt to understand the role of this volcanic, non-
local material in regional exchange systems and the possible
origins of this material. The final article, by Alison Elgart and
Bob Carr, summarizes the small collection of human remains
identified from the Miami Circle.
While research on the site continues it is worth noting the
other efforts underway to protect and interpret the site. The
Bureau of Archaeological Research revamped its website on the
Miami Circle, including many photos of the site and its arti-
facts, a virtual tour, summaries of the various research projects,
and time lines of major events in the history of the site-the
website has its own address: www.miamicirclesite.com. The
question of future management of the Miami Circle is being
addressed in a study by the National Park Service. While drafts
of the study suggest that the site may not come under federal
auspices, the Florida Division of Historical Resources has
begun discussing the possibility that the Historical Museum of
Southern Florida will assume this important role. The museum
has had a long relationship with the Miami Circle, including an
exhibit dedicated to the site and the first inhabitants of southern
Florida. If the museum assumes management of the site they
will be able to develop an interpretive plan for the site and
expand on the conceptual designs presented in the National
Park Service study. Part of the bid to make the Miami Circle
part of the National Park Service included a nomination as a
National Historic Landmark. While initially meeting with some
resistance, this effort has produced a thematic study of the
Tequesta Indians and their ancestors that can be used to
nominate sites as Natioanl Historic Landmarks or for listing on
the National Register of Historic Places, as well as a successful
landmark designation for the Mud Lake Canal-a prehistoric
canoe canal located deep in the Florida Everglades. Much of
the research present here and in previous Special Issues on the
Miami Circle have been crucial in the landmark nomination.
The Florida Division of Historical Resources also continues its
plan to replace the deteriorating seawall at the site, a compli-
cated process that involves countless permits from federal,
state, county and local entities, as well as considerable sums of
money. The Division has maintained a steadfast commitment
to this project, despite the many obstacles regarding a major
construction project in downtown Miami. A visit to downtown
Miami today gives a vastly different impression than it would
have eight years ago when the Miami Circle was first discov-
ered. Many buildings from 20 to 50 years ago have succumbed
to new construction, including the IconBrickell rising next to
the Miami Circle parcel and the Epic, located across the river
where the funky old DuPont Plaza used to sit. A lot of this
construction has brought new archaeological digs and new
discoveries. Hopefully we will read about them all in future
issues of The Florida Anthropologist.
Perhaps it is auspicious that this issue, dedicated to the
Miami Circle, also will be my final as Editor of The Florida
Anthropologist. I began my tenure as Editor in December,
1999 just as I was finishing fieldwork for the state-sponsored
investigation of the Miami Circle. Serving as editor has been
a lot of work but also a great honor; I have enjoyed working
with the FAS board members, contributors and reviewers, as
well as the kind folks at Durra-Print and Capital City Mailing.
Keith Ashley and Dan Hughes were great in their role as Book
Review Editor and my friend and colleague George Luer was
always ready to lend a hand with editing and general encourage-
ment. Members of the Florida Anthropological Society should
be very proud of our journal-it would not be possible without
their continued support. I am proud that I can add my name to
the list of editors and I am also a little sad to give up my role as
Editor. It is, however, with great pleasure that I introduce
Deborah Mullins and Andrea White who have volunteered to
serve as Co-Editors of The Florida Anthropologist.
With family roots deep inthe southeast, Deborah R. Mullins
is a proud Cracker from Deland, Florida who began her training
in archaeology over a decade ago by contacting (i.e.: politely
harassing) long time FAS member and avocational archaeolo-
gist Dorothy Moore for volunteer opportunities in Deborah's
home county of Volusia. One dirt parking lot, one series of
shovel tests, and one scattered shell midden later, and she was
hooked. Deborah took her B.A. in Anthropology at the
University of Florida and her M.A. in Historical Archaeology
VOL. 59(3-4) THE FLORIDA ANThROPOLoGIST SEPTEMBER-DECEMBER 2096
VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
SEPTEMBER-DECEMBER 2006
TEE FLORIDA ANTHROPOLOGIST 2006 VOL.59(34)
from the University of West Florida. Currently conducting
doctoral research at the University of Florida with an emphasis
in historic sites archaeology, Deborah's academic research has
lead her (thus far) to Veracruz, Mexico, San Juan, Puerto Rico,
and Pensacola and St. Augustine, Florida. She has also had the
great pleasure in participating on numerous prehistoric and
colonial era archaeological research programs across Florida as
well as in Haiti, Puerto Rico, Turks and Caicos, and the
Bahamas. Throughout her tenure in archaeology, Deborah has
learned from and worked alongside several outstanding
professionals and avocationalists working across the state of
Florida in our communities and cultural management programs.
Andrea P. White comes to the Florida Anthropologist with
a passion for Florida archaeology. Knowing she wanted to
pursue a career in archaeology, Andrea career shadowed with
Bonnie McEwan at Mission San Luis while still in high school.
After washing countless artifacts she was still enthusiastic and
crazy enough to pursue a B.A. in Anthropology from the
University of Florida. In quest of additional historical archaeo-
logical training, Andrea took her M. A. in Anthropology from
the College of William and Mary in Williamsburg, VA. After
constantly correcting her historical archaeology professors that
it was Florida, not Virginia, who has the oldest colonial
archaeology in the U.S., she realized she had to return to her
home state. With over a decade of experience in all facets of
Florida and Southeastern archaeology, she has participated in
numerous projects on prehistoric, historical, and submerged
archaeological sites, especially in St. Augustine where she was
assistant City Archaeologist. Currently she is the Training and
Outreach Coordinator for the Florida Bureau ofArchaeological
Research. As an archaeologist, she has been honored and
delighted to learn from various knowledgeable professionals
and FAS members committed to Florida Archaeology.
It is with greatpleasure that Deborah andAndrea commence
serving The Florida Anthropologist as Co-Editors. Please
submit inquiries and manuscripts to:
Deborah Mullins
The Florida Anthropologist, Co-Editor
P.O. Box 357605
Gainesville, FL 32635-7605
E-mail: dmullins.fl.anthropologist@gmail.com
This issue also includes a summary of the 2006 FAS
Awards; an obituary on Dan Laxson, long-time FAS member
and contributor to The Florida Anthropologist; and three book
reviews byJon Endonino, Christian Wells, and Uzi Baram. We
hope that all readers find something of interest!
RYAN J. WHEELER
December, 2006
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
ANALYSIS OF CERAMICS FROM BRICKELL POINT, 8DA12
ROBERT S. CARR
Archaeological and Historical Conservancy, Inc., 4800 SW 64†Avenue, Suite 107, Davie, FL 33314
E-mail: archlgcl@bellsouth. net
Introduction
In this paper an analysis of the ceramics recovered from the
Brickell Point component of Miami Midden No. 2 (8DA12) is
conducted. The principal goals of this analysis were to identify
the ceramic types associated with the site, to compare the
Brickell Point ceramic assemblage with other sites in the area,
and to identify non-local wares that indicate possible exchange
across the region. Another goal was to use the ceramic assem-
blage to provide relative dating for deposits associated with the
Miami Circle and across the site and also to compare the
assemblage to other sites in the area. Ceramics were sorted by
type but no attempt was made to conduct a chemical, PIXE, or
microscopic analysis of paste.
Research Design
One of the important questions about the feature known as
the Miami Circle, which most archaeologists agree is the
footprint of a circular structure, is its age. The feature is
characterized by 24 basins cut into the limestone bedrock. Each
basin has multiple circular postholes cut into the rock at the
basin bottom. There is no direct way to date the feature since
it cannot be assumed that the midden mantle above the Circle
is necessarily younger. In theory, it is possible that when the
structure was built that existing sediments could have been dug
up and the basins and holes then cut into the exposed bedrock.
If that were the case, one might anticipate some type of mixture
of cultural materials within the basins and their vicinity.
A review of the field records and an analysis of associated
ceramic types might indicate whether reverse stratification
occurred in the basin areas. In contrast, the hypothesis that the
Circle’s creation predates the deposition of all or most of its
overlying sediments would support the likelihood that ceramic
types occur in a “correct†sequence (with Glades I types below
and Glades I] and Glades HI types above) or that all the ceramic
types within the basins would reflect a limited temporal range.
Of course, even if the latter hypothesis is correct, habitation
sites are subject to numerous disturbances from daily prehis-
toric activities (such as from digging pits and bioturbation) and
particularly in this case, the construction associated with over
700 hundred postholes uncovered in Block 1.
Complicating this review is the partial removal of, and
disturbances to, the soil mantle in Block 1. This occurred
during the construction of the Brickell Apartments in 1949, and
the subsequent mechanical removal of much of the sediments
above the basins by archaeologists with the Miami-Dade
County Historic Preservation Division (MDC) and the Archaeo-
VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
logical and Historical Conservancy (AHC). Archaeologists
were under pressure to finish uncovering the circle feature when
the developer gave them only two weeks to complete the
excavations in August 1998.
Any attempt to associate the Circle’s ceramic assemblage
with the age of its construction and use is fraught with chal-
lenges because of probable disturbances to the sediments during
the Circle’s construction, not to mention during and after its
use. However, the one component that is the least likely to be
disturbed are those sediments trapped within the individual
basins that characterize the Circle’s circumference. These
basins vary from 56 to 74 cm in length, and are cut into the
bedrock 30 to 45 cm below the bedrock surface. Serving as
footer trenches, the basins have smaller circular cut holes at the
basin bottom, suggesting that each basin contained numerous
wooden posts. Helping to support these posts were numerous
limestone rocks pushed and wedged into the basins, many of
which are cemented together and concreted from ground water.
In addition, shell refuse, faunal bones, and pottery occur in the
sediments within the basins, and in some instances artifacts and
other cultural material are also cemented onto the limestone
bedrock or individual rocks. The general appearance of the |
cemented materials and the duracrust within the basins suggests
considerable antiquity, although no method for accurately
dating the concretion has yet been identified, in part, because
the concretion represents secondary material redeposited from
unknown sources of unknown age.
It is hypothesized that the basin sediments and associated
cultural material are largely coeval with the Circle’s construc-
tion. The basins provide a protective environment for any
materials deposited within them, and particularly from bioturba-
tion as the basins are horizontally sealed by their construction
into the limestone bedrock. However, any post-Circle holes
dug directly on top of a basin could introduce new materials
and/or remove associated materials.
It is hypothesized that sediments within the basins largely
contain materials including pottery of a similar age as the
Circle, which is hypothesized to be approximately 2000 years
old. In contrast, it is hypothesized that sediments above the
basins and in other parts of Block 1 contain a wider variety of
ceramic types reflecting more recent use of the site.
Contexts
This assessment reports the ceramics recovered from three
different archaeological investigations of the Miami Circle
parcel (Figure 1). A total of 26,281 prehistoric pottery sherds
are récorded for the site, of which 23,326 are sand-tempered
SEPTEMBER-DECEMBER 2006
THE FLORIDA ANTHROPOLOGIST
Figure 1. Brickell Point (8DA12), showing areas of major investigations.
2006 VOL. 59(3-4)
CARR BRICKEU. POUIT CERAMIC AI'JALYSIS
Figure 2. Brickell Point (8DA12), showing MDC test units.
CARR
BRICKELL POINT CERAMIC ANALYSIS
THE FLOlUDA ANTHROPOLOGIST 2006 VOL 59(34)
plain (STP). The first investigation was conducted jointly by
the Miami-Dade CountyHistoric Preservation Division (MDC)
and the Archaeological & Historical Conservancy (AHC) from
August 1998 to January 1999. The second investigation was a
systematic phase 1 assessment conducted across the parcel by
Florida's Bureau of Archaeological Research (FBAR) from
October to November 1999. The third investigation was
conducted by the University of Houston during June and
August 2000 (Widmer 2004). The University of Houston
projects were funded by the Archaeological & Historical
Conservancy.
MDC/AHC
This initial investigation was the most intensive of the three,
resulting in the excavation of about 175 square meters of
sediments (Figure 2). Excavations were conducted in only
three areas, Blocks 1, 2, and 3. Block 1 encompasses the
Miami Circle feature. Block 2 is located about 10 meters
northwest of Block 1 on top of the historic bank of the Miami
River, an area now covered with fill and about 10 meters south
of the modem riverbank. Block 3 comprises a single unit dug
to the west of Blocks 1 and 2 that was never expanded.
Excavations were confined to Blocks 1 and 2 by order of the
developer, who did not want any expanded investigations that
could lead to additional discoveries, which could further
complicate the proposed development.
Excavations directed by the Miami-Dade County Archaeol-
ogist yielded 21,275 pottery sherds, of which 170 were deco-
rated. Pottery was always collected during the excavations,
including sherds found on the surface, and during monitoring of
the removal of the foundations of the Brickell Apartments.
FBAR
documented over 500 cut post holes.
Of the 1,688 sherds recovered from the Operation 3
excavation, 16 are described as temporally diagnostic by
Widmer (2004:30). Decorated sherds all date from Late Glades
I through Glades II Early periods and include the following:
four Fort Drum Punctate, six Opa Locka Incised, and one
Gordons Pass Incised. The four Fort Drum Punctate sherds are
associated with Features 87 and 93, and the Opa Locka Incised
sherd with Feature 4.
Widmer's work was pivotal in determining the age of the
site. With a total of 19 radiocarbon dates secured from Lucina
shell, most from within cut postholes, a range of from 760 B.C.
to A.D. 200 was determined for the occupation of the structures
from that part of the site. Widmer hypothesized that the overall
site, and the Circle in particular, were a similar age (Widmer
2004).
Field Collections Methodology
Although field excavation techniques varied between the
three investigations conducted at Brickell Point, collecting
strategies were similar for all. Investigations by MDC and
AHC focused on two blocks. Excavations were conducted in
5 foot squares and collections also were made from demolition
trenches and spoil piles across the parcel. All unit excavations
were subject to wet-screening through 1/4 and 1/8 inch mesh.
Instructions to excavators were to collect all ceramic sherds.
Thus, it is likely that only a small percentage of pottery was
missed. A similar comprehensive recovery of ceramic sherds
was carried out during the FBAR and Operation 3 phases.
Local Ceramic Types
Fort Drum Incised
Phase I testing across the parcel was conducted by FBAR
using shovel and auger testing augmented by some block
excavations between October 20 through November 24, 1999
(Wheeler 2000). The FBAR investigation is importantbecause,
to date, it is the only systematic survey of the site, thus provid-
ing a generally representative sample of materials from across
the site.
A total of 3,318 ceramic sherds was recovered, of which 25
were decorated. Ceramic types were diverse, ranging in
chronology from the Late Archaic to the Glades II periods.
Decorated types include Fort Drum, Opa Locka Incised, St.
Johns simple stamped, and unclassified incised types. Non-
local wares include sherds tempered with shell, mica, and grog.
Operation 3
Investigations by Randolph Widmer of the University of
Houston focused on one block measuring 8 by 8 meters located
in the northeast corer of the parcel about 30 meters northeast
of the Circle (Figure 1). Here, fill covered a thin mantle of
midden soils. Upper levels of sediment were highly disturbed
or had been removed, probably during the construction of the
Brickell Apartments in 1949. The Operation 3 excavations
This type is characterized by vertical or diagonal incisions
on and just below the lip. The incisions vary from gracile, light
incisions to bold, deep notches on the lip (sometimes errone-
ously classified as Glades Tooled). Other variants include
incisions on the rim only. Goggin (1944) was first to classify
this type. This design appears on sand- tempered plain wares
and was assigned by Goggin to the Late Glades I Period (1944).
Griffin notes that the use of the type name, Fort Drum Ticked
Rim, by some scholars is probably a subset of Fort Drum
Incised (Griffin 2002:84).
A total of seven Fort Drum Incised sherds was recovered
from the three work phases at the site (see Figures 3 and 4).
One sherd is from the FBAR survey and six are from the MDC
excavations. During the MDC investigation two sherds were
found in Block 1 and one from Block 2. At least one of the
Circle specimens (FS 559) was uncovered from what appeared
to be an undisturbed context in Unit 36 from Level 2 at 6.12
feet above the National Geodetic Vertical Datum (NGVD) of
1929. The others were found in a spoil pile.
Fort Drum Punctate
Fort Drum Punctate was recovered from all Brickell Point
THE FLomIA ANTHROPOLOGIST
2006 VoL. 59(3-4)
CARR
BRICKELL POINT CERAMIC ANALYSIS
137
325 330 335
1507 a r —
3 | |
OFS ersce |6=— gt, 7 mu
4 66 44 45
-. @FS579 Se i
(@FS55 6 34s! ff NN 5
1 | 36 48 47 49 50 Ngo
gy A Fs 559 © FS 583 - on 435
@ FS 162 FS 547
107, 14 | 62 63 61
| *
t— y — — — 130
24 82 21 0
= b 125
¢ 120
115
110
105
295 300 305 310 315 320 325
A Fort Drum Incised ee | aaa
@ Fort Drum Punctate 0 8 10 feet
Figure 3. Distribution of Fort Drum decorated pottery at the Miami Circle.
work phases. A total of 21 sherds was recovered (see Figures
3 and 4). Four sherds (representing two vessels) were recov-
ered from the Operation 3 work area east of the Circle, and
three from the FBAR survey. Fifteen sherds were recovered
from the MDC excavations, of which ten were recovered from
within or adjacent to the Circle in Block | and three from Block
2. Most of these sherds occurred in Levels 1 and 2 in Block 1
close to the fill interface where no clearly undisturbed contexts
were identified. Likewise, Operation 3 sherds are all from
Level 1, except for one (OP3.1.709.6) which was recovered
from Feature 87.
Opa Locka Incised
This type is characterized by small downward opening arcs
made with fingernail-like incisions below the vessel lip.
Incisions are usually bold and well defined, sometimes occur-
ring in vertical rows or clusters below the lip and on the rim.
138 THE FLORA ANTHROPOLOGIST
2006 VoL. 59(3-4)
Figure 4. Fort Drum ceramic sherds: a) MDC.1.162.8; b)
MDC.1.36.2; c) MDC.1.1000.3; d-e) MDC.1.36.2;
MDC.1.638.9.
Goggin (1944) was first to characterize this type, and
subsequent descriptions were offered during his Matecumbe
Key investigations (Goggin and Sommer 1949:40). Goggin
noted this type’s similarity to Key Largo Incised, but the exact
nature of their relationship is unclear (Goggin n.d.). Tempo-
rally Opa Locka Incised precedes but also overlaps with Key
Largo Incised (Griffin 2002:83).
Opa Locka Incised is the most common decorated type (63
sherds) at Brickell Point, representing 30% of the decorated
sherds (see Figures 5 and 6). The Brickell Point specimens are
characterized with variations from two to thirteen incisions
occurring in columns on the rim below the lip. Lips are both
rounded and flattened.
Based on a review of the field notes, eight specimens were
recovered from within the Block 1 midden, all from levels 1
and 2; both of these levels were probably disturbed during the
construction of the Brickell Apartments. One specimen (FS
259) was recovered from a deeper level (Level 3).
Opa Locka Incised was more common in Block 2, nearer
the river. It produced 12 specimens, which were recovered
within a 40 square foot area. At least seven of these sherds
were in undisturbed contexts in Levels 6 to 8.
Recovery during systematic testing across the site by FBAR
produced eleven specimens. Widmer’s Operation 3 yielded six
specimens, five of which were recovered from Feature 4.
Key Largo Incised
This type is assigned to the Glades JIA and IIB periods
(Goggin 1944; Griffin 2002). Key Largo Incised is character-
ized by downward opening incised arcs or loops on the rim and
shoulder below the lip. The arcs are generally bold incisions,
often overlapping with the adjacent arcs and continuing around
the circumference of the pot. The design is a single series of
loops around the rim, but multiple series of loops are not
uncommon. Lips are rounded and some rare variants occur
with incisions on the lip (Goggin 1950b:240). Vessels are
generally medium-sized with slight to medium orifices (Griffin
2002).
A total of 22 Key Largo Incised sherds was recovered from
the Brickell Point excavations, all from the MDC phase (see
Figures 7 and 8). Of these, six were recovered from within or
adjacent to the Circle in Block 1, and fourteen from Block 2.
An additional two specimens were collected from spoil. Of the
five specimens collected from the Circle Feature in Block 1,
only one (MDC.1.500) is clearly assigned to an undisturbed
context of Level 4. Within Block 2 all but two of the 14 sherds
uncovered were from undisturbed contexts.
Dade Incised
This type is characterized by an incised design of upward
opening arcs located below the lip. Sometimes the pattern
occurs as a singular element but, more often, it is a series of
three elements. Sometimes the arcs cross the neighboring arc
or sometimes they are separated by a space. Goggin defined
this type (Goggin and Sommer 1949:35-36), and Griffin notes
that it is fairly rare (2002:82) and that the type dates from the
Glades II Period.
Only five examples were found at Brickell Point: four by
MDC and one by FBAR (Figure 9). The MDC examples
include three from Block 1, of which only one (a possible
Miami-Dade Incised variant) was found in Unit 36, Level 1, in
a probable disturbed context. The other two sherds were
recovered from spoil piles. Block 2 yielded a single specimen
(MDC. 1.108) from Level 7 in an undisturbed context.
Miami Incised
This type is represented by a series of parallel lines incised
diagonally on the rim. The lines occur in groups of two to six
incisions. Goggin defined this type (1944:6) and dates it from
the Glades II Period; Griffin provides a summary of this type’s
attributes (2002:82). A total of 17 sherds was recovered from
the project parcel (Figures 9 and 10). Of these, 16 were from
the MDC group, seven from Block 1 and nine from Block 2.
Four of the Block | specimens were from units, and two from
spoil from the trench. All of the unit sherds were from Levels
1 and 2 with at least two of the Block | sherds recovered from
disturbed contexts.
Sanibel Incised/Gordons Pass Incised
The Sanibel Incised type has a series of incised ticks that
CARR Biuci~u. POINT CERAMIC ANALYSIS
* FS 97
145
140
135
62
24
48 47 49
63
23
34 41
eFS
57
52 28
61 64
FS 527
19 22
42
32 21
--1---. 25
53 56
120
FS 541
115
N
300 305
310 315 320
* Opa-Locka Incised
325
0 5 10 feet
Figure 5. Distribution of Opa Locka Incised pottery at the Miami Circle.
form diagonal lines or a running "V" design on the rim or
shoulder. Goggin (1944) was first to classify this type and
noted its similarity and probable relationship to Gordons Pass
Incised, a type with similar ticks but also accompanied by a
parallel incised line creating a feathered appearance (Goggin
1944, 1950a:75); Griffin also describes Sanibel Incised
(2002:84).
The author's observations indicate that this type is uncom-
mon in southeastern Florida although it is a common marker in
southwestern Florida, specifically in the Ten Thousand Islands
and Big Cypress areas. Only one specimen (MDC.1.108.12)
was found in the Brickell Point investigations, from Block 2,
Unit 6, Level 7; this level has probably been disturbed (Figure
1 c). A possible Gordons Pass Incised (FS 737.11) sherd was
found during the Operation 3 work.
Matecumbe Incised
Characterized by intersecting, deep, diagonal incisions or
cross hatchings across the bowl exterior, this type was defined
44 45
CARR
BRICKELL POINT CERAMIC ANALYSIS
TIlE FLORIDA ANTHROPOLOGIST 2006 VOL. 59(3-4)
Glades Noded
Figure 6. Opa Locka Incised ceramic sherds: a) MDC.1.100
MDC.1.159.3; c) MDC.1.108.15; d) MDC.1.259.4.
by Goggin (1944) during his Upper Matecumbe investigations
(Goggin and Sommer 1949). Other investigators have also
illustrated this type (Willey 1949:Plates 3, 14; Griffin
2002:181, Figure 4 a-c).
Only one definite specimen of Matecumbe Incised was
found during these investigations (Figure 11 a). That specimen
(MDC. 1.122.6) was found in Block 2, Unit 7, Level 8. A
second possible specimen (MDC. 1.822.3) was recovered from
Feature 577, Unit 47.
Glades Tooled
This is a distinctive type characterized by tooling or
grooving on a thickened lip. The vessel shapes (as noted by
Goggin) include two types, one a hemispherical bowl with a
slightly in-facing rim, and a second vessel type which has a
constricted mouth with a grooved or incised lip (Goggin
1950b:235; Goggin and Sommer 1949:Plate 1; Griffin2002:77-
78, Figure 4:2).
Only one specimen (MDC.1.214.13) was recovered from
the project parcel, specifically from Block 2, Unit 11, Level 7
(Figure 1 Id). It is worth noting that this level is just below the
Brickell era (ca. 1910) road fill, and that Level 7 soils had
nineteenth century artifacts mixed with prehistoric materials,
and that at lower depths, no historic materials were encoun-
tered. Two "early" or crimped varieties of rim tooling
(MDC.1.157.8.11 and MDC.1.182.10) also were uncovered,
but these are not the classic Glades Tooled as commonly
understood.
The author has observed several specimens of this
type in southern Florida during the past thirty years.
Griffin notes its occurrence at the Granada site.
Glades Noded is characterized by embossments
below the exterior lip. The nodes are created by a
reed or tubular bone being pushed from the bowl
interior toward the exterior just below the lip. A
single specimen (MDC. 1.108.13) was found in Block
2, Unit 6, Level 7, a disturbed context with a mixture
of Glades II and III ceramic types (Figure 1 Ib).
Ceramic Pipes
Of particular interest are the ceramic smoking
pipe fragments from the project parcel (Figures 12
and 13). Six pieces from three different pipes were
found-all from the MDC phase. Four are bowl
fragments and two are stem sherds. Three of the
pieces (MDC.1.450.1) are from a single bowl; two
sherds fit together. This group was found in the spoil
from the trench dug on the northwest circumference
of the Circle, thus no exact depth or context is known,
other than it being the midden context above the
0.5; b) bedrock. The pipe is made with fine-grained, sand-
tempered paste and may be of local manufacture.
Another bowl fragment (MDC.1.114.14) is made
from a coarse sand-tempered paste, but the paste colors (a buff
exterior and grey interior) are suggestive of an intermediate
chalkyware. Overall, the paste is Belle Glade-like in its
appearance. The fragment was found in Block 2, Unit 7, Level
5, which appears to be a disturbed context.
Specimen MDC. 1.68.7 is unusual in that it represents two
pieces that fit together, with a form suggesting a hole for
inhalation on the side or top of a rounded tube that opens to an
adjacent bowl. A small piece of carbonized residue adheres to
the interior bowl. The pipe is tan with coarse sand tempering
and was found in Block 1, Unit 4, Level 1, at a location about
five to seven feet northwest of the circle circumference.
Specimen MDC. 1.989.6 is a bowl fragment measuring 2.8 by
4.8 cm and was found on the surface, probably in the vicinity of
the Circle. It may be of importance, or simply a coincidence,
that all of the provenienced pipe fragments occurs within a 10
meter square area on the northwest side of the Circle.
Non-Local Ceramics
A small but interesting assemblage of non-local sherds was
recovered. These wares are plain and decorated wares with
tempering not typical of South Florida "sand tempering." Non-
local tempers observed in the collection include grog, shell,
limestone, and quartzite-all of which are tempering materials
rarely observed at South Florida sites. Other clay types include
micaceous clay and sponge spicule clays-materials not
documented as being local to South Florida. Other attributes
reflecting ceramic diversity and non-local wares, such as
chemical elements, etc., probably occur, but that level of
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
CARR B1UCKELL POINT CERAMIC ANALYSIS
4
SFS 55
1
40
49
62 63 61 64
24 23 19 22
34 41
18 39 31 3
0 FS 615
57 27
52 28
w^
300 305
O Key Largo Incised
0 FS 4444
2 21
OFS 500
53 56 .
145
140
135
130
125
42 60
J~7i
4';
lo S
120
115
N
325
0 5 10 feet
Figure 7. Distribution of Key Largo Incised pottery at the Miami Circle.
analysis is beyond the scope of this paper.
Deptford Series
This series is described by Willey based on its common
occurrence in northwest Florida and the Gulf Coast (Milanich
1994; Willey 1949). The types have been defined with exten-
sive analysis (Bense 1985; Caldwell and Waring 1939a, 1939b;
DePratter 1979). Deptford pottery is characterized as sand-
tempered, often with mica inclusions, which is how the type
was originally defined in Gulf Coast Florida (Milanich 1994;
Willey 1949). Three types are noted: Deptford Linear Check
Stamped, Deptford Simple Stamped, and Deptford Check
Stamped. The series dates from ca. 500 B.C. to A.D. 100,
placing it in the Early Woodland Period and its southern Florida
correlate, the Glades I Period.
Deptford pottery is rare in southeast Florida and occurs in
low frequency at several sites. Specimens were observed by the
author while monitoring the construction of the Holiday Inn at
Brickell Park in 1981 and more recently during Icon-Brickell
CARR
BMiCKELL POINT CERAMIC ANALYSIS
-
29 j
THE FLORIDA ANTHROPOLOGIST 2006 VOL.59(3-4)
Figure 8. Key Largo Incised ceramic sherds: a) MDC.1.114.9; b) MDC.1.500.9; c) MDC.1.239.26.
investigations within Brickell Park (report in progress).
At Brickell Point, a total of thirteen specimens were
identified from the three field operations, twelve of which were
uncovered by MDC and a single specimen recovered by FBAR
(FBAR.1.14.2) (Table 1 and Figures 14, 15 and 16). One
Deptford Simple Stamped sherd (FBAR.1.14.2) was found
during the FBAR assessment in Unit 1, Trench 11. Three
Deptford Linear Check Stamped sherds were documented: one
was found in Block 1, Unit 13 (MDC.1.228.9), another
(MDC. 1.628.6) from Unit 62, and another during monitoring of
demolition (MDC.1.28.1). Two sherds of Deptford Check
Stamped were found in a Circle basin feature. MDC.1.903.8
and MDC.1.903.10 were found in Feature 484, Unit 45.
Specimen MDC. 1.228.9 was uncovered from Unit 13, Level 7.
Specimen MDC. 1.60.1 was recovered from Unit 1, Level 4,just
above the bedrock. This sherd had been misidentified as St.
Johns Check Stamped in the field notes. Of particular interest
was the uncovering of six associated sherds (MDC. 1.427) from
Unit 25 lying directly atop the bedrock (Figure 15).
St. Johns Series
This type is a soft chalky ware first described by Goggin in
1940 as the Biscayne Chalky series, which he later described in
detail in his manuscript, with the changed name of St. Johns
(Goggin n.d.:449; Willey 1949:408). Griffin provides a
summary of several of the controversies associated with this
type. Aside from noting Goggin's change in his nomenclature,
he also notes that Goggin believed the type was not locally
made (Griffin 2002:87-89). A master's thesis by McGregor
(1974) on a study of ceramic chronology adds to the confusion
surrounding the type's name by using the discarded Biscayne
series nomenclature.
The debate as to whether the St. Johns Series represents a
local versus non-local ware was again fueled by Espenshade
who argues that it is a local "muckware" (Espenshade 1983).
Griffin voices his doubts about that hypothesis based, in part,
on the fact that the decorations that occur on St. Johns ware do
not occur on the sand-tempered local Glades wares. Griffin
believes that a crossover would be expected if both types were
local and from the same cultural group (Griffin 2002:88-89).
The author's observations, based on his experiences in
South Florida beginning in 1980, is that the St. Johns Series has
two temporal occurrences in southeast Florida. The first occurs
during the early Glades I Period (ca. 750 B.C. A.D. 500),
when St. Johns Plain occurs in association with undecorated
sand-tempered plain, and semi-fiber- tempered pottery, such as
at the Atlantis Site, DA1082, and the Santa Maria site,
8DA2132 (Carr 1981). The second temporal occurrence of St.
Johns ware, characterized by plain ware and St. Johns Check
Stamped, begins ca. A.D. 1200 and continues through European
contact, ca. A.D. 1700.
More recently as a result of the author's investigations at
Brickell Point and Dupont Plaza on the Miami River and Hyde
Park on the New River, it is apparent that the first temporal
occurrence of St. Johns ware includes decorated types in the
Deptford tradition not previously recognized in southeast
Florida. Three types of stamped St. Johns ware were identified
in the Brickell Point collection: simple stamped, linear check
stamped and check stamped. The simple stamped and linear
check stamped are similar in design to the Deptford types.
The Deptford Series is described by Milanich and Fairbanks
(1980:77-83) with regional variations of clay and temper forthe
Gulf subregion and the Atlantic Coast. They describe large
2006 VOL. 59(3-4)
TnE FUoRIiDA ANTiimOPOLOGIST
CARR BRICEELL POINT CERAMIC ANALYSIS
4 66
40
18 39
57
52
295 300 305
* Miami Incised
36
A FS 558
62
9 FS 629
44
48 47 49 50 %9
63 61 64 3 6
24 23 19
34 41 42
3FS 604-
31 3 4
27
22
60
32 21
53 56
FS 611
i N
296
N
310 315 320 325
0 5 10 feet
A Dade Incised
Figure 9. Distribution of Miami Incised and Dade Incised pottery at the Miami Circle.
quartz inclusions for the Atlantic coast, smaller quartz inclu-
sions for the peninsular Gulf, and the absence of mica inclu-
sions in the peninsular Deptford, while sherds occurring on the
Atlantic coast and in Northwest Florida tend to have mica in the
paste. The Brickell Point examples are similar in design but are
applied on chalkyware vessels, most with sponge spicules. For
the purposes of this article I will refer to these forms as St.
Johns simple stamped and St. Johns linear check stamped.
These Deptford-influenced types of St. Johns ware were
first recognized by John Goggin. Goggin (1952:47) notes the
occurrence of "local incised and stamped forms" in northeast
Florida during the St. Johns I Period but adds that they are
relatively rare. He describes their occurrence during the St.
Johns Ia Early Period as follows:
CARR
BRICKELL POINT CERAMIC ANALYSIS
145
140
135
130
125
120
115
110
TEE FLORIDA ANTHROPOLOGIST 2006 VOL. 59(34)
Figure 10. Miami Incised ceramic sherds: a) MDC.1.629.5; b) MDC.1.551.6; c) MDC.1.449.12.
Figure 11. Variety of decorated Glades ceramic types: a) Matecumbe Incised, MDC.1.122.6; b) Glades Noded, MDC.1.108.13;
c) Sanibel Incised, MDC.1.108.12; d) Glades Tooled, MDC.1.214.13.
Trade wares include Deptford Linear Check Stamped,
Deptford Bold Check Stamped, Deptford Simple Stamped,
and an associated cord-marked type. Tetrapodal base forms
of the two former types have been found. More significant
perhaps are the local attempts at copying these Deptford
forms. At Tick Island (both midden and mound) there have
been found chalky ware sherds with simple stamping, linear
check stamping, and cord marking. These forms are very rare,
only 34 sherds occurring among the 661 sherds and vessels
from the site, and they appear at few other sites. For this
reason it may be best not to give them type names at the
present time. [Goggin 1952:49]
A total of 481 St. Johns sherds was identified from the
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
CARR BRICKELL POINT CERAMIC ANALYSIS
Figures 12 and 13. Top: Ceramic pipes, obverse views: a-b) MDC.1.450.1; c) MDC.1.114.14; d) MDC.1.68.7; Bottom:
Ceramic pipes, reverse views: a-b) MDC.1.450.1; c) MDC.1.114.14; d) MDC.1.68.7.
Brickell Point site (Table 2 and Figures 17, 18, 19 and 20).
This includes 427 St. Johns Plain sherds and 44 St. Johns
stamped sherds of which 30 are in the Deptford style. It is the
author's opinion, based on the presence of these Deptford-style
stamped decorations on St. Johns ware, that it is important to
describe these sherds in more detail, but not as formal types.
Goggin (1952:99) recognized a number of early St. Johns
Series forms that had not been given formal names, including
CARR
BRICKELL POINT CERAMIC ANALYSIS
TRE FLORIDA MffHROPOLOGIST 2006 VOL. 59(34)
325 330 335
150 0- 1 4
o FS 104
5
FS 903.8
145
SFS 74
1 36 48 47 49 50
A FS 60
--- ------ 135
N FS 628
10 14 62 63 61 64 6
130
20 24 23 !9 22 2 21
-125
40 34 4.1 42 60 53 56 I
FS 363 120
18 39 31 31
115
2 FS 4277
57 27 3 29^
110
52 28
105 N
295 300 305 310 315 320 325
A Deptford Check Stamped N Deptford Linear Check 0 5 10 feet
Stamped
Figure 14. Distribution of Deptford Series pottery at the Miami Circle.
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
CABJI BIUCKEU. POINT CERAMIC ANALYSIS
Figure 15. Deptford Linear Check Stamped pottery, MDC.1.427, in situ on top of bedrock in Unit 25. Photograph by John
Beriault.
Figure 16. Deptford Linear Check Stamped ceramics: a) MDC.1.903.8; b) MDC.1.28.1; c) MDC.1.60.1; d) MDC.1.628.6.
CARR
BRICKELL POINT CERAMIC ANALYSIS
THE FLORIDA ANTHROPOLOGIST 2006 VOL 59(34)
Table 1. Deptford Series Pottery.
Field Specimen Provenience Deptford Deptford Linear Deptford Total
Simple Check Stamped Check
___________ _____Stamped Stamped___
MDC 28.1 West of Trench 1 1
1
MDC 228.9 Unit 13 1 1
MDC 628.6 Unit 62 1 1
Level 1
MDC 60.1 Unit 1 1
Level 4
MDC 427 Unit 25 6 6_____
MDC 903.8 Unit 45 1 1
Feature 484
MDC 903.10 Unit 45 1 1
Feature 484
FBAR 14.2 ---- 1 1____
Totals -- 1 9 3 13
Table 2. St. Johns Series Decorated Pottery.
Field Specimen Provenience St Johns St. Johns linear St Johns Total
simple check stamped Check
stamped Stamped
MDC 26.7 Trench 7 1 1
MDC 53.2 Spoil 1 1
MDC 65 Surface I 1
MDC 74 Unit 4 1 1
MDC 80.10 Unit 4 2 2
MDC 90 Unit 3 1 1
MDC 91.7 Unit 3 1 1
MDC 95.6 Surface 1 1
FBAR 96 ---- 3 3
MDC 98.6 Unit 5 1 1
MDC 99 Unit 5 1 1 2
MDC 101.2 Surface 1 1
MDC 117.15 Unit 6 1 1
MDC 122.2 Unit 7 1 1
MDC 134 Unit 7 1 1
MDC 141 ----
MDC 192 Unit 11 2 2
MDC 204.2 Unit 12/Fea. 20 1 1
MDC 313 Feature 77 2 2
MDC 315 Feature 79 1 1
MDC 328 Feature 96 1 1
MDC 332 Feature 101 4 4
MDC 372 Feature 95 1 1
MDC 564 Unit 36 2 2
MDC 575 Unit 42 1 1
MDC 586 Unit 48 4 4
MDC 632 Unit 64 1 1
MDC 985 NW Circle Spoil 1 1
MDC 1009 NE Circle Spoil 4 4
MDC 1014.5 Block 1 Spoil 2 2
Totals ---- 15 17 14 46
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
CARR BRICKELL PoINT CERAMIC ANALYSIS
150 AFS 96
62 63 61
24 23 19
SFS 575
42
34 41
18 39 31 3
57 9 27
52 28
300 305 310
* St. Johns Checked Stamped
Sherd
A St. Johns Simple Stamped
* St. Johns Linear Checked
Stamped
0 5 10 feet
V Dunns Creek Red
Figure 17. Distribution of St. Johns Series pottery at the Miami Circle.
fine simple-stamped and linear check-stamped, which dated to
a brief time in the St. Johns la Period (ca. A.D. 100-500).
St. Johns simple stamped is characterized by linear parallel
lands covering the vessel wall. The stamping appears to be
vertical or at a slight oblique angle to the rim. A total of 15 St.
Johns simple stamped sherds was recovered (Figures 17 and
19). Of these, 13 were from Block 1, seven of which were
recovered from midden spoil from the northeast and northwest
AFS 74
4
SFS 80
66
45 47
47 49 50
47 49 50
145
140
135
130
125
120
A FS 632
64
VFS 437
22
60
32 21
53 56 6
N
CARR
BRICKELL POINT CERAMIC ANALYSIS
TniE FLORIDA ANTHROPOLOGIST 2006 VOL 59(3-4)
Figure 18. St. Johns Check Stamped ceramics: a-b) MDC.1.80.10; c) MDC.1.122.2; d) MDC.1.91.7.
Figure 19. St. Johns simple stamped ceramics: a) MDC.1.99.7; b-d) MDC.1.1009.4.
TIIE FLORIIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
CARR BIUCKELL POINT CERAMIC ANALYSIS
Figure 20. St. Johns linear check stamped ceramics: a) MDC.1313.1; b) MDC.1372.8; c) MDC.1.586.9.
circumference of the Circle which were uncovered during the
removal of soil from above the basins. Four sherds
(MDC.1.332) were recovered from Basin Feature 101 and
another four sherds (MDC.1.372) from Basin 95. One sherd
(FS 632) was recovered from Unit 64, and three (FS 96) from
Unit 5. Another three sherds (FBAR.1.96.11) were found
during the FBAR survey.
St. Johns linear check stamped is characterized by a check-
stamped design similar to Deptford Linear Check Stamped,
with bold horizontal lands and thinner smaller perpendicular
lands creating a ladder-like effect. This design and other
stamped designs were created by a carved wooden paddle being
impressed against the vessel exterior. Many of the Brickell
Point examples are well executed with sharply defined lands
encompassing rectangular recessed areas (Figure 20). A total
of 15 St. Johns linear check stamped sherds was found. Twelve
were found in Block 1, three in Block 2, and one on the surface.
Four of the sherds were recovered from the Circle's cut basins,
specifically Feature 77 (MDC. 1.313), Feature 79 (MDC. 1.315),
and Feature 95 (MDC.1.372). Three sherds (MDC.1.564,
1.575, 1.586) were found in Units 36,42, and 48, respectively.
It should be noted that two other "check stamped" sherds were
listed in the field notes for Feature 77, but they were not found
during this assessment.
A total of 14 St. Johns Check Stamped sherds was found at
the site (Figures 17 and 18). All but one were found in the
MDC investigation, and one from the FBAR survey
(FBAR.1.96). Seven sherds were uncovered in Block 1: one
from Unit 5 (MDC.1.98.6), two from Unit 4 (MDC.1.80.10),
one problematic sherd (MDC. 1.204) from Unit 12, Feature 20
(which is a Circle basin), and three from spoil from the Block
1 area. Block 2 yielded two sherds, one from Unit 6
(MDC.1.117.15) and one from Unit 7 (MDC.1.122.2). The
other sherds were surface finds. Sherd MDC. 1.985, a tetrapod
foot uncovered in redeposited soil near the northwest quadrant
of the Circle, is of particular interest because of its similarity to
Deptford Check Stamped.
Dunns Creek Red
This type is characterized by a red slip on either or both the
interior and exterior on St. Johns paste. Griffin describes this
type as a subset of the St. Johns series, with no particular age
assignment (Griffin 2002:89-90). Two specimens were found;
both from the MDC operation (Figure 17). One specimen
(MDC.1.437.6) was found during the MDC investigation in
Block 1, Unit 22, Level 1. Red clay inclusions occur through-
out the sherd. This specimen has an orange surface, without
slip. (George Luer observed that it may be color derived
residue from the slip, which might have eroded away). The
second specimen (MDC. 1.47.3) was found during monitoring.
Gulf Check Stamped
Four sherds ofthis type (Willey 1949:387-388) were found;
all were from Block 1 (Figure 21). Specimen MDC.1.629.4 is
characterized by fine grain sand with mica inclusions and was
found in Unit 62, Level 2. Specimen MDC. 1.902.1 also was
recovered from Unit 62 atop the bedrock, adjacent to Feature
626. These two sherds appear to represent different vessels.
CARR
ftic(ELL POINT CERAMIC ANALYSIS
325 330 335
150-
5
A FS 104 FS 457
_--- 145
4 66 da 44 45
AFS 74
S_- -140
** FS 832 FS 594
1 36 48 47 49 50 9
*FS 59 N
----- 135
*FS 16Qp AFS629 *FS631
106 14 A^S 2 63 61 64 6
t| NFS 904
130
f FS 395
7 20 24 23 9 22 2 21
*FS 601 125
40 34 41 42 60 53 56
--- 120
18 39 31 3
115
57 27 29
110
52 28
105 N
295 300 305 310 315 320 325
A Gulf Check Stamped N Unclassified Zoned
Punctate/Incised 0 5 10 feet
0 Cord Marked
Figure 21. Distribution of miscellaneous, non-local pottery at the Miami Circle.
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
CARR B1ucIu~LL PoINT CERAMIC ANALYSIS
Figure 22. Miscellaneous non-local, decorated ceramics: a) MDC.1.594.4; b) MDC.1.216.11; c) MDC.1.631.9; d)
MDC.1.52.7.
Figure 23. Miscellaneous non-local, zoned punctated ceramics: a) FBAR.1.11.4; b) MDC.1.904.1; c) MDC.1.220.3; d)
FBAR.1.201.5; e) FBAR.1.63.7; f) MDC.1.8063; g) MDC.1.440.2; h) MDC.1.166.16.
~Khbl
r.m
CARR
BRICKELL POINT CERAMIC ANALYSIS
Table 3. Comparison of Ceramics from the Granada Site (8DA11) and Brickell Point (8DA12).
Granada Site Brickell Point Brickell Point
Type (DAll) Block 1 (DA12) Other (DA12)
sand-tempered plain 7,472 23,326
Glades Tooled 1,147 0 1
Surfside Incised 34 0 0
St. Johns Check Stamped 1,723 7 7
St. Johns Plain 563 108+ 427+
St. Johns linear check stamped 0 17 3
St. Johns simple stamped 0 13 2
Key Largo Incised 650 6 16
Dade Incised 64 3 2
Miami Incised 182 7 10
Opa Locka Incised 183 8 55
Matecumbe Incised 81 1 1
Ft. Drum Punctate 82 10 11
Ft. Drum Incised 18 2 5
Glades Embossed/Noded 7 0 1
Sarasota Incised 1 1 1
Deptford Series 0 5 1
Sanibel Incised 0 0 1
Dunn Creek Red 0 2 0
Gulf Check Stamped 0 4 0
Pinellas Incised 0 1 0
Cord Marked 0 8 0
Belle Glade Plain 412 0 1
*from Griffin et al. 1983:Table 5.
Another two sherds (MDC.1.74 and MDC. 1.104) were found
northwest of the Circle in Units 4 and 5, respectively.
CordMarked
Eight cord-marked sherds were identified. George Luer
identified some of these as possibly West Florida Cord Marked
(Willey 1949:388-389). All were found in Block 1 (Figure 21).
Specimen MDC. 1.631.9 was found in Unit 63, Level 2 (Figure
22c). This sherd is characterized by mica inclusions in the
paste. Another sherd, MDC.1.601.9, was recovered from Unit
53, Level4. Sherd MDC.1.161.24 was found in Block 2, Unit
8, Level 8. Four sherds (MDC.1.395.3)-three of which fit
together-were found in Block 1, Unit 22. One larger sherd
(MDC.1.594.4) measuring 5.9 cm by 4.9 cm was recovered
from Unit 49, Level 3 (Figure 22a).
Belle Glade Plain
This pottery type is associated with the Belle Glade Culture
of Lake Okeechobee. It is characterized as a plainware, with a
sand-tempered and spiculate paste, often with the exterior
shaved. This type is most often associated with Glades III
occupations in southeast Florida. Only a single specimen
(FBAR. 1.100.5) was found.
Miscellaneous Incised/ Zoned Punctated
Eleven non-local incised/ zoned punctated sherds that could
not easily be assigned to types were present (Figures 21 and
23). Some of these are suggestive of various Florida Gulf Coast
types, but they are not classified in this study.
One sherd (MDC.1.220.3) has possible Pinellas paste
(Figure 23c). It was recovered from Block 2, Unit 11, Level 9.
The sherd has an intricate, well executed design of angled lines
and punctated zones.
Two possible Sarasota Incised sherds were found. Speci-
men MDC.1.52.7 is a zoned-punctated sherd decorated on
chalky ware (Figure 22d). This sherd was a surface find. A
second sherd (MDC.1.1043.1) also was found.
Three unclassified sherds have some similarities in design
(Figure 23d, f). Each sherd has a linear recessed groove on the
exterior surface with punctations both in the groove and on the
adjacentfield. SpecimenMDC.1.59.1 wasrecoveredfromUnit
1, Level 3, near the outside of the northwest circumference of
the Circle. A second sherd (MDC.1.806.3) was found in basin
Feature 457 that straddles Units 45 and 46. A third sherd
(FBAR. 1.201.5) was found during the FBAR investigations east
of the Circle in Basin MM.
Specimen MDC. 1.63.7 was found in Unit 1 at a depth of5.2
feet above NGVD 1929 (Figure 23e). George Luer notes its
similarity to various St. Simons types (see Williams 1977:145,
TnE nORIDA ANTuROPOLOcIS;T
2006 VOL. 59(3-4)
BRICKELL POINT CERAMIC ANALYSIS
Table 4. Miami Circle Basin Ceramic Types.
* = Collected material not accessible
DC = Deptford Check Stamped
STP = sand-tempered plain
SJ = St. Johns Plain
SJS = St. Johns simple stamped
SJCS = St. Johns Check Stamped
SJL = St. Johns linear check stamped
UC = unclassified chalky ware / plain
UCI = unclassified incised
UC = unclassified
Figure 44d).
One sherd (MDC. 1.166.16) that is similar to a Crystal River
type, with grooves and zones ofpunctates decorating a smooth
sand tempered ware, was found in Unit 10, Level 2 on the
outside of the Circle's northwest basins (Figure 23h).
Another sherd (MDC.1.440.2) with a recessed zone of
punctations, was found in Unit 26, Level 5 (Figure 23g). It may
have been from within a large natural solution feature that is
located on the Circle's southeast circumference.
Decorated with punctations within a zone bordered by a
curvilinear incision, specimen MDC.1.904.1 was recovered
from Unit 63 from Feature 511, a shallow solution hole (Figure
23b). A sherd of micaceous paste with zoned punctations,
specimen MDC.1.832.6, was found in Feature 461 in Unit 48.
Discussion
An analysis of ceramics associated with the Brickell Point
site (8DA12) indicates a wide variety of types representing a
chronological range from the early Glades I through the Glades
III periods, or an approximate span of about 2000 years.
Despite this general duration of site activities, the data is more
revealing regarding the chronology of the site's most intensive
habitation, indicating that the highest incidence of ceramic
types date from the early Glades I through II Periods, suggest-
ing an intensive occupation from 750 B.C. through ca. A.D.
1200. A paucity of Glades III types, such as Surfside Incised,
Matecumbe Incised, and Glades Tooled, suggests that intensive
occupation had ceased by the late Glades II Period.
It has been suggested that the Glades III Period is poorly
Basin FS DC UC UCI STP SJP SJS SJCS SJL Total
20 204 1 1
56 245* 0
66 281* 0
71 307* 0
77 313 1 1
79 315 14 5 1 20
80 316* 0
81 317* 0
92 327* 0
95 372 1 14 8 1 24
100 331 4 4
101 332 3 4
102 333* 0
103 334 1 10 6 17
109 378*______ 0
110 335 0
112 336 0
114 338 0
139 362*, 955 0
140 363 1 31 35 67
141 364 1 10 10
163 400,401
514,957 1 6 2 8
172 409 15 10 _25
457 806 1 4 9 14
484 903 2 ____1 3
Totals 5 1 108 75 3 4 199
CARR
THE FLOlIIDA ANTHIlOPoLOGIsT 2006 VOL 59(3-4)
represented because of a data bias caused by site distur-
bances-specifically the construction of the Brickell Point
Apartments in 1949, which leveled much of the site and
removed the most recent cultural horizon (presumably off-site).
However, there are several lines of evidence that indicate that
a significant Glades III horizon never existed on the Miami
Circle or anywhere on the overall Brickell Point site. If con-
struction activities had impacted and removed a Glades III
horizon, then it would be anticipated that the numerous areas of
redeposited midden thatwere encountered across the site during
excavations and monitoring would include some representation
of the Glades III ceramic assemblage. That was not the case
during the three investigations conducted on the Circle parcel.
In fact, only a single Glades Tooled sherd and a single
Matecumbe Incised sherd were found, despite extensive
collections of redeposited midden areas during monitoring and
the overall phase 1 assessment by FBAR.
Reinforcing these observations is the large quantity of
redeposited and in-situ nineteenth century artifacts found in a
horizon that would have been on top of, and probably physi-
cally associated with, the Glades III horizon. If the Glades III
horizon had been destroyed or removed off-site, then the
nineteenth century assemblage should have suffered a similar
fate, but that is not the case. Disturbances to Block 1 encom-
passing the Miami Circle were extensive, with an estimated 20
to 30 centimeters of midden removed by 1949 apartment
construction activities, but even these disturbances did not
remove all of the numerous nineteenth century historic artifacts
uncovered there, even though no significant quantities of
Glades III ceramics were encountered.
Block 2 encompasses the historic riverbank, part of which
was sealed beneath fill and a limestone rock road that were
placed there during Flagler's dredging of the mouth of the
Miami River ca. 1900. There is no evidence of the removal of
in-situ sediments since that time. Thus, large quantities of
historic artifacts were recovered from the upper levels of Block
2 units during the MDC excavations. However, no Glades III
markers were uncovered, only Late Glades I and Glades II
types such as Key Largo Incised, Opa Locka Incised, Miami
Incised, and Dade Incised.
Further evidence of the chronology of the overall Brickell
Point site (DA12) and sites along the south bank of the Miami
River are the numerous investigations of undisturbed midden
components and the monitoring of construction there that have
yielded no significant evidence of Glades III occupation
anywhere on the south bank of the Miami River. Monitoring of
the Holiday Inn construction at Brickell Point in 1981 and
Phase III investigations conducted after the hotel demolition in
2005 (report pending), and at Brickell Park (Carr et al. 2002)
indicate intensive Glades I and II activities, but no Glades III
Period artifact assemblage.
This overall 8DA12 site chronology is important because it
differs from the 8DA11 site chronology on the opposite, north
side of the Miami River (Table 3). Site 8DA 1 has extensive
evidence of a Glades III occupation. It is mentioned in various
historic accounts as being the town of Tequesta, which was
occupied from the sixteenth through eighteenth centuries
(Griffin et al. 1983). It also has yielded a large quantity of
Glades III pottery types excavated at the Granada and Dupont
Plaza site components (Griffin et al. 1983).
In regard to the specific age of the Circle feature, a review
of the ceramics recovered from 21 of the Circle's 24 basins
indicate that of 224 pottery sherds only fifteen (7%) are
decorated (Figure 24). Only three decorated sherds from the
basins had been identified during the field investigations
according to the field notes (Features 77, 78, 81). A total of
fifteen decorated sherds was documented during the artifact
cataloguing. During the artifact cataloguing, an additional
decorated sherd (MDC. 1.363) was identified from Basin 140.
It is worth noting the Basin 79 decorated sherd (MDC.1.315)
was not recognized by the cataloguers and consequently was
missing from the initial ceramic analysis. Seven decorated
sherds are in the Deptford tradition, specifically two Deptford
Check Stamped (erroneously identified as St. Johns Check
Stamped in the field notes) and the others being St. Johns
simple stamped (n=3) and St. Johns linear check stamped (n=5).
Of particular interest is the number and percentage of St.
Johns ware (n=83; 41%) to sand-tempered plain ware from the
basins (n=108; 55%) (see Table 4). This ratio of St. Johns
Plain is exceptionally high for ceramic assemblages from
southeast Florida, with the exception of the several Late
Archaic/Glades I transitional sites on Biscayne Bay such as the
Santa Maria (DA2132) and Atlantis (DA1082) sites. It should
be noted that the cataloguers identified only 13 St. Johns Plain
in the basins compared to a count of 75 in the final review by
the author, suggesting that the actual number of St. Johns Plain
was under-represented in the initial count and may be four or
five times higher across the site.
In contrast, the percentage of St. Johns sherds at the
Granada site is 17%, which largely represents a Glades II and
II assemblage. Griffin (Griffin et al. 1983:44-45) notes that
the sample is largely biased, since about 90% of the sample was
discarded because the sherds were considered too small or
plain. Thus, the St. Johns ceramic assemblage at the Granada
site is probably less than 10%.
One unusual sherd (MDC. 1.985), a chalky ware tetrapod
with check stamping, was found in the spoil in Block 1 near the
Circle. This tetrapod is a classic Deptford trait. A similar
tetrapod has been found in the Oak Knoll Burial Mound
8LL729 (Dickel and Carr 1991:Figure 13). Ceramic pipes, like
those from near the Circle, also are known in Florida on the
Deptford horizon (Luer 1995).
This high incidence of St. Johns ware, the Deptford influ-
enced decorated types, and the absence of Glades series
decorated pottery associated with the basins strongly suggests
an age range of ca. 700 B.C. to A.D. 200 for the Circle's
construction. The Deptford types represent almost 27% of the
site's decorated wares. This supports the author's hypothesis
that the basins contain cultural material of a limited temporal
range that probably represent the age of the Circle's construc-
tion or activities occurring shortly after its construction.
The Deptford influence at Brickell Point, particularly with
the construction and use of the Miami Circle structure, is of
particular interest since Deptford ceramic assemblages have
been rarely reported in southern Florida. Among the few sites
2006 VOL. 59(3-4)
THE nflOWDA ANTHROPOLOGISr
325 330 335
150 FS 903.7
5 FS 903.8
FS 315 FS806
FS 313 FS 86 /FS 903.10
145
FS 372
4 66 44 45 FS 332
140
1 36 48 47 49 50 9
135
10 14 62 63 61 64
130
S 20 24 23 19 22 2 219
125
FS3640 34 41 42 60 53 56
FS 204 120
18 39 31 33
115
57 27 B29-
110
52 28
105 N
295 300 305 310 315 320 325
A Deptford Check Stamped 0 St. Johns Linear Check
Stamped 0 5 10 feet
Deptford Simple Stamped V St. Johns Check Stamped
[ Unclassified Zoned
St. Johns Simple Stamped Punctate
Figure 24. Distribution of decorated pottery within the Miami Circle basins.
with Deptford ceramic assemblages are the Oak Knoll Mound Mound (8HG676) in Highlands County (Austin 1993), the
(8LL729) in Lee County (Dickel and Carr 1991), the Royce Goodland site (8CR45) in Collier County, and most recently,
CARR
BRICKELL POINT CERAMIC ANALYSIS
THEFLOIDAANTROPLOIST200 VO.5934
8BD259 and 8BD3341, both located on the New River in
Broward County.
Conclusion
The ceramic analysis of 26,281 sherds from excavations of
the Miami Circle and surrounding portion of site 8DA12
indicates two interesting trends. The collection is dominated by
sand-tempered plain pottery, with a very small array of deco-
rated Glades Series sherds and St. Johns Plain sherds, suggest-
ing occupation from early Glades I through II periods (ca. 750
B.C. through ca. A.D. 1200). Analysis of sherds from the
Circle's carved basins indicate a more restricted time range for
construction of the Miami Circle feature, likely ca. 700 B.C. to
A.D. 200. The presence of Deptford Series, Gulf Check
Stamped, ceramic platform pipes, and extra-local miscellaneous
incised and zoned punctated sherds that evoke early Gulf Coast
types all suggest a Middle Woodland element to the site
occupation. This is supported by other artifacts and features at
the site, which also suggests a Middle Woodland affiliation; for
example, the galena artifacts; the size and configuration of the
Miami Circle feature, and a link between the chipped stone
artifacts of the site and Fort Center (Austin 2004:128; Carr and
Ricisak 2000:281-282; Wheeler 2000:312). In this way, the
ceramic assemblage is unusual but complements the other
Middle Woodland elements present at the site.
Acknowledgments
The author is indebted to a number of people who made this
analysis possible. I am particularly grateful to Stephanie Faulkner
who drew and measured many of the decorated sherds; her drawings
greatly facilitated my work. Edda Rodriquez also worked on the
individual sherd descriptions. Comments and identifications from Dr.
Ryan Wheeler, George Luer, and Ann Cordell were of exceptional
value, particularly in regard to identifying many of the non-local
ceramics. I also am indebted to the Historical Museum of Southern
Florida, particularly to Jorge Zamanillo, who provided access to the
Brickell Point collection. Invaluable help also was provided by Joe
Mankowski, who photographed the collection, and John Beriault, who
provided valuable information about ceramics found at the Circle.
Also, my thanks to Melinda and Bob Copper who provided illustra-
tions of many of the diagnostic ceramics.
References Cited
Austin, Robert J.
1993 The Royce Mound: Middle Woodland Exchange and
Mortuary Customs in South Florida. The Florida Anthro-
pologist 46(4):291-309.
2004 Chipped Stone Artifacts from the Miami Circle Excavations
at Brickell Point. The Florida Anthropologist 57(1-2):85-
131.
Bense, Judith A. (editor)
1985 Hawkshaw: Prehistory and History in an Urban Neighbor-
hood in Pensacola, Florida. Report of Investigations 7,
Office of Cultural and Archaeological Research, University
of West Florida, Pensacola.
Caldwell, Joseph R., and A. J. Waring, Jr.
1939a Pottery Type Descriptions. Southeastern Archaeological
Conference Newsletter 1(5):4-12.
1939b Pottery Type Descriptions. Southeastern Archaeological
Conference Newsletter 1(6):1-9.
Carr, Robert S.
1981 Dade County Historic Survey, Final Report: The Archaeo-
logical Survey. Metropolitan Dade County Office of
Community and Economic Development, Miami.
2002 The Archaeology of Everglades Tree Islands. In Tree
Islands of the Everglades, edited by Fred H. Sklar and
Arnold Van der Valk, pp. 187-206. Kluwer Academic
Publishers, Boston.
Carr, Robert S., and John G. Beriault
1984 Prehistoric Man in Southern Florida. In Environments of
South Florida: Present and Past, revised ed., edited by
Patrick J. Gleason, pp. 1-14. Miami Geological Society
Memoir No. 2, Coral Gables.
Carr, Robert S., and John Ricisak
2000 Preliminary Report on Salvage Archaeological Investiga-
tions of the Brickell Point Site (8DA12), Including the
Miami Circle. The Florida Anthropologist 53(4):260-284.
DePratter, Chester B.
1979 Ceramics. In The Anthropology ofSt. Catherines Island 2.
The Rufuge-Deptford Mortuary Complex, edited by David
Hurst Thomas and Clark Spencer Larsen, pp. 109-132.
Anthropological Papers ofthe American Museum ofNatural
History, vol. 56, pt. 1. New York.
Dickel, David, and Robert S. Carr
1991 Archaeological Investigations of the Oak Knoll Mound,
8LL729, Lee County, Florida. ArchaeologicalandHistori-
cal Conservancy Technical Report #21.
Goggin, John M.
1939 A Ceramic Sequence in South Florida. New Mexico Anthro-
pologist 3:35-40.
1940 The Distribution of Pottery Wares in the Glades Archaeo-
logical Area of South Florida. New Mexico Anthropologist
4:22-33.
1944 A Tentative Formulation ofPottery Types for the Glades
Area. Mimeographed. Yale University, New Haven.
1950a Cultural Occupation at Goodland Point, Florida. The
Florida Anthropologist 2:65-91.
1950b Stratigraphic Tests in the Everglades National Park. Ameri-
can Antiquity 15:228-246.
1952 Space and Time Perspectives in Northern St. Johns Archeol-
ogy, Florida. Yale University Publications in Anthropology
47, New Haven.
n.d. The Archeology ofthe Glades Area, Southern Florida, circa
1949, with additions. Unpublished typescript.
Goggin, John M., and Frank H. Sommer III
1949 Excavations on Upper Matecumbe Key, Florida. Yale
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
University Publications in Anthropology 41, New Haven.
Griffin, John W.
2002 Archaeology of the Everglades. Edited by Jerald T.
Milanich and James J. Miller. University of Press Florida,
Gainesville.
Griffin, John W., Sue B. Richardson, Mary Pohl, Carl D. McMurray,
C. Margaret Scarry, Suszanne K. Fish, Elizabeth S. Wing, L. Jill
Loucks, and Marcia K. Welch
1983 Excavations at the Granada Site. Archaeology and History
of the Granada Site, Vol. 1. Florida Division of Archives,
History and Records Management, Tallahassee.
Luer, George M.
1995 Pipe Fragments from Ortona, Southern Florida: Comments
on Platform Pipe Styles, Functions, and Middle Woodland
Exchange. The Florida Anthropologist 48:301-308.
Milanich, Jerald T.
1994 Archaeology ofPrecolumbian Florida. University of Press
Florida, Gainesville.
Wheeler, Ryan J.
2000 The Archaeology of Brickell Point and the Miami Circle.
The Florida Anthropologist 53(4):294-322.
Widmer, Randolph J.
2004 Archaeological Investigations at the Brickell Point Site,
8DA12, Operation 3. The Florida Anthropologist 57(1-
2): 11-57.
Willey, Gordon R.
1949 Archeology ofthe Florida GulfCoast. Smithsonian Miscel-
laneous Collections 113. Smithsonian Institution, Washing-
ton, D.C.
Williams, Stephen (editor)
1977 The Waring Papers: The Collected Works of Antonio J
Waring, Jr. Papers of the Peabody Museum of Archaeology
and Ethnology, Volume 58, Harvard University, Cambridge.
CARR
BRICKELL POINT CERAMIC ANALYSIS
BACK ISSUES
of
The Florida Anthropologist
are available from the
Palm Beach Museum of Natural
History:
www.pbmnh.org
i I
INTEGRATED SPATIAL TECHNOLOGIES:
HIGH DEFINITION DOCUMENTATION OF THE MIAMI CIRCLE AND ROYAL PALM CIRCLES
LORI COLLINS' AND TRAVIS DOERING2, WITH ADDITIONAL COMMENTARY BY ROBERT S. CARR3
Department ofAnthropology, University of South Florida, 4202 E. Fowler Ave., SOC 107, Tampa, FL 33620-8100
'E-mail: lcollins@cas.usf edu; 'E-mail: tdoering@mail.usf edu
3Archaeological and Historical Conservancy, Inc., Davie Professional Plaza, 4800 Davie Rd., Suite 107, Davie, FL 33314
E-mail: archlgcl@bellsouth.net
The Miami Circle represents the most complete remains to be
identified archaeologically in southeastern Florida of what
appears to be a Native American architectural structure. As
such, it is a rare and possibly unique cultural resource.
[Weisman et. al. 2000]
The contemporary Miami skyline is defined by the modem
skyscrapers that line the entrance to the Miami River represent-
ing one end of at least a 2,000 year sequence of human occupa-
tion. At the opposite end of this continuum are the prehistoric
Miami Circle, a component of the Brickell Point site (8DA12)
on the south side of the Miami River, and the Royal Palm
Circle, which is part of the Miami Midden #1 (8DA11), on the
opposite north bank, also known as the Granada site (Figure 1).
Both of these areas represent landscapes that have been
repeatedly modified through time, from prehistoric Tequesta
occupation to the modem real estate development. Here, where
the Biscayne Bay shoreline merges with the Miami River,
prehistoric settlements, historic trading posts, military forts,
grand hotels, and home sites have been supplanted by modem
apartment complexes, commercial buildings, streets, and
parking lots (Florida Division of Historical Resources 2004a;
Weisman et. al. 2000; Wheeler and Carr 2004). Understanding
how this ancient landscape evolved and connects to the present
is a significant issue in the ongoing archaeological study of this
area.
The investigation and resulting interpretations of the
prehistoric features are dependent upon accurate and complete
data. Through the use of an integrated series of cutting edge
spatial technologies, data were acquired with an intensity,
speed, and accuracy not previously attainable. The multiple
datasets create a baseline for highly precise measurement,
analyses and landscape consideration. The databases cumula-
tively represent a 'living document' that combines the results of
previous work with the present surveys and is expandable and
upgradeable to include any future investigation of the area.
The work described here was performed by a team of
archaeologists and technical specialists from the Alliance for
Integrated Spatial Technologies at the University of South
Florida. The Miami Circle Project was conducted at the request
of State Archaeologist Dr. Ryan Wheeler, a lead investigator of
the Circle and larger Brickell Point site projects. The Royal
Palm Circle project was performed for Robert Carr and the
AHCI, as part of a larger compliance effort conducted to assess
the impacts to cultural resources from high rise condominium
development on the north side of the river. Both of these
projects employed new spatial survey and documentation
techniques that featured three-dimensional laser scanning. This
article focuses on how these technologies assisted archaeologi-
cal research, documentation, analysis, and ultimately benefit
public understanding and appreciation of archaeological
landscapes.
High Definition Documentation Survey (HDDS)
High Definition Documentation Survey (HDDS) is a series
of integrated spatial recording technologies that provide
extremely accurate data more rapidly, efficiently, and effec-
tively than is possible with traditional survey techniques (Frei
et al. 2000). At the center of HDDS is three-dimensional laser
scanning, a technique that produces a stream of precisely
measured points of a surface that are then processed into three-
dimensional spatial models. This non-contact, non-invasive
technique allows the analysis, measurement, and manipulation
of objects that range in size from individual handheld artifacts,
to architectural structures, to entire terrains and landscapes.
Raw data collected by the laser scanning process are
referred to as "point clouds." The cloud is produced by millions
of individually referenced measurements, each with x,y,z
coordinates that are relative to all other collected point loca-
tions. Depending on the specific scanning technique used
accuracy can range from 50 microns to a millimeter. Visualiza-
tion of the data is immediate and in real time as it is collected.
Three-dimensional software packages allow furtherrefinement
and post processing of the data allows the creation of accurate
3D models. The data can be exported to numerous formats that
can be viewed in more traditional 2D software platforms such
as ArcGIS and AutoCAD. Data collected from conventional
surveys, such as tabular spreadsheet information with artifact
provenience, can be integrated with HDDS datasets. Field
documentation of excavation areas, mapped features, site plans
and grids, and artifact concentrations, can be performed during
HDDS. Attribute information can then be linked to the collected
spatial locations during post-processing and database develop-
ment.
A primary objective of the USF team's efforts was to
rapidly and accurately document the sites from a variety of
perspectives and to generate a broad analytical baseline. The
spatial datasets provide a means for reliable long-term analysis
THE FLORIDA ANTHROPOLOGIST
VOL. 59(3-4)
SEPTEMBER-DECEMBER 2006
THE FLORIDA ANTHROPOLOGIST 2006 VOL 59(34)
U U
--1_r-"LJ lMeters S \I\ [ ,IT\ OF
A Recorded Archaeological Sts 0 50 100 200 S T Fl' I
Figure 1. Map depicting recorded archaeological sites in the vicinity of the Miami Circle and Royal Palm Circle features.
and virtual preservation of information. Each component of the
site can be examined in meticulous detail as well as analyzed
diachronically relative to its surroundings. Thus, the HDDS
data can be used to address evolving research questions and to
assist with interpretation, visualization, and three-dimensional
representation for the public and for on-going archaeological
inquiry.
For both the Miami Circle and the subsequent Royal Palm
Circle HDDS projects, three-dimensional laser scanning
applications were integrated with other survey techniques to
produce richer and more complete datasets. Additional spatial
reference data were collected with robotic total stations and
survey grade, Real Time Kinetic Global Positioning System
(RTK-GPS) equipment that provided sub-centimeter vertical
and horizontal spatial control at the site areas (Figure 2).
Additionally, the Miami Circle HDDS project used data
previously collected through conventional transit survey to
further improve the analytical capabilities. A Geographical
Information Systems (GIS) database synthesized all previous
datasets and available imagery with the spatial control from the
current work to combine, visualize, and query the multiple
types of archaeological information collected (Collins and
Wheeler 2006; Weisman and Collins 2003).
The integration of the Royal Palm Circle data with those
from the Miami Circle demonstrated how HDDS techniques can
document individual features as well as merge discrete datasets
for comparison and analysis across space. By viewing these
sites from differing scales of analysis, the broader landscape at
the mouth of the Miami River is brought into focus and a
clearer understanding of individual features and relationships
with larger scale site areas is possible (Figure 3). Taken
together, these surveys allow cross-comparison and examina-
tion of the area's settlement history and demonstrate the
benefits of rapid inventory techniques. HDDS can virtually
preserve features and sites even in the shadow of high-rise
development (Collins and Wheeler 2006).
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
COLLINS, DOEI1ING, AND CARR INTEGRATED SPATIAL TECHNOLOGIES
Figure 2. Equipment used for High Definition Documentation Surveys at the Miami and Royal Palm Circle features. A robotic
total station (LEFT) and survey grade GPS equipment (RIGHT) were integrated with three-dimensional laser scanning
equipment (CENTER) for data acquisition. Equipment is shown at the Royal Palm Circle project.
The Miami Circle
The Miami Circle is a feature within the Brickell Point site
(8DA12) that is composed of a series of various sized and
shaped holes carved into the limestone bedrock. A series of 24
semi-rectangular basin-like elements that were cut into the
limestone form the pattern for this circular feature. In addition
to the larger basins there are numerous other round & ovoid
holes, many of which contain smaller secondary holes (Florida
Division of Historical Resources 2004a; Wheeler and Carr
2004). The initial phase of the HDDS project was carried out
on July 22 and 23, 2003, when the team from the University of
South Florida conducted a survey of the Brickell Point site
(8DA12). A primary focus was on the Miami Circle feature,
which was documented by three-dimensional laser scanning.
Researchers have proposed thatthe Miami Circle represents
the foundation of a 2,000 year old prehistoric structure, and that
further analysis of the Circle's holes and basins could reveal
more information about little-known Native American architec-
ture in Florida (Florida Division of Historical Resources 2004a;
Weisman 2000; Wheeler 2000b). The antiquity of the site has
been demonstrated through geophysical analysis. A laminated
patina or duricrust, which forms at an average rate of 1mm per
1000 years, coats the limestone facing of the circular holes and
basins associated with the ancient construction (Means and
Scott 2000). Researchers believe the Miami Circle represents
only a portion ofa larger settlement which remains unexcavated
and covered by earth at the Brickell Point site and along the
ancient shorelines of the Biscayne Bay at mouth of the Miami
River (Carr and Ricisak 2000; Florida Division of Historical
Resources 2004a; Weisman et. al. 2000; Wheeler 2000b;
Wheeler and Carr 2004; Widmer 2004). This hypothesis was
supported by recent investigations conducted on the north side
of the Miami River, directly across from the Miami Circle,
where another circle feature, along with human interments, and
prehistoric middens suggest contemporaneity or continuity of
occupation (Florida Division of Historical Resources 2004a;
Widmer 2004; Robert S. Carr, personal communication, 2006).
The Royal Palm Circle
On-going archaeological investigations and analyses by the
COLLINS, DOERING, AND CARR
INTEGRATED SPATIAL TECHNOLOGIES
164 THE FLORIDA ANTHROPOLOGIST 2006 VOL. 59(3-4)
samwweaeeseunernmesntene areata ecennitenecantnanecncenaeameansaeteeatninni niet neeretntttnanottniatitass isaac CEOS AE
Pane
rere ila
yee 7
Figure 3a and 3b. Viewsheds of the Miami River landscape, showing the Miami Circle amidst current and projected urban
development.
Archaeological and Historical Conservancy, Inc. (AHCI), under Robert S. Carr, personal communication, 2006). Dubbed the
the direction of Robert Carr, revealed an additional prehistoric Royal Palm Circle by the USF team, the feature is referred to as
circle feature, dense midden accumulations, and numerous the DA11 Circle, indicating its location within the previously
ancient burial areas and ossuaries (Associated Press 2006; recorded Miami Midden #1 site area (Griffin et al. 1982)
COLLINS, DOEIUNG, AND CAliR INTEGRATED SPATIAL TECHNOLOGIES
Figure 4. The Royal Palm Circle feature (DAll Circle), showing historic trench and foundational walls of the Royal Palm
Hotel in close proximity. Photo courtesy of Richard Haiduven.
(Figure 4). This circular feature is comprised of a series of three
concentric rings carved into the limestone on the north bank of
the river, directly opposite the Miami Circle. The construction
techniques, types and sizes of the holes, their pattern, and
feature dimensions at both sites share a high degree of correla-
tion (Figure 5). The duricrust formation, documented in the
prehistoric cavities at the Miami Circle, is present at the Royal
Palm Circle which, along with the inter-site uniformity, dense
midden accumulations, and imbedded faunal remains suggests
a mutual and corresponding antiquity (Robert S. Carr, personal
communication, 2006) (Figure 6).
At both circle features there is evidence of repeated intru-
sions created by historic construction events. In this respect, the
ground surface on both sides of the entrance to the Miami River
provides researchers with palimpsests of landscape develop-
ment. Implementation of the HDDS, specificallythe capabilities
of three-dimensional laser scanning, allows the historic and
prehistoric features to be detected, isolated, and studied
independently. For example, within the area of the Royal Palm
Circle feature are intrusive foundations and infrastructure of
Henry Flagler's Royal Palm Hotel, built in 1897, which were
documented along with numerous historic artifacts from the
Fort Dallas occupation (1837-1858) within the same area of the
prehistoric Miami Midden #1 (8DA11), and the Royal Palm
Circle feature (AHCI 2006; Schwartz 2004). At the Miami
Circle, it has been suggested that the historic intrusions are the
result of twentieth century construction activities related to the
Palm Tree Inn and Brickell Point Apartments (Carr and Ricisak
2000; Collins and Wheeler 2006; Florida Division of Historical
Resources 2004a; Wheeler 2000a, b; Wheeler and Carr 2004).
These historic elements are also able to be isolated and studied
independently from the prehistoric feature construction (Figure
7).
Survey Methods and Analysis
At the Miami Circle coordinates had been previously
established on known control points and were provided to USF
by a local surveyor. These coordinates were used for total
station set-ups and became the basis for logistical control
during the HDDS project. The data points were acquired in the
same coordinate system, which allowed for all new and
previously collected data to be integrated and displayed
together with real-world locational placement.
The laser scanning process began with the placement of a
series of targets around the area to be scanned. These target
locations were recorded using a robotic total station and survey-
grade GPS that established a control network of known
reference points. The precise location of these points was
critical to allow data from multiple scans to be properly and
accurately meshed (Figure 8) (Weisman and Collins 2003).
Multiple scans were conducted from a variety of locations to
ensure that the feature area was captured from a broad perspec-
tive in order to provide the most robust datasets possible. Laser
COLLINS, DOERING, AND CARR
INTEGRATED SPATIAL TECHNOLOGIES
Tm FLORIDA ANThROPOLOGIST 2006 VOL. 59(3-4)
Figure 5. Overhead photograph of the Royal Palm Circle feature
concentric rings and its location in relation to the Miami Circle 1
courtesy of Richard Haiduven.
technology is based on line of sight; therefore, multiple scans
minimize any obscured areas or "shadows" that may occur in
the data (Frei et al. 2000). Two scans were performed from
opposing sides of the Circle, approximately six feet above
ground level, and two others were made from a stabilized
platform 60 feet above the circle feature (Figure 9).
Data from all the scans were integrated into an extraordi-
narily dense, geo-referenced "point cloud" comprised of
millions of x, y, z coordinate positions. Once a scene is cap-
tured at this point density, analysis can take place in a near
'virtual' environment, and eliminates the need to return and re-
survey the site in the future. This capability is significant
because the Miami Circle has since been covered with
geotextile and gravel and re-buried. A com-
plete 3D model of the site is available and
allows for continued exploration without the
necessity of uncovering the site again. The
ability to view the Circle in a virtual environ-
ment, or recreate an exact copy is important
from a public involvement standpoint, espe-
cially in light of the various objections raised
to the Circle's re-burial (Merzer 2003).
The ability to virtually preserve features
was important at the 2005-2006 HDDS pro-
ject across the river at the Royal Palm Circle
feature, where development has proceeded
and preservation did not occur. Similar work-
flow methodologies were used, with target
placement surrounding the circular feature
and excavated surfaces, and laser scan data
were collected from multiple ground based
positions. The need for rapid assessment and
acquisition of HDDS data was imperative at
this feature. High rise construction was en-
croaching and the site was being flooded in
order to keep the construction area dry (Fig-
ure 10). Laser scanning revealed features that
were highly consistent with the Miami Circle
S in its size and layout, although basin holes and
cuts were smaller and more ovoid rather than
rectangular, researchers believe the feature is
contemporaneous to the Miami Circle and
may represent a 'blueprint' for construction
(Robert S. Carr, personal communication,
2006; Merzer 2005).
S Imaging of the laser scan data of the
S excavated area showed a series of three con-
centric circles (Figure 11). Comparative
analysis between the circle features on the
north and south sides of the river show nearly
identical spatial layouts, with the radius of the
Miami Circle measuring 18.12 feet, and the
radius of the most distinctive central concen-
tric ring of the Royal Palm Circle at 18.02 feet
(Figure 12). The Miami Circle feature also
e showing the contained evidence for an outer concentric
feature. Photo circle (Weisman 2000). These similarities
help to confirm researcher's hypotheses that
there was a construction typology used by
Southeastern Native Americans for structures such as council
houses, chiefresidences and other similar constructions (Robert
S. Carr, personal communication, 2006; Collins and Wheeler
2006; Merzer 2005; Weisman 2000; Ryan J. Wheeler, personal
communication, 2006).
Post Processing and Data Integration
The initial HDDS post processing steps at both the Miami
Circle and Royal Palm projects involvedthe 3D laser scan data.
The density of the point cloud (more than 4,000,000 points)
was reduced, to allow for a more manageable analytical dataset
of the carved basin cuts and ancillary holes that comprised the
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
of points were mathematically selected in the
scanner software to achieve a workable dataset.
For the Miami Circle HDDS project, a total of
136,766 vertical, horizontal and elevational
points were exported to a text file format for use
inthe production of surface maps in. a variety of
L survey and GIS software platforms (Figure 13).
As a point of technological comparison, the
conventional transit survey collected nearly
4,000 points over a period of 2 weeks in the
U field and required 120 work hours (Ryan J.
I Wheeler, personal communication, 2003). The
-laser scan survey collected in excess of
.---- 4,000,000 points in less than 16 work hours. In
addition to the quantitative disparity, each ofthe
scan data points are accurate to within a milli-
meter, which conventional survey cannot ap-
proximate.
The processed HDDS data also provided
precise and reliable measurements and depic-
tions that were useful in the reformulation and
Figure 6. Cut hole forming part of the Royal Palm Circle feature with analysis of the Miami Circle's layout and di-
associated faunal material present, mensions. Previous maps of the feature had
been based on an aerial photograph that was
taken from east of the Circle's center, a position
that caused a parallax distortion of the Circle's
true shape and size. The error created by this
type of distortion is amplified with distance and
it introduces unanticipated errors into the depic-
tions and, in turn, into the interpretations of the
feature (Figure 14). The scan data were used to
critically address issues of central location,
pattern occurrence, shape, and spatial associa-
tions (Figure 15). When the previously plotted
features were compared against the HDDS data,
the distortion caused by the aforementioned
parallax phenomena became apparent (Figure
16). Because the laser scanner collects spatial
data that are all relative to a single point, distor-
tion does not occur (Alessandri et al. 2005).
A number of other visualizations and dia-
grams were produced through various process-
ing techniques using the Miami Circle HDDS
data. A Triangulated Irregular Network (TIN)
model was generated that proved effective in
the examination of site-specific details such as
slope, aspect, and line of sight, and the model
was used to produce elevation contours across
the feature. Vertical cut marks, similar to those
made by shell tools in replication studies
(Florida Division of Historical Resources
Figure 7. Three dimensional scan data of Miami Circle showing detail of 2004a; Wheeler and Carr 2004) were observed
terrain and intrusive historic elements (i.e., Septic tank, planter footing and in many of the carved holes and basins. The
trench cut, thought associated with the former Palm Tree Inn and Brickell shape, morphology, and dimensions of each of
Point Apartments. these cavities were able to be carefully exam-
ined in 3D profile (Figure 17). The differences
between the linear historic feature cuts and the
circular features. This procedure does not mean that any of the randomness of the prehistoric basin cuts were clearly
original raw data is discarded, but simply that a lesser number discernable in the new visualizations and demonstrated distinc-
COLLINS, D)OERING, AND CARR
INTEGRATED SPATIAL TECHNOLOGIES
THE FLOlUDA ANTHROPOLOGIST 2006 VOL 59(34)
The Circle
\
. I.-
0%,
Figure 8. Scanning diagram image showing locational target placement at the Miami Circle feature. Targets are used to
allow meshing of scan data from multiple positions and are spatially registered using GPS and total station, which provide
precise vertical and horizontal control.
r :' L, ,
Figure 9. Laser scanning from above the Miami Circle utilized a stabilized platform elevated 60 feet above the feature.
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
COLLINS, DOElUNG, AND CARR INTEGRATED SPATIAL TECHNOLOGIES
Figure 10. Collins and Carr assess water inundation at the Royal Palm Circle site that occurred during the HDDS survey.
tive differences in construction (Collins and Wheeler 2006)
(Figure 18).
Development of a Project-Wide
Geographic Information System (GIS)
All available HDDS and traditionally collected spatial and
archaeological datasets were entered into a Geographic Infor-
mation System (GIS) that produced a virtual model of the
Miami Circle feature. This GIS serves as a 'living document'
that can be expanded and improved as new information is
discovered and documented. The addition of data from the
recently discovered Royal Palm features on the opposite
riverbank is an excellent example of this inclusive and compre-
hensive capability.
Wheeler provided USF with additional datasets and
information collected in the earlier mapping and excavation
project conducted at the site. Data from conventional transit
surveys showed site boundaries, the placement of archaeologi-
cal excavations (Beriault 2000; Carr and Ricisak 2000), and the
locations of recovered lithic, shell and bone artifacts (Austin
2004; Wheeler 2004a, b; Wheeler and Carr 2004). All of these
previously collected and documented data were able to be
included in the current HDDS analysis project (Figures 19 and
20). More extensive portions of the outer perimeter of the site
had been documented during earlier survey, but were not
exposed when the HDDS was conducted. Thus, the previously
collected transit survey data provided valuable details and
supplementary information that was integrated with the HDDS
to produce a fuller more robust GIS dataset. These maps also
provide for an interpretation of artifact values, and give a
quantitative indication of distribution across the site surface.
The ability to integrate data from multiple sources and
survey techniques and then use the cumulative data to produce
depictions at multiple scales and perspectives is significant.
Dynamic maps and media can assist in the making of informed
decisions about future resource management and in understand-
ing changes to the cultural landscape through time. The
feasibility and successful application of these HDDS technolo-
gies has been demonstrated by the results of these projects, both
individually and collectively.
Future Applications
In October of 2003, the Miami Circle was covered with
geotextile fabric and gravel following recommendations from
archaeologists due to site erosion processes that were occurring
(Wheeler and Carr 2004). This conservation measure has led to
criticisms regarding the inaccessibility of the site to the public
and the lack of interpretation (Merzer 2003). The HDDS
conducted in this present study has virtually preserved the site
by providing a scaleable, precise three-dimensional representa-
tion that will benefit not only future researchers, but can be
used to develop a cost-effective public interpretation and model
of the site. Scanning data and the developed surface models can
now be exported to five-axis routers and three-dimensional
COLLINS, DOERING, AND CARR
INTEGRATED SPATIAL TECHNOLOGIES
Figure 11. Wire frame image developed from scan data collected at the Royal Palm Circle feature. Lines indicate concentric circular cut features revealed during
survey.
CoLLINS, DOERING, AND CARR INTEGRATED SPATIAL TECHNOLOGIES
Figure 12. Radius measurements generated from the scan survey data using definitive features from the middle concentric
ring on the Royal Palm Circle feature (RIGHT) On the left is the radius of cut basins from the Miami Circle feature.
S.' >.' A -'
9- ^.2" it' "
.w' 4 "-I00
A .A a- I IC E.'
Figure 13. Image mesh created by merged scan data collected from various positional set-ups. The density and richness
of the data collected have the appearance of a photograph, but are actually highly accurate x, y and z spatial position
points.
COLLINS, DOERING, AND CARR
INTEGRATED SPATIAL TECHNOLOGIES
TuE FLORIDA ANTIIROPOLOGIST 2006 VOL.59(3-4)
Figure 14. Overlay of hand drawn, mapped features (Weisman et. al. 2000: 344) on scan data contour map. Note the
parallax phenomenon causes drift from left to right, with the most evident distortion along the eastern half of the Circle.
Figure 15. Map showing feature locations drawn with standard techniques against contour map produced from scan data.
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
COLLINS, DOEIUNG, AND CAlm INTEGRATED SPATIAL TECHNOLOGIES
Figure 16. Digital Terrain Model (DTM) produced from the scan data. North and south medial axis lines have been added,
and lines of previously examined linear hole patterns depicted.
printing equipment to produce highly accurate models and exact
reproductions of the site at a variety of scales and perspectives.
These technologies offer an economical means of capturing and
modeling the data, which can be used for on or off-site interpre-
tation and documentation (Weisman and Collins 2003). Web
site development, including digital media files and video
documentation of the scanning process has been produced with
data and information collected from surveys and excavations
(Florida Division of Historical Resources n.d.). Scan images
and multi-media representation of the site can be used in a
variety of ways to enhance visitor experience through the
development of kiosks, interpretive, and other public heritage
presentations.
In the case of the Royal Palm Circle documentation, the
HDDS stands as the most detailed spatial record documentation
of the feature. The virtual digital preservation, along with more
traditional field sketch maps and photographs, is all that
remains for future analysis and comparison, as the feature is
now obscured by the Metropolitan Miami development (Figure
22). On-going work and analysis by ACHI and Robert Carr will
utilize the HDDS to visualize and understand the complex
terrain and modifications to the landscape through time.
Robert Carr:
Comments and Observations from an End-User
The challenge for the archaeologists and technicians
conducting the compliance project on portions of the Miami
Midden #1 including the Royal Palm Circle, was how to
efficiently create an accurate and effective record of the
significant features while under the severe time and financial
constraints that characterize many salvage-oriented projects.
Prior to the involvement of the University of South Florida
researchers, conventional transit and compass mapping tech-
niques were used to record the Royal Palm Circle and associ-
ated excavation areas. These maps were created by field
archaeologists and a professional surveyor. The maps focused
on the principal cut holes and basins that characterize the circle
and other terrain features, but other minor or smaller holes were
omitted from some of the drawings. The conventional methods
alone were incapable of capturing the complexity of the
limestone bedrock terrain in its entirety. In contrast, the maps
generated with the laser scanning data captures the intricacy of
the site's features in extreme detail and accuracy. The precise
renderings allowed for comprehensive, indepth analysis.
In the case of the Miami Circle, the conventional maps
made prior to USF researcher involvement were largely artistic
interpretations. These maps offered an easily understandable
display of features, but lacked accuracy of details and dimen-
sions. The surveyor's data and benchmark positions provided
locational control at both the circles, giving a degree of
dimensional accuracy. These data, however, simplified the
basin hole shapes into schematic forms that could be mislead-
ing. Laser mapping, by comparison, provided a precise and
accurate rendering of the true form and dimensions of the
specific features and other "background" features (both natural
and cultural). The scan data provided consistent visual continu-
ity allowing the observer to compare the greater landscape with
the smaller site elements.
It is estimated that the cost of the conventional and manual
COLLINS, DOERING, AND CARR
INTEGRATED SPATIAL TECHNOLOGIES
TH LRD ATRPLGIT20 OL 934
bi.
E*Iaame...m*
--
M*d
Dd Ms
Figure 17. Three-dimensional cross-section, feature extraction and measurement (RIGHT), are easily accomplished with
HDDS datasets using software such as CycloneTM.
Figure 18. Prehistoric construction methodologies as seen through cut analysis, angle, measurement, and view comparison
can be examined using HDDS datasets. Here Feature 363 is examined for morphology and metrology.
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
-8 J ,' s- --- f -- -- -- "-- -
a ML 4 J'b 9 Zr. re
COLUMNS, DOERING, AND CARR INTEGRATED SPATIAL TECHNOLOGIES
Figure 19. Chart showing data capture synthesis survey methods. HDDS uses an integration of data collected from a
variety of survey and visualization techniques to produce final spatial outputs.
1
-I
Men ,, H
o 25 25
Figure 20. Shell, bone and lithic artifact locations, feature locations, and previous excavation unit locations, shown by
designated numbers, are able to be considered with the scan spatial data.
COLLINS, DoERImG, AND CARR
INMEGRATED SPATUAL TECHNOLOGIES
TiEFLRDAATHOOOGS 20 VL5934
Figure 21. HDDS survey of the Royal Palm Circle Feature included use of RTK-GPS, Robotic total station, and long-range
three-dimensional scanner.
mapping of the Royal Palm Circle and the Metropolitan
development parcel bedrock greatly exceeded the cost of the
laser mapping. Conventional field mapping and documentation
on this project took much more time and resulted in much less
detail than did the laser scanning survey. While laser scanning
does not replace the need for detailed site and feature
recordation by conventional methods, it can be integrated with
more standard documentation techniques to produce stream-
lined, efficient, accurate and rich datasets, from which future
analysis will benefit. Cost alone is not necessarily an argument
to use laser scanning over conventional techniques, especially
when an integration of methods is ultimately most beneficial.
On the Royal Palm and Miami Circle projects it was this
synthesis of approaches that demonstrate intrinsic values of
eachmethod. High Definition Documentation Survey combines
methodologies to include three-dimensional laser mapping, and
in these cases, proved to be a cost effective approach when
absolute deadlines and budget constraints were a project reality.
The virtual preservation through this type of complete docu-
mentation will allow continuing research and analysis without
the need to return to the field. This is an important consider-
ation when, as in the case of the Royal Palm Circle and Metro-
politan developmentproject, features documented no longer are
available for research consideration.
Conclusion
Accidents sometimes happen in historic preservation, with
some sites and locations preserved and others, although similar,
lost. The constantly changing landscape of downtown Miami
exemplifies this situation and points out the need for archaeo-
logical methods that allow for rapid assessment of large-scale
areas and for analysis to include the landscape rather than a
site-by-site consideration. HDDS techniques are cost effective,
highly accurate, and are performed rapidly to meet the demands
ofprojects which require detailed spatial survey documentation
(Alessandri et al. 2005).
University of South Florida's Alliance for Integrated Spatial
Technologies use of innovative and cutting edge digital
documentation at these two circular features, demonstrates
clearly how laser scanning and HDDS can provide an afford-
able and rapid means of survey and assessment of large scale
landscapes and features. These HDDS projects illustrate how
documentation time can be reduced and how dynamic three-
dimensional models can be used to address archaeological
research questions. Remote measurement, differentiation of
depth, contour and elevational modeling, cross-section detail
and analysis, and integration of previous documentation, offer
an almost limitless potential for these rich three-dimensional
datasets. HDDS techniques are poised as method ofdocumen-
tation and data integration that allows the complex, palimpsest
of landscapes to be considered and researched today and in the
future.
Notes
All photos and images by the authors, unless otherwise stated. For
more information about three-dimensional laser scanning, contact: The
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
6 -1. 1
COLLINS, DOEIUNG, AND CARR INTEGRATED SPATIAL TECHNOLOGIES
Alliance for Integrated Spatial Technologies, University of South
Florida, Lori Collins and Travis Doering, Coordinators, 813-974-
0613.
Acknowledgments
The Authors wish to thank Chris Branas with Phillips and Jordan,
Inc. who provided technical assistance and support for these projects.
Chris Bell, former University of South Florida graduate student
assisted with the Miami Circle field work. Thanks also to Ryan J.
Wheeler for his help and direction on the Miami Circle project.
Special thanks also to Brent Weisman for his guidance and input.
References Cited
Archaeological and Historical Conservancy, Inc.
2006 www.flarchaeology.org.
Alessandri, C., M. Balzani, N. Zaltron and F. Uccelli
2005 A 3D Laser Scanner model as virtual database for integrated
analyses, September 11-14, 2005. Paper presented at the
22nd International Symposium on Automation and Robotics
in Construction ISARC 2005 Ferrara, Italy.
Associated Press
2006 Indian Burial Grounds Discovered in Miami. In
Foxnews.com, January 16, 2006.
Austin, R. J.
2004 Chipped Stone Artifacts from the Miami Circle Excavations
at Brickell Point. The Florida Anthropologist 57(1-2):85-
131.
Beriault, J. G.
2000 Map of excavation units at the Miami Circle. In: Carr,
Robert S. and John Ricisak, "Preliminary report on salvage
archaeological investigations of the Brickell Point Site
(8DA12), including the Miami Circle. The Florida Anthro-
pologist 53(4):270.
Carr, R. S. and J. Ricisak
2000 Preliminary Report on Salvage Archaeological Investiga-
tions of the Brickell Point Site (8DA12), Including the
Miami Circle. The Florida Anthropologist 53(4):260-284.
Collins, L. and R. J. Wheeler
2006 High Definition Documentation Survey ofthe Miami Circle,
Brickell Point Site, Miami, Florida. Paper presented at the
Paper presented at the Society of American Archaeologists
Conference San Juan, Puerto Rico.
Florida Division of Historical Resources
2004a Miami Circle at BrickellPoint National Historic Landmark
Nomination. Florida Department of State.
n.d. www.MiamiCircleSite.com.
Frei, E., J. K. Bukowski and R. Bukowski
2000 High-Definition Surveying (HDS): A New Era in Reality
Capture. International Archives of Photogrammetry,
Remote Sensing, and Spatial Information Sciences, Vol
XXXVI- 8/W2.
Griffin, J. W., S. B. Richardson, M. Pohl, C. D. McMurray, C. M.
Scarry, S. K. Fish, E. S. Wing, J. Loucks and M. K. Welch
1982 Excavations at the Granada Site: Archaeology and History
of the Granada Site, Vol. 1. Prepared for the City of Miami
by Florida Division of Archives, History and Records
Management, Tallahassee.
Means, G. H., and and T. Scott
2000 A Geological Assessment of the Miami Circle Site. The
Florida Anthropologist 53(4):324-326.
Merzer, M.
2003 Miami Circle archaeological find will be reburied. In The
Miami Herald.
2005 More Tequesta 'circles' found. In The Miami Herald,
February 23, page lB.
Schwartz, N.
2004 Archaeologists hurry to excavate remains of Henry Flagler
hotel in Miami. In South Florida Sun-Sentinel, April 21,
2004.
Weisman, B. R1 and L. Collins
2003 Report on the Three-dimensional laser scanning at the
Miami Circle Brickell Point Site, Miami, Florida. Ms on
file, Florida Division of Historical Resources.
Weisman, B. R, H.E. Shepard, and G. Luer
2000 The Origin and Significance of the Brickell Point Site
(8DA12), also known as the Miami Circle. The Florida
Anthropologist 53(4):342-346.
Wheeler, R. J.
2000a The Archaeology of Brickell Point and the Miami Circle.
The Florida Anthropologist 53(4):294-322.
2000b Cultural Resource Assessment of Brickell Point and the
Miami Circle. In Florida Archaeological Reports No.10.
Bureau of Archaeological Research, Division of Historical
Resources, Tallahassee.
2004a Bone Artifacts from the Miami Circle at Brickell Point
(8DA12). The Florida Anthropologist 57(1-2):133-158.
2004b Shell Artifacts from the Miami Circle at Brickell Point
(8DA12). The Florida Anthropologist 57(1-2):159-186.
Wheeler, R. J. and R. S. Carr
2004 The Miami Circle: Fieldwork, Research and Analysis. The
Florida Anthropologist 57(1-2):3-10.
Widmer, R. J.
2004 Archaeological Investigations at the Brickell Point Site,
8DA12, Operation 3. The Florida Anthropologist 57(1-
2):57.
COLuINS, DOERING, AND CARR
INTEGRATED SPATUAL TECHNOLOGIES
Florida Anthropological Society Chapters
9 7 C LA .10
o 7
14
1) Archaeological Society of Southern Florida
2495 NW 35th Ave., Miami, FL 33142
2) Gold Coast Anthropological Society
106 West Camino Real, #109, Boca Raton, FL 33432
3) Central Florida Anthropological Society .
P.O. Box 947544, Maitland, FL 32794-7544 4 w
4) Central Gulf Coast Archaeological Society 6 1
P.O. Box 82255, St. Petersburg, FL 33682
5) Indian River Anthropological Society --
3705 S. Tropical Trail, Merritt Island, FL 32952 2
6) Kissimmee Valley Archaeological and Historical Conservancy
195 Huntley Oaks Blvd., Lake Placid, FL 33852 12
7) Emerald Coast Archaeological Society
333 Persimmon Street, Freeport, FL 32435
8) Panhandle Archaeological Society at Tallahassee v
c/o The Tallahassee Trust for Historic Preservation :
423 E. Virginia Street, Tallahassee, FL 32301 ""
9) Pensacola Archaeological Society
P.O. Box 13251, Pensacola, FL 32591 13) Time Sifters Archaeology Society
P.O. Box 25883, Sarasota, FL 34277-2883
10) St. Augustine Archaeological Association
P.O. Box 1301, St. Augustine, FL 32085 14) Volusia Anthropological Society
P.O. Box 1881, Ormond Beach, FL 32175
11) Southeast Florida Archaeological Society
P.O. Box 2875, Stuart, FL 34995 15) Warm Mineral Springs Archaeological Society
P.O. Box 7797, North Port, FL 34287
12) Southwest Florida Archaeological Society
P.O. Box 9965, Naples, FL 34101
THE ANIMAL INTERMENTS AT THE MIAMI CIRCLE AT BRICKELL POINT SITE (8DA12)
ALISON A. ELGART
Department ofBiological Sciences, Florida Gulf Coast University, 10501 FGCU Blvd., Fort Myers, FL 33965
E-mail: aelgart@fgcu.edu
The 2000-year old Miami Circle (a component of 8DA12),
located on Brickell Point in Miami, derives its name from a
circular feature of holes and basins 38 ft (11.6 m) in diameter,
hewn into the bedrock by the Tequesta Indians or their ances-
tors. It has been suggested that these holes were footings for
posts for a circular structure (e.g., Widmer 2004) dating to 700
B.C to A.D. 200. Among the more unusual features found at
the site were interments of a shark (Carcharinus sp.), a dolphin
cranium (Tursiops truncatus), and a sea turtle carapace (Caretta
carreta) (for a complete faunal analysis of the site, see
Quitmyer and Kennedy 2002). Radiocarbon dating indicates
that these animal remains are from a more recent time period
(ca. A.D. 1330-1680) than the date of the Miami Circle feature
itself. Widmer's (2004) study concludes that post A.D. 200,
there is no evidence of architectural features at the site, so these
interments may postdate the habitation of the site.
Animal Use in Southern Florida
Brickell Point is located at the confluence of the Miami
River and Biscayne Bay. The bay is located in Eastern Miami-
Dade County, and is a prime habitat for marine life as it is very
shallow. Bottlenose dolphins (Tursiops truncatus) and mana-
tees (Trichechus manatus) are common, and pilot whales
(Globiocephala macrorhynca) may enter the bay on occasion.
There are seven species of "ground sharks" or carcharhinids
found in Biscayne Bay (Ault et al. 2001).
Prehistoric coastal southern Floridians relied heavily on
marine resources. Fishing was a year-round activity, particu-
larly in lagoons, while cetaceans were hunted in winter and
turtles were hunted in summer. Shark and turtle remains are
common refuse in southern Florida middens, although com-
plete, or nearly complete animals are unusual. In contrast, sea
mammals are not common refuse items, but it is possible that
most of the remains, which include heavy bones, were left on
the beach. It is likely, after examining habitat and swimming
speed, that only a few species of whales and dolphins were
hunted (Larson 1980). Bottlenose dolphins were likely hunted
by the Tequesta, as were manatees, although manatee remains
are rare in Florida sites. Marine mammals may have provided
much needed oil derived from animal fat.
Miami Circle Animal Features
Bottlenose Dolphin Cranium
The remains of a dolphin cranium (Feature 218) were found
within the Miami Circle feature, on the far eastern side,
approximately two meters west of the Circle's eastern circum-
ference (Figure 1). The fragments, composed of most of the
cranium without the rostrum and mandible, were located in
between two units, 32 and 33, in the northeast quadrant of the
Miami Circle. The remains were articulated but crushed, with
the interior of the skull oriented superiorly within Level 2, at
1.98-1.86 m (6.5- 6.1 ft) NGVD' in a soil matrix of black dirt
midden surrounded by a ring of gray-white ashy soil. This level
contained faunal bone and shell fragments as well as prehistoric
artifacts such as decorated and undecorated pottery and historic
artifacts. Three prehistoric potsherds were found within 25 cm
of the dolphin bones, but a nineteenth-century copper nail was
recovered in association with the cranial remains and a historic
iron spike was found approximately 6 cm southwest of it.
Fragments of the skull were radiocarbon dated to 69040 B.P.
(A.D. 1530-1680,2 sigma calibrated age range), indicating that
the dolphin remains date to the European Contact Period and
that the historic artifacts may have been intrusive to the level.
The bottlenose dolphin skull is about fifty percent complete
(Figures 2 and 3; see Appendix 1 for cranial details). Much of
the vault of the skull is present, but most of the rostrum and all
of the teeth are missing. The left tympanoperiotic (auditory
bulla; earbone) is present, which was used to identify the
species as a bottlenose dolphin (Tursiops truncatus) (James
Mead, personal communication, 2001). The skull is quite small
in size in comparison to a modem adult dolphin skull (Figure
3), but the prehistoric specimen does not appear to be juvenile
as its cranial sutures are closed, indicating maturity. A possible
cut mark, coronally-directed, is present on the superior aspect
of the left parietal bone. The right end of the cut is shallow,
with rounded edges, indicating either pre-mortem healing or
erosion. Many of the bones of the skull have eroded edges.
No artifacts specific to the hunting of marine mammals
(e.g., harpoons) have been found in an archaeological context
in Florida, but several cetacean remains have been found in
archaeological sites in Florida and Georgia. The faunal remains
of dolphins, likely representing food remains, are not uncom-
mon at southern Florida sites. However common dolphin refuse
items are, the dolphin cranium is a very rare find in the south-
eastern United States. In addition to the cranium, a perforated
bottlenose dolphin tooth also was found at the Miami Circle
(Wheeler 2004a:149), lending credence to Wheeler's sugges-
tion that the dolphin cranium was kept in order to extract these
teeth (Wheeler 2004b:33).
Shark Interment
The interred shark skeleton (Feature 234), consisting of
teeth, dermal denticles (from the skin), and a vertebral column
(the only ossified portions of the skeleton), was located in the
VOL. 59(3-4) THE FLORIDA ANTHROPOLOGIST SEPTEMBER-DECEMBER 2006
VOL. 59(3-4)
THIE FLORuDA ANTHROPOLOGIST
SEPTEmBER-DECEMBER 2006
300 305
320 321
nium I ,
120
IS
110
N
0 5 0et
Figure 1. Map of the Miami Circle excavation area showing locations of animal interments within units (derived from map by John G. Beriault, in Carr
and Ricisak 2000:Figure 8).
ELGART MIAMI CaCLE ANiMAL FEATURES
Figure 2. Frontal view of the reconstructed dolphin cranium (Fe
218) from the Miami Circle at Brickell Point (8DA12).
southeastern quadrant of the Miami Circle (Figures 1 and 4).
The remains were oriented east-west in units 30 and 35 at a
depth of 2.02 m (6.64 ft) NGVD, within the midden, 9 cm
below the top of the midden layer (Carr and Ricisak 2000).
The teeth representing the position of the head were located
toward the west end and the vertebrae extended eastward. A
shallow footing for a twentieth-century apartment complex lay
atop the shark remains, which seemed to be intact below it.
Morphology of the teeth indicates that the remains belong to a
requiem shark of the genus Carcharhinus (Wheeler 2004b).
The feature measured 1.3 m in length. A sand-tempered plain
pottery sherd was found directly below the vertebrae at an
elevation of 2.04 m (6.7 ft). Bone collagen from the shark
yielded a radiocarbon date of 670 30 B.P. (A.D. 1560-1680,
2 sigma calibrated age range). Nearly all coastal sites in
southern Florida contain shark teeth and vertebrae, but recover-
ing remnants of an entire shark is fairly rare.
Sea Turtle Carapace
A large marine turtle carapace (Feature 689; FS #1028),
lying dorsally, was recovered in the southwest part of Unit 64
in the northeast quadrant of the Miami Circle (Figures 1 and 5).
The rim of the carapace was found at an elevation of approxi-
mately 2.0 m (6.56 ft) NGVD (Level 1), and the carapace was
20 cm deep. The carapace was surrounded by black dirt
midden, but the southeast corner of the unit was disturbed by
former construction work. The whole carapace was removed en
bloc and was excavated at a later date by the author.
The dorsal, inverted position of the carapace and the
contents within the carapace indicate that it served as some kind
of container or repository. The carapace contained an unstrati-
fied deposit consisting of ash, vertebral and limb bones from
other turtles, rusted square iron nails, and rusted metal
fragments. Some nails and a rusted metal sheet were
adhering to the bone. A fragment of clear glass also
was present. Although the presence of the ash and the
turtle bone suggests that the carapace may have been
used as a cooking hearth, the shell itself is not charred.
The fact that historic artifacts were attached to the bone
in the deepest part of the deposit (the apex of the
carapace) indicates that they were not intrusive. A
radiocarbon date of 420 80 B.P. (A.D. 1330-1650, 2
sigma calibrated age range) dated charcoal recovered
from the deposits within the carapace. However, this
does not actually date the carapace itself, nor does it
accurately date the deposits within the carapace, which
definitely date to the recent past. The age of the
carapace remains in question.
The turtle carapace has been tentatively identified
as belonging to a sub-adult loggerhead sea turtle
(Caretta caretta) (Wheeler 2004b:32). The carapace
measures 65 cm in length and 50 cm in width, which is
small considering that adult female loggerheads nesting
in Broward County measured 97-100 cm in length with
nature an average of 68 cm in width (Fletemeyer 1984:32 in
Wheeler 2004b). Adult loggerheads average about 136
km in weight, and are exceeded in size by leatherback
sea turtles and green turtles (Ernst and Barbour 1972:231). All
of the sea turtles are edible, and the loggerhead is exploited for
its food value (Ernst and Barbour 1972:238). In historic times,
as in the present, the closest beach to the Miami Circle where
loggerheads would have nested was Key Biscayne (Florida
Marine Research Institute 2002).
Another possible interment at the Miami Circle site was
uncovered in Trench 11, located northwest of the Miami Circle
feature. It consists of the articulated lower portion of a raccoon
(Procyon lotor), and was only partially excavated since the
remainder of the remains extended into the trench wall
(Wheeler 2000a:305). Because it remains to be seen whether
this is an interment, it will not be dealt with in this paper.
Discussion
Archaeological Perspectives on Animal Burials
For animal remains to be classified as an interment, a
complete or partial body of an animal must demonstrate
intentional burial in an archaeological context and be devoid of
any type of modification that would suggest a utilitarian role.
An animal burial will be in anatomical position, unlike the
faunal bone found in middens, which is randomly deposited,
with disparate elements of different species commingled. Such
mortuary treatment of remains can include burial within a pit
specifically prepared for the body of the animal, or burial in a
special location, such as a structure or a human grave. In North
America, animal burials date from as early as 8,500 B.P. at the
Koster site in the Illinois River Valley (Morey and Wiant
1992). Dog burials are the most common type of burial in
archaeological sites, and are more often buried with humans
than any other taxon. Canid burials have been found world-
ELGART
MIAMI CIRCLE ANIMAL FEATURES
THE FLORIDA ANTHROPOLOGIST 2006 VOL. 59(3-4)
Figure 3. Comparison of the reconstructed prehistoric dolphin cranium with a modern dolphin skull, lateral view.
wide (e.g., Haag and Heizer 1953; Lyon 1996:100; McMillan
1970; Mellink 1966; Morey and Wiant 1992) and were com
mon by the Archaic period in the eastern and midwestern U.S.
(McMillan 1970). It is likely that dogs were afforded special
mortuary treatment because of their closeness with humans and
high status as pets. It is also possible that in some instances,
they served as substitutes for human sacrifice (Hill 2000:386-
387).
Hill (2000) studied 164 interments of birds, canids, and
bears dating from A.D. 200-1450 in the southwestern U.S. and
determined that animal burials fell into three categories:
1)"animal sacrifice and disposal as ceremonial trash;"
2)"dedicatory interment," as an offering for special occasions
(e.g., building a structure); and 3) "simple interment or expedi-
ent disposal," for all other interments (Hill 2000:363-364).
Walker (1995) applied the term "ceremonial trash" to worn-
out ritual objects found in an archaeological context. These
objects were disposed of in a manner suitable to sacred objects,
a practice which appears to be cross-cultural and continues
today. Many animal burials would be classified under Hill's
first category as ceremonial trash or animal sacrifice. These
interments show intentional killing before disposal and are
recovered either in association with human burials (Haag and
Heizer 1953; Heizer and Hewes 1940; Hill 2000; Mellink
1966) or from other sacred locations outside habitation areas
(Hill 2000). After the animals are used for a ceremonial
purpose, they are ritually disposed of in sacred ground (hence,
"ceremonial trash"). For example, Hill studied 459 avian
interments and concluded that they were chiefly ceremonial
trash. She argues that the birds were sacrificed for their
feathers and were buried because their use-life had ended (Hill
2000:378). Other animal burials are associated with human
burials and may constitute substitutes for human sacrifices. For
example, at sites located in central California dating to the Late
Culture, a fully articulated bear skeleton (Ursus americanus)
was found at the west edge of a cemetery area of site C.138, a
badger (Taxidea taxus neglecta) was found within a few inches
of a flexed human burial at site S.60, and several coyotes
(Canis latrans) were found near or within human burials
(Heizer and Hewes 1940). These interments are regarded as
ceremonial, particularly because many of the animals were
transported long distances and they are in association with a
sacred site, which is a burial or a cemetery in this case.
Ethnographic accounts reveal that central California tribes kept
coyotes, deer, and bear as captive pets and may have had
emotional attachments to them.
Florida Animal Interments
Most animal burials recorded in Florida also may be
categorized as ceremonial trash. As is true elsewhere, dog
interments are the most common type of animal interment in
Florida (Table 1), and dogs are thought to have been kept as
pets by Florida tribes (Wheeler 1992b). This contrasts with
what Hill (2000) found in the Southwest: out of 90 canid
interments, which contained 115 individuals, she concluded that
most were dedicatory offerings, found in pits or under the floors
ofkivas or other structures (Hill 2000:379, 386).
North of Lake Okeechobee, animal interments, mostly dog
burials, are known from several sites. For instance, at the
Lighthouse Mound (8NA3) on Amelia Island, Moore (1896, in
Mitchem 1999:38) recovered a dog skeleton in the burial
mound with many human bones. At the Hutchinson Island site
(8MT37) in Martin County, seven dog burials were found
within a burial mound (Carr and Steele 1993:12).
There are six sites in southern Florida with animal burials
(Table 1). Numerous animal burials were reported from the
Palmer Burial Mound (8SO2) at the Palmer site complex near
Sarasota, Florida (Bullen and Bullen 1976). Four dog burials
and a"ceremonial" alligator interment were found among about
400 human burials. Three of the dog burials were in close
proximity to one another and were located three to seven feet
northwest of the alligator burial. They were laid in shallow pits
on their side, and one burial contained two dogs. The alligator
was laid in a shallow pit above two human burials and con-
tained several grave goods. Two strings of sawfish vertebrae
2006 VOL. 59(3-4)
TTtE FUoRiDA ANTHROPOiOGIST
Table 1. Prehistoric sites in Florida with recorded animal interments (entire animals only).
Site Date Interment description Source
Broward County
Margate-Blount (8BD41) Glades I-III Alligator, snakes, raccoons, turtle in Felmley 1991
"ceremonial locus" with ornamental
artifacts
Martin County
Hutchinson Island Site Glades III 7 dogs within burial mound Carr and Steele
(8MT37) 1993
Miami-Dade County
Granada site (8DA11) Glades I-III one dog in association with secondary Bob Carr, Personal
human burials communication,
2005
Surfside Mound (8DA22) Glades III indeterminate animal interments next to Lamme n.d. in
mound Felmley 1991
Trail Site (8DA33) Glades I-III Component B-3 whole turtle shells; Coleman 1989;
north side of Tamiami Trail: dog burial Elgart-Berry and
with scattered human remains Rombola 2003
Nassau County
Lighthouse Mound (8NA3) St. Johns Ib Dog in burial mound with many human Moore 1896 in
and II (post bones Mitchem, 1999
A.D. 500)
Palm Beach County
Riviera Complex (8PB30) Glades III Entire dog and shark skeleton in midden Wheeler 1992b
Sarasota County
Palmer Site (8SO2) Glades III Four dog burials and an alligator in Bullen and Bullen
association with about 400 human burials 1976
beads were found parallel to the body of the alligator. The
interment dates to a Late Glades period, probably earlier than
A.D. 1300. The burial may represent a totemic relationship,
such as an "alligator clan" to which the people belonged
(Bullen and Bullen 1976).
There is at least some evidence that the Tequesta or their
ancestors ceremonially interred animals. In Miami-Dade
County, across the river from the Miami Circle, one prehistoric
dog burial in association with numerous secondary human
burials was reported from recent excavations at 8DA11 (Bob
Carr, personal communication, January, 2005). Farther west in
the Everglades, the Trail site (8DA33) contained a dog burial
and possible turtle burials. The dog burial was located on the
west side of the site among scattered human bones (Elgart-
Berry and Rombola 2003). In Component A (a portion of the
site apart from Component B) of the Trail site, an alligator skull
was found on top of apelvis of a partial human burial, and three
"whole turtle shells" were found in Component B (Coleman
1989:258, Figure 3). At the Margate-Blount site in Broward
County (directly north of Miami-Dade County), interments of
an alligator, snakes, raccoons, and turtles were uncovered in a
"ceremonial locus" separate from the habitation and mortuary
areas (Felmley 1991:68; Wheeler 1992a:95).
The other sites where animal burials were identified in
southern Florida contain "simple interments" (Table 1) as the
burials do not seem to be ceremonial or dedicatory interments.
In Palm Beach County, a dog and shark were interred within a
midden context at the Riviera Complex (8PB30) (Wheeler
1992b:6).
Thus far, no dedicatory interments have been recorded at
Florida sites. The three animal interments uncovered at the
Miami Circle are undoubtedly unusual, but do they signify
ritual burial by the Tequesta?
Bottlenose Dolphin Cranium. Dolphin post-cranial bones
within refuse are not uncommon in Florida sites. Common
dolphin (Delphinus delphis L.) bones were found at the Glades
Ill-period Surfside site in Miami-Dade County (Willey
1949:79-83), which is roughly contemporaneous with the
Miami Circle dolphin cranium. However, not much is known
about this find, because the dolphin bones were reported with
the other faunal remains, and no further information on them is
presented. Atlantic bottlenose dolphin bones were found at
Green Mound in Volusia County, which is located eight miles
south-southeast of Daytona Beach (Bullen and Sleight
1960:31). The site dates to the St. Johns II to III periods, and
the remains, of which there were few, were found in a midden
ELGART
MLAMw CIRmL ANmMAL FFATuREs
THE FLORIDA ANTHROPOLOGIST
Figure 4. The shark (Carcharhinus sp.) interment (Feature 234) from the Miami Circle at Brickell Point (8DA12).
context. Atlantic bottlenose dolphin bones also were found at
the Garden Patch site in Dixie County (Kohler 1975:77) and at
Jupiter Inlet 1 (8PB34) (Wheeler et al. 2002). Wheeler and
colleagues (2002:181-182) report that bottlenose dolphin and
baleen whale auditory bullae were found within midden
material at Jupiter Inlet 1. Occupation of the site dates between
A.D. 750 and A.D. 1200, which is transitional between Glades
II and III periods.
There are only two other sites in the southeastern U.S.
where dolphin cranial remains may have been found, but these
cannot be verified. The first was at the Marineland site, located
near St. Augustine, which was excavated by State Archaeologist
Vernon Lamme. In a letter to Smithsonian Institution archaeol-
ogist Matthew Stirling, Lamme (1940) states, "both turtle and
the skull of a porpoise [read: dolphin] have been unearthed
buried as the humans with oyster shell around the body and
clam shells around the skulls." Unfortunately, this is the only
reference to a "porpoise skull" in Lamme's writings on Marine-
land. In his other writings on Marineland, his only mention of
"porpoise" bones was within a midden context (Lamme 1939,
1940, 1941a:7, 1941b, 1973). The second mention of remains
of a dolphin skull comes from Florida artist Hermann
Trappman (Ryan Wheeler, personal communication, December,
2003), who stated that pieces of the upper and lower jaws of a
dolphin were found at the Tierra Verde burial mound near St.
Petersburg. At the Granada site (8DA 11), located on the north
bank of the Miami River directly across from the Miami Circle,
two bottlenose dolphin elements, a tooth and an atlas, were
found (Wing and Loucks 1982:310).
Although dolphin meat was likely consumed by the
Tequesta, the completeness ofthe dolphin cranium suggests that
it was not from a discarded carcass, but rather was intentionally
buried. In addition, it is doubtful that the dolphin head was a
source of meat, which indicates that it served some other
purpose. It may have been kept in part for its valuable teeth,
and one dolphin tooth pendant was recovered from the Miami
Circle site (Wheeler 2004a: 149). Ethnographic evidence from
Tequesta sites and material evidence from Calusa sites suggests
that dolphins were revered in Tequesta culture.
Dolphins in Tequesta culture. Unfortunately, few ethnographic
accounts exist on the Tequesta. Although the burial of animals
was not witnessed during the European Contact Period, some
historical accounts do mention the ceremonial use of animals.
In 1743, the Jesuits Joseph Maria Monaco and Joseph Xavier
de Alafia proposed Christianizing the "Indians of the south
Florida Keys," and a small earthen fort was constructed in
Miami for their protection (Childers 2003; Sturtevant 1978).
Describing the belief system of the Tequesta, the Jesuits related
that the dead were feared and were placed some distance from
the village, and "so that they do not do them evil, they place the
skulls of stags, turtles, barracuda, and other animals around (the
cemetery)" (Childers 2003:77). A known prehistoric mortuary
area, Brickell Park, lies approximately 500 ft south of Brickell
Point, and a burial mound, Miami Sand Mound #4 (8DA13)
was located about 200 ft south of the park (Elgart and Carr, in
this issue). Prehistoric burials have been recovered in the park,
which were assigned the time frame "Glades Period" (Carr et al.
2001).
It is possible that the dolphin skull was used for the purpose
of warding off evil spirits from the nearby cemetery and then
later ritually interred as trash at the Miami Circle. Both Robert
Carr and Randolph Widmer (Randolph Widmer, personal
communication, April, 2003) suggest that after A.D. 500, the
2006 VOL. 59(3-4)
ELGART MIAMi CIRCLE AIIUMAI. FEATURES
Figure 5. The sea turtle (Caretta caretta) carapace (Feature 61
#1028) post excavation en bloc, ventral side of carapace in view.
village center moved from Brickell Point to 8DA11 located
across the river, but that the Miami Circle site continued to
serve some ceremonial purpose. Ethnographic accounts (Harm
1991:314), recent excavations at 8DA 11, and analysis of Miami
Circle artifacts seem to confirm this movement to the north side
of the river for habitation. Therefore, an alternative hypothesis
may be posited based on the proposed sanctity of the Miami
Circle area during the later periods: the Tequesta may have
ritually buried their ceremonial trash in this "hallowed ground".
No ethnographic accounts of the Tequesta specifically
mention dolphins, but Lopez de Velasco, whose writings in
1569 were based on a review of other accounts, does mention
the capture of a marine mammal and an associated mortuary
ritual. In Hann's (1991:319) translation of de Velasco, he
writes that after the Tequesta captured a whale, "the first thing
that they do to it is they open the head and extract two bones
that it has in the skull and they throw these two bones in the
chest (caja) in which they place the deceased [chief] and this
they adore." In Swanton's (1946:722) translation of the same
passage, he determines that it is not a whale in the ritual, but a
manatee. Larson (1980) speculates that the whale may actually
be a pilot whale, which is a member of the dolphin family
Delphinidae. He argues that the two bones that were removed
from the skull were the tympanic bones (the auditory bullae),
which resemble small human heads. Because the auditory
bullae were recovered with the Miami Circle dolphin cranium,
I do not believe that it was used in this type of mortuary ritual,
as some have suggested.
Alternatively, the dolphin skull may have been used for
medicinal purposes. The medicine kit of the shaman recovered
from Key Marco contained the skulls of opossum and
weasels (Cushing 1896). However, these skulls are
obviously much smaller than a dolphin skull and
there is no further evidence of skulls being used in
this manner.
Animal imagery is a common artistic theme found
in Glades-area (Stirling 1936) artifacts dating to the
European Contact era, and dolphins were occasion-
ally depicted. This "terminal Glades complex"
(Wheeler 2000b) temporally extended from approxi-
mately the mid sixteenth to the mid eighteenth cen-
tury and involves the incorporation of European
material such as metal, clothing, and glass beads into
indigenous artifacts. According to radiocarbon dates,
the animal burials from the Miami Circle probably
dated to this time period. The animal iconography of
the terminal Glades complex falls into two general
categories: one, for use in public places, such as in
mortuary ponds, ceremonies, and dances and the
other for personal use. The former includes large
wooden carvings, woodenmasks, and plaques and the
latter includes small effigies in the form of pendants,
but both categories depict the same animals. In-
stances of dolphin imagery in Calusa artifacts from
the Terminal Glades complex include a wooden tablet
89; F.S. in the form of a duck or roseate spoonbill with an
incised dolphin at the top, and a square turtle plastron
piece incised with two wheeling dolphins, both from
the Key Marco site (Gilliland 1975:80; Wheeler
2000b:148-149). Cushing (1896:377) considers the latter
artifact to be ceremonial, and states that it was "used in sacred
games, or perhaps in processes of divination."
The Shark Skeleton. Shark remains are common in southern
Florida sites, but the only other example of an entire shark in an
archaeological context is the carcass of a "large shark" repre-
sented by 20 articulated vertebrae that was recovered within
food refuse at the Riviera Complex (Wheeler 1992b:6).
Unfortunately, this is all of the information available for the
Riviera shark, but does raise the question of whether both the
Riviera and the Miami Circle sharks should be considered
burials, or whether they are just refuse. Across the Miami River
from the Miami Circle at the Granada site (8DA 1), shark and
ray remains were abundant, comprising 53.6% of the faunal
meat biomass (Wing and Loucks 1982:324-327). Similarly,
many shark remains were discovered at the Palmer site (8SO2)
near Osprey, Florida, including articulated sections of shark
centra and burned denticles. Kozuch (1998:183-185) concludes
that sharks were caught by hooks, lines, and weirs by prehis-
toric Floridians and that most of the carcass was brought back
to the habitation site to be butchered. Many parts of the shark
would have been used, including the skin, which would have
been dried for use as shagreen ("sandpaper"), the liver, which
would have been boiled for the oil, and teeth, which were vital
for tool-making in the absence of stone. Larson (1980:99) adds
that catching sharks offshore must have been very time consum-
ing and labor intensive, but sharks were invaluable for their
parts, and so worth the time and energy involved.
ELGART
MLIAw CIRCLE ANIMAL FFATuREs
THE ~ ~ ~ ~ ~ __ FLlIAATRPLGIT20 O 934
The majority of bone artifacts found at the Miami Circle
were shark-derived (Wheeler 2004a:134-135). The 103
perforated shark or ray centra and 131 modified shark teeth
recovered are predominantly from several species of requiem
sharks, as well as from lemon sharks (Negaprion brevirostris),
tiger sharks (Galeocerdo cuvieri), and sand-tiger sharks
(Odontaspis taurus). These two types of artifacts are common
in southern Florida sites, and represent practical uses for shark
remains: perforated centra may have been used as beads or
earspools (Richardson and Pohl 1982:93), and shark teeth were
probably hafted and used as tools and weapons (Wheeler
2004a:137).
The shark skeleton recovered from the Miami Circle
constitutes an animal burial, but it is difficult to determine
whether it falls under the animal sacrifice/ceremonial trash or
simple interment category. The fact that whole shark carcasses
were transported to sites in order to utilize many parts and the
occurrence of shark carcasses at other sites that are not "inter-
ments" supports the "simple interment" categorization. Further-
more, the shark was interred in midden, not in any prepared pit,
and is not associated with any obvious structures.
However, there is evidence that sharks were revered and
highly valued in Tequesta culture and this shark was not utilized
for its parts (evident by the remains present), which would give
credence to categorizing it as an animal sacrifice or ceremonial
trash. Widmer (2004:52-53) recovered a "dedication cache"
consisting of a shark steak (represented by six vertebrae) placed
inside of a ceramic vessel at the Miami Circle. According to
Widmer, the vessel was placed at the base of a posthole prior to
the insertion of a post and represents ritual behavior surround-
ing sharks. Shark teeth were also used as adornments as well as
tools and were placed within human burials in both Calusa and
Tequesta cultures during the terminal Glades complex. Grave
goods at many southern Florida sites have included single-
drilled and double-drilled shark teeth of both extant and fossil
sharks (Carr et al. n.d.; Wheeler 2004a). Metal pendants in the
form of shark teeth, constructed from either sheet silver or
larger pieces of silver, have been found at several sites in
Glades, Palm Beach, Osceola, and Charlotte Counties (Wheeler
2000b and references therein). At the Granada site (8DA11),
located on the north side of the Miami River, a bone pin was
recovered, the head of which consists of an effigy of shark
vertebrae (Wheeler 1996:144). No human burials were
recovered from the Miami Circle site (Elgart and Carr, this
issue), but if the area was considered sacred or used ceremoni-
ally by the Tequesta, as Widmer and Carr suggest, then perhaps
the shark was ceremonially trashed here.
Sea Turtle Carapace. The sea turtle carapace does not appear
to be an animal burial. The carapace contains a single, thick
layer of ash, rusted metal, glass, and turtle bone, but the bone
itself was not charred, indicating that it was not used for
cooking nor was it burned. The rusted square nails and rusted
pieces of metal were not intrusive, they were found throughout
the deposit, and some of this metal was still adhering to the
shell. The radiocarbon date of420 + 80 B.P. (A.D. 1330-1650,
2 sigma calibrated age range) from charcoal within the ash
indicates that it is possible that the carapace served some ritual
or utilitarian purpose during the Late Glades period, but it is
more likely that the carapace never served any purpose until
historic times, when it was used as a receptacle. Either way, it
was not buried until after nails were deposited into it during the
historic period.
Sea turtle bone served different purposes in prehistoric
American cultures. At El Meco, Mexico, a small, ceremonial-
habitational site dating to the early Classic period, turtle
carapaces were used as containers (Andrews 1986:69).
Hawksbill tortoiseshell, which is composed of the epidermal
scutes that overlie the bony carapace, was used for hairpins and
ornaments in the southeastern U.S. (Frazier 2003:19).
Marine turtle remains are very common in Florida and
Caribbean sites (Frazier 2003), and most likely represent food
refuse. At the Granada site, remains of 63 sea turtles were
found (Wing and Loucks 1982:277), and at the Riviera Com-
plex, a large portion of a turtle was uncovered in the cooking
area (Wheeler 1992b:5). Several turtle interments have been
found as well. Two sea turtle crania, one of which was from a
loggerhead, were found within a human burial at the Palm
Beach 3 Burial Mound (8PB26) (Johnson 1952:36), and parts
of a loggerhead turtle were found at the Santa Lucea site
(8MT37) in St. Lucie County (Carr and Steele 1993:14). Atthe
Golden Rock site on St. Eustatius in the Caribbean, the skeleton
of a hawksbill turtle (Eretmochelys imbricata) was found
upside down, similar to the orientation of the carapace found at
the Miami Circle, but it was much more complete (Van der
Klift 1992:74).
Conclusions
Interments of a dolphin cranium, a shark, and a sea turtle
carapace recovered at site 8DA12 were discussed. The
radiocarbon dates for the dolphin and shark interments place
them within the late Glades III to European Contact periods.
The date of the sea turtle carapace is uncertain, but associated
artifacts attest to its use during the historic period.
Stratigraphically, the interments were located in the upper
levels of the site, which is a disturbed level, containing both
prehistoric and historic deposits. The interments post-date the
extensive midden deposition at the site. Analysis of the Miami
Circle site artifacts and ethnographic evidence suggest that the
Tequesta moved the center of their village to the north side of
the river (to 8DA11) by the Glades II-I period. However,
after the Miami Circle site no longer served as a focal point for
habitation, the Brickell Point area may still have been used for
ceremonial purposes.
The capture of dolphins and sharks required advanced
technology, and was most likely a costly but valuable pursuit.
These animals were apparently prized for their teeth, which
were worn ornamentally, and used for tools as well.
Ethnographic evidence from Tequesta sites, and material
evidence from Calusa sites suggests that dolphins and sharks
were revered by the Tequesta.
The animal interments were assessed using Hill's (2000)
categorization of animal burials and were compared to other
interments in North America and Florida. The survey of animal
interments in Florida and the U.S. described here reveals that
many of the Florida interments, although not all, fall into Hill's
ceremonial trash and animal sacrifice category because they
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
ELGART MIAMI CIRCLE M~1MAL FEATURES
were used ceremonially and were in association with or were
near ceremonial areas.
The dolphin cranium and shark skeleton from the Miami
Circle are unlike many of the recorded animal burials in
Florida, which are associated with human burials, although,
they qualify as ceremonial trash. Before being buried, the
dolphin skull may have been kept for extracting teeth, which are
all missing from the cranium (Wheeler 2004b:33), or for
warding off evil spirits from the burial grounds located immedi-
ately south of the Miami Circle site.
The shark interment is tentatively termed an animal sacrifice
or ceremonial trash only on tenuous evidence of ritual use of
sharks within Tequesta and Calusa culture. It was found within
disturbed midden and is similar to the shark interment uncov-
ered at the Riviera Complex in Palm Beach County, and there
is evidence of ritual activity with shark parts at the Miami
Circle site (Widmer 2004). Often, whole sharks were brought
to living areas because so much of the shark was used. It is
likely that this shark's meat was not consumed because the skin,
teeth and vertebrae remain, and nothing was extracted for use
as tools, jewelry, or weapons.
The sea turtle carapace was determined not to be an
intentional animal interment, but instead it is postulated that it
served as a receptacle that was used during the historic period.
This assessment is based on the presence of historic artifacts
adhering to the carapace.
None of the animal interments recovered from the Miami
Circle were considered dedicatory interments. This is because
they date to over 1000 years after the main construction of the
Miami Circle feature, post-date the main occupation period of
the site, and were not found in prepared pits or in association
with any structures.
Two of the three animal interments described here were
deemed to be associated with ceremonialism. It is possible that
they were interred as ritual trash at the Miami Circle site
because it was considered sacred ground to the Tequesta during
the Glades III Period, after the main period of habitation.
Animal interments may have been present at other Tequesta
sites, but they were not recognized as such in the past. In the
future, it is vital to treat any animal found whole or nearly
whole in an archaeological context as a potential animal burial.
Notes
All elevations were measured in feet and tenths of feet, using the
National Geodetic Vertical Datum (NGVD) of 1929 as a datum.
Acknowledgments
Thanks to James Mead of the Smithsonian Institution, who first
identified the species of the dolphin. Blair Mays of the Southeast
Fisheries Science Center (National Oceanic and Atmospheric
Administration) graciously lent me a dolphin skull. Jorge Zamanillo
from the Historical Museum of Southern Florida helped with the
photographs and allowed me to examine the dolphin and turtle
numerous times. Ryan Wheeler, William Straight, and Jim Clupper of
the Monroe County Library helped a great deal with background
research. James Pepe provided useful comments on the manuscript.
Finally, the Archaeological and Historical Conservancy allowed me to
reconstruct and excavate, where needed, the interments themselves.
References Cited
Andrews, A.P.
1986 La fauno arqueol6gico de El Meco. In Excavaciones
Arqueol6gicas en El Meco, Quintana Roo, 1977, edited by
A.P. Andrews and F. Robles C., pp. 67-75. Institute
Nacional de Antropologia e Historia. Serie Arqueologia.
Ault, Jerald S., Steven G. Smith, Geoffrey A. Meester, Jiangang Luo,
and James A. Bohnsack
2001 Site characterization for Biscayne National Park: Assess-
ment of fisheries resources and habits. NOAA Technical
MemoNMFS-SEFSC-468. U.S. Department ofCommerce,
National Oceanic and Atmospheric Administration. Miami,
Florida.
Bullen, Ripley P. and Adelaide K. Bullen
1976 The Palmer Site. Florida Anthropological Society Publica-
tion Number 8, Gainesville.
Bullen, Ripley P., and Frederick W. Sleight
1960 Archeological investigations of Green Mound, Florida.
The William L. Bryant Foundation American Studies
Report #2. Orlando.
Carr, Robert S., JeffRansom, and Alison Elgart-Berry
n.d. Archaeological Investigations and Monitoring ofthe Long
Lakes Estates Parcel Broward County, Florida. Technical
Report of the Archaeological and Historical Conservancy
(in preparation). Davie, Florida.
Carr, Robert S., Jeff Ransom, Mark Lance, and Alison Elgart-Berry
2001 A Due Diligence Archaeological Assessment ofBrickell
Park, Miami, Florida. TechnicalReportoftheArchaeolog-
ical and Historical Conservancy #312. Davie, Florida.
Carr, Robert S., and John Ricisak
2000 Preliminary report on salvage archaeological investigations
of the Brickell Point Site (8DA12), including the Miami
Circle. The Florida Anthropologist 53(4):260-285.
Carr, Robert S., and Willard Steele
1993 An Archaeological Assessment of the Hutchinson Island
Site (8MT37), Martin County, Florida. Technical Report of
the Archaeological and Historical Conservancy #68. Davie,
Florida.
Childers, R. Wayne
2003 Historical Notes and Documents: Life in Miami and the
Keys: Two reports and a map from the Monaco-Alafia
Mission, 1743. Florida Historical Quarterly Vol. 82 (1):59-
81.
Coleman, Wesley F.
1989 Salvage excavations at the Trail Site, Dade County, Florida.
The Florida Anthropologist 42(3):257-262.
Cushing, Frank Hamilton
1896 Exploration ofAncient Key Dwellers'Remains on the Gulf
Coast ofFlorida. Proceedings of the American Philosophi-
cal Society 35.
Elgart, Alison A., and Robert S. Canr
2006 An analysis of the prehistoric human remains found at the
Miami Circle at Brickell Point site (8DA12). The Florida
ELGART
MuMIA CIRCLE ANmMAL FEATu~REs
THE FLORIDA ANThROPOLOGIST 2006 VOL 59(34)
Anthropologist 59(3-4):241-249 (this issue).
Elgart-Berry, Alison, and William Rombola,
2003 Archaeological Monitoring of the C-4 Canal Expansion,
Miami-Dade County, Florida. Technical Report of the
Archaeological and Historical Conservancy #44. Davie,
Florida.
Ernst, Carl H., and Roger W. Barbour
1972 Turtles ofthe United States. University Press of Kentucky,
Lexington, Kentucky.
Felmley, Amy
1991 Prehistoric Mortuary Practices in the Everglades Cultural
Area, Florida. Unpublished Master's thesis, Department of
Anthropology, Florida Atlantic University, Boca Raton.
Florida Marine Research Institute
2002 Reported Nesting Activity of the Loggerhead Turtle
(Caretta caretta) in Florida, 1993-2001. Florida Marine
Research Institute, Florida Fish and Wildlife Conservation
Commission, St. Petersburg, Florida.
Frazier, Jack
2003 Prehistoric and ancient historic interaction between humans
and marine turtles. In The Biology of Sea Turtles, Vol. II,
edited by Peter Lutz, John Musick and Jeanette Wyneken,
pp. 1-38. CRC Press, Boca Raton.
Gilliland, Marion S.
1975 The Material Culture of Key Marco, Florida. University
Presses of Florida, Gainesville.
Haag, William G., and Robert F. Heizer
1953 A dog burial from the Sacramento Valley. American
Antiquity 18(3):263-265.
Hann, John H.
1991 Missions to the Calusa. University Presses of Florida,
Gainesville.
Heizer, Robert F., and Gordon W. Hewes
1940 Animal ceremonialism in central California in the light of
archaeology. American Anthropologist 42(4):587-603.
Hill, Erica
2000 The contextual analysis of animal interments and ritual
practice in southwest North America. Kiva 65(4):361-398.
Iablokov, A.V., V.M. Belkovich, and V.I. Borisov
1974 Whales and Dolphins Part I Joint Publications Research
Service, Arlington, VA.
Johnson, William R.
1952 Lepidochelys kempii and Caretta c. caretta from a south
Florida Indian Mound. Herpetologica 8:36.
Kohler, Tim A.
1975 The Garden Patch Site: A Minor Weeden Island Ceremonial
Center on the North PeninsularFlorida GulfCoast. Unpub-
lished Master's thesis, University of Florida, Gaines-
ville.
Kozuch, Laura
1998 Faunal remains from the Palmer Site (8SO2) with a focus on
shark remains. The FloridaAnthropologist 51(4):177-192.
Lamme, Vernon
1939 Ancient Skeleton is Found in Mound near Marineland. St.
Augustine Record, November 19, 1939.
1940 Indian Burials at Marineland. Daytona Beach Observer,
September 7, 1940.
1941a Archaeological Investigation at Marineland, Florida. On
file, Florida Historical Society, Alma Clyde Field Library,
Melbourne.
1941b Dead Men Do Tell Tales. Florida Highways, October,
1941.
1973 Florida Lore Not Found in the History Books! Star Publi-
cations, Boynton Beach.
Larson, L.H.
1980 Aboriginal Subsistence Technology on the Southeastern
Coastal Plain during the Late Prehistoric Period. Univer-
sity Presses of Florida, Gainesville.
Lyon, Edwin A.
1996 A New Dealfor Southeastern Archaeology. University of
Alabama Press, Tuscaloosa.
McMillan, R. Bruce
1970 Early canid burial from the Western Ozark Highlands.
Science (new series) 167(3922):1246-1247.
Mellink, Machteld J.
1966 Archaeology in Asia Minor: Addenda. American Journal
ofArchaeology 70(3):279-282.
Mitchem, Jeffrey M.
1999 The East Florida Expeditions of Clarence Bloomfield
Moore. University of Alabama Press, Tuscaloosa.
Morey, Darcy E., and Michael D. Wiant
1992 Early Holocene domestic dog burials from the North
American Midwest. CurrentAnthropology 33(2):224-229.
Quitmyer, Irvy R., and Erin E. Kennedy
2002 The Zooarchaeology of the Miami Circle, Brickell Point
(8DA12), Florida. EnvironmentalArchaeology Laboratory,
Department of Natural History, Florida Museum of Natural
History, Gainesville.
Richardson, Sue B., and Mary Pohl
1982 The Bone Tool Industry from the Granada Site. In Excava-
tions at the Granada Site: Archaeology and History of the
Granada Site, Volume 1, edited by John W. Griffin, pp. 83-
170. Florida Division of Archives, History and Records
Management, Tallahassee.
Sturtevant, William C.
1978 The last of the South Florida Aborigines. In Tacachale:
Essays on the Indians ofFlorida and Southeastern Georgia
during the Historic Period, edited by Jerald Milanich and
Samuel Proctor, pp. 141-162. University of Florida Presses,
Gainesville.
Stirling, Matthew W.
2006 VOL. 59(3-4)
THE LORIDA ANTHROPOLOGIST
ELGART MIAMI CIRCLE ANIMAL FEATURES
1936 Florida Cultural Affiliations in Relationship to Adjacent
Areas. In Essays in Anthropology in Honor ofAlfred Louis
Kroeber, pp. 351-357. University of California Press,
Berkeley.
Swanton, John R.
1946 The Indians of the Southeast U.S. Smithsonian Institution
Bureau of American Ethnology Bulletin 137.
Van der Klift, Heleen M.
1992 Faunal remains of Golden Rock. In The Archaeology ofSt.
Eustatius: the Golden Rock Site, edited by Aad H. Versteeg
and Kees Schinkel, pp. 74-83. Publication No. 2, St.
Eustatius Historical Foundation, St. Eustatius and Publica-
tion No. 131, Foundation for Scientific Research in the
Caribbean Region, Amsterdam.
Walker, W.H.
1995 Ceremonial Trash? In Expanding Archaeology, edited by
James M. Skibo, W.H. Walker, and A.E. Nielsen, pp. 67-79.
University of Utah Press, Salt Lake City.
Wheeler, Ryan J.
1992a Time, Space and Aesthetics: Decorated Bone Artifacts from
Florida. Unpublished Master's thesis, Department of
Anthropology, University of Florida, Gainesville.
1992b The Riviera Complex: an East Okeechobee archaeological
area settlement. The Florida Anthropologist 45(1):5-17.
1996 Ancient Art of the Florida Peninsula: 500 B.C. toA.D.
1763. Ph.D. dissertation, Department of Anthropology,
University of Florida, Gainesville.
2000a The Archaeology of Brickell Point and the Miami Circle.
The Florida Anthropologist 53(4):294-322.
2000b Treasure ofthe Calusa. Monographs in Florida Archaeol-
ogy Number 1. Rose Printing, Inc., Tallahassee.
2004a Bone artifacts from the Miami Circle at Brickell Point
(8DA12). The Florida Anthropologist 57(1-2):133-158.
2004b Miami Circle atBrickell Point: National Historic Landmark
Nomination. Prepared by the Florida Division of Historical
Resources, Florida Department of State, Tallahassee.
Wheeler, Ryan J., and Robert S. Carr
2004 The Miami Circle: Fieldwork, Research and Analysis. The
Florida Anthropologist 57 (1-2):3-10.
Wheeler, Ryan J., James P. Pepe, and William J. Kennedy
2002 The Archaeology of Jupiter Inlet 1 (8PB34). The Florida
Anthropologist 55(3-4):157-191.
Widmer, Randolph J.
2004 Archaeological investigations at the Brickell Point Site,
8DA12, Operation 3. The Florida Anthropologist 57 (1-
2):11-58.
Willey, Gordon R
1949 ExcavationsinSoutheastFlorida. Yale University Publica-
tions in Anthropology #42, New Haven, Connecticut.
1982 Granada Site faunal analysis. In Excavationsatthe Granada
Site: Archaeology and History ofthe Granada Site, Volume
1, edited by John W. Griffin, pp. 259-345. Florida Division
ofArchives, History and Records Management, Tallahassee.
Appendix 1: Dolphin cranium details
Most of the ventral-facing vault of the skull, consisting of
the maxilla, frontal, and parietal bones is present, but most of
the rostrum (consisting of the maxilla and premaxilla superi-
orly, and mandible inferiorly) and all of the teeth are missing.
The mid section of the cranium is missing, but portions of the
maxillary-frontal suture are present, and the suture is fused,
indicating maturity. The inferior aspect ofthe left frontal bone
is present, creating the superior orbital. Both of the comma-
shaped, asymmetrical nasal bones are present on the frontal
aspect of the skull. The left nasal bone is bigger, which is not
unusual for dolphins (Iablokov et al. 1974). Laterally, part of
the right temporal and the superior aspect of the left temporal
are present. Most of the right and left parietal bones on the
superior aspect of the skull are present. The width across the
parietals is approximately 120 mm. The dorsal skull, consisting
of the occipital bone, is nearly complete. Both occipital
condyles are present, but the foramen magnum is incomplete.
There are fragments of the inferior portion ofthe skull, but none
articulate to the vault.
Wing, Elizabeth S., and L. Jill Loucks
ELGART
MmIAn CIRCLE ANmMAL FYATuREs
THE FLORIDA ANTHROPOLOGIST
ENDOWMENT
Support production of
The Florida Anthropologist,
the scholarly journal published quarterly by
The Florida Anthropological Society since 1948.
Donations are now being accepted from
individuals, corporations, and foundations.
Inquiries and gifts may be directed to:
The Editor
The Florida Anthropologist
P.O. Box 6356
Tallahassee, FL 32314
The Florida Anthropological Society is a non-profit organization
under section 501(c)(3) of the Internal Revenue Code.
Contributions are tax-deductible as provided by
section 170 of the code.
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE AT BRICKELL POINT (8DA12)
RYAN J. WHEELER
2026 Chuli Nene, Tallahassee, FL 32301
E-mail: rwheeler@nettally.com
Introduction
Review of materials recovered from excavations at the
Miami Circle (8DA12) revealed that a large number of pumice
artifacts (n=121) had been found. Pumice is a form of volcanic
glass, expelled from volcanoes during explosive eruptions
(Mottana et al. 1978:315). It consists of gas bubbles trapped in
a groundmass of fragile volcanic glass and minerals. The
numerous gas bubbles make pumice light-weight and buoyant
in water. It is not native to Florida, and represents a raw
material with an extra-local origin. The nearest sources include
volcanoes in the Lesser Antilles island arc and Mexico (Simkin
1981) (Figure 1). Pumice artifacts are occasionally found at
sites in Florida, especially in southern Florida, usually in small
numbers. The large amount of pumice from the Miami Circle
offered a good opportunity to study this rare and possibly
important artifact type. Questions addressed in this study
include the place of pumice artifacts in the toolkit of Florida
Indian peoples; a companion study uses several different
geochemical approaches to determine origins and relationships
of this material (see Kish 2006, this issue).
Previous Research
Archaeologist John Goggin (n.d.:615-616) speculated that
pumice in southern Florida sites was drift material carried by
the Gulf Stream from perhaps the Lesser Antilles. He cites an
early source (Guppy 1917:164) that suggests pumice found on
the eastern side of the Turks Islands may be from the eastern
Atlantic, perhaps the Azores. Goggin (n.d.:616) counters by
suggesting that some pumice could have originated in Peru,
carried from the mountains to the sea by the Amazon River.
Other authors have favored the Caribbean and Lesser Antilles
as the source for pumice in Florida sites (Griffin 1982:69,
1988:99; Willey 1949:37). Harley Means, of the Florida
Geological Survey (personal communication, 2002), notes that,
in recent years, pumice occasionally washes ashore along the
Florida coast. Dove (1745:318-319) provides a fascinating
account of an extensive raft of pumice encountered by shipping
traffic in 1724 in the southern Atlantic. Bryan (1972) analyzed
pumice that washed ashore in Palm Beach County in the late
1960s and concluded that the Lesser Antilles was a likely
source. These reports confirm that ocean currents and not
human activity are responsible for transporting the pumice from
its source (or sources) to Florida. Donovan (1999:323)
indicates that pumice may survive on open beaches for as long
as 100 years, and that pumice clasts can float for up to a year
before becoming saturated and sinking.
Pumice artifacts are rarely found in Florida sites, and when
found are usually present in small quantities. A review of the
literature indicates that when found, pumice artifacts were
usually only mentioned in tables of artifacts, and not studied for
patterns of use wear or artifact form. Some authors have
described pumice artifacts as abraders, or noted perforated
examples. In some cases, pumice artifacts are misidentified as
fossils or limestone (see Steinen 1982:Figure 6.2h, Table 6.2),
while some objects identified as pumice turn out to be wood,
limestone, slag, or modern landscaping materials. A survey of
collections held by the Historical Museum of Southern Florida
(HMSF), the Florida Bureau of Archaeological Research
(FBAR), and Florida Museum of Natural History (FLMNH)
added additional specimens to our sample of pumice artifacts
(see Table 1 and Appendix A).
Geographically the pattern seems to indicate that pumice is
most frequently encountered at sites in southeastern Florida
(including numerous Everglades tree island sites), with occa-
sional examples along the Atlantic coast and around Lake
Okeechobee, and rare examples on the southwestern Florida
coast (see Appendix A). Temporal patterns are more difficult
to discern, though many examples of pumice come from sites
dating to the Glades I period. There are clear examples of
pumice from Glades II and Glades III contexts as well, includ-
ing several specimens from sites of the European contact period
(Glades IIIc period). Paul Heinrich (personal communication,
2002), geologist with the Louisiana Geological Survey,
recovered pumice clasts from the Bayou Jasmine mound site in
Louisiana and reports that there are some occurrences of
pumice in Texas archaeological sites as well.
Study Sample
A total of 157 pumice artifacts were examined for this
study. The bulk of these are from the Miami Circle site
(8DA12). Samples from other sites held in museum collections
were studied as well for comparative purposes.
Pumice from the Miami Circle
Examples of pumice used in this study come from all three
phases of excavations at the Miami Circle site: the original
joint investigation by the Miami-Dade Office of Historic
Preservation and the Archaeological and Historical Conser-
vancy (1998-1999); the State of Florida investigation in late
1999; and the University of Houston field school in summer,
2000. A total of 173 pieces of pumice, representing 121
artifacts, were recovered from 91 proveniences across the site.
VOL 59(3-4) THE FLORIDA ANTHROPOLOGIST SEPTEMBER-DECEMBER 2006
VOL. 59(3-4)
THIE ftowiDA ANTHROPOLOGiST
SEPTEmBER-DECEMBER 2006
THE FLORIDA ANTHROPOLOGIST 2006 VOL 59(3-4)
Figure I. Volcanoes in the Gulf of Mexico, Caribbean, and Atlantic regions (after Simkin 1981).
In some cases several fragments from the same piece or clast
were found together, thus the discrepancy between number of
pieces and number of artifacts.
Museum Collections
Thirty-six pumice artifacts were borrowed from the Histori-
cal Museum of Southern Florida (n= 1), the Florida Museum
of Natural History (n= 11), and the Florida Bureau of Archaeo-
logical Research (n= 14). Table I provides a list of the pumice
artifacts obtained from each collection. These artifacts were
analyzed for all macroscopic categories described below, but
are not included in the statistical results discussed.
Macroscopic Analysis
Data on each pumice artifact was recorded on a separate
analysis sheet. Scaled drawings and photographs were made
for most of the pumice specimens. The following categories of
information, which are typically used in analysis of obsidian
(another volcanic glass) artifacts, were recorded for each
specimen:
Color: Hand Specimen
Colors noted for the weathered surface and at any breaks
were recorded using a Geological Society of America Rock-
Color Chart, which follows the Munsell system of describing
hue, value, and chroma. Colors observed ranged from light
pale yellow (2.5Y 7/2) to dark gray (10YR 4/1). Common
colors (those with 5 or more examples) include dark gray
(10YR 4/1), gray (10YR 5/1), gray to light gray (10YR 6/1),
pale brown (10YR 6/3), light gray (10YR 7/1 to 10YR 7/2),
very pale brown (1 OYR 7/3 to I OYR 7/4), brown (7.5YR 5/2 to
7.5YR 5/4), pinkish gray (7.5YR 6.5/2, 7.5YR 6/2), and light
brown (7.5YR 6/3 to 7.5YR 6/4).
Color: Texture
Color distribution, either on the surface or at fresh breaks,
was noted for each artifact. Typical variations include uniform
(color is evenly distributed), mottled (a variegated or patchy
distribution), banded (either distinct or indistinct bands of
color). In most cases color was uniformly distributed on the
specimens examined.
Surface Luster
Surface luster was examined at fresh breaks when possible,
though often no such breaks were present. Two major variants
of luster were noted--chatoyant, which is an iridescent or pearl-
like sheen (Figure 2), and earthy, which is a general lack of
2006 VOL 59(3-4)
THE FLORIDA ANTHROPOLOGIST
WHEELER PUMICE ARTIFACTS FROM TIlE MIAMI CIRCLE
Table 1. Pumice artifacts from museum collections.
Site Name Site No. Description Catalog No.
Higgs IR24 rounded clast FBAR 75.62.10.1
Higgs IR24 well-rounded clast, cup- FBAR 75.62.102.1
shaped depression,
deep groove
Stock Island M02 well-rounded clast FBAR 74.2.10.8
Indian Key M015 rounded clast, flat FBAR 72.20.534.1
surfaces
Indian Key MO15 rounded clast FBAR 72.20.555.1
Indian Key MO15 well-rounded clast FBAR 72.20.888.1
Indian Key M015 rounded clast FBAR 72.20.832.1
De Soto Grove B BR82 rounded clast FBAR 94.123.135.1
Deering Estate Midden DA6519 rounded clast FBAR 00.14.39.8
Deering Estate Midden DA6519 sub-angular clast FBAR 00.14.39.8
Deering Estate Midden DA6519 rounded clast FBAR 00.14.47.1
Whitebelt 1 PB220 rounded clast, one flat FBAR 99.65.7.1
surface and one deep
groove
Shark Butchery SL37 rounded clast, dome- FBAR 76.209.1.1
shaped w/ one flat
surface
Honey Hill DA411 rounded clast, HMSF 2471
perforated w/ cup-
shaped depression
Honey Hill DA411 rounded clast HMSF 2942.1
Brickell Point DA98 well-rounded clast HMSF 865.3
Brickell Point DA98 angular clast HMSF 582.1
Brickell Point DA98 rounded clast HMSF 559.1
Brickell Point DA98 rounded clast, cup- HMSF 687.1
shaped depression & 3
grooves
Brickell Point DA98 rounded clast, cup- HMSF 492
shaped depression and
shallow groove
Brickell Point DA98 rounded clast, cup- HMSF 896
shaped depression, one
flat surface, and some
scratches
Granada DA11 rounded clast, shallow HMSF 1018.18
groove
Sutton DA1034 rounded clast HMSF 1874.1
Custom House DA1064 angular to sub-angular HMSF 427.5
clast__
luster, often associated with a matte or grainy texture. Weath-
ered surfaces of the pumice examined often have earthy luster,
while fresh breaks exhibited the chatoyant luster.
Surface Texture
Surface texture varied among the pumice artifacts analyzed,
with most examples best described as "tube pumice" with
elongated vesicles or tubes of glass forming a distinctive
striated surface texture. Tube pumice forms when material is
expelled from volcanoes under high pressure. Some ofthe tube
pumice was very dense with tightly packed, narrow vesicles,
while other examples were very spongy with broad, poorly
organized vesicles. Other variations include a grainy or sugary
texture, which appears to be related to a large number of
inclusions; examples with dense, chalky texture; and those with
a spongy texture, composed of large and poorly organized
vesicles. Rare examples had popcorn-like surfaces, composed
WHEELER
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
THE FLORIDA A1~CrliROPOLOGIST 2006 VOL 59(34)
Table 1. Pumice artifacts from museum collections., continued.
of broken, perfectly round glass vesicles. Most specimens
examined in this study are examples of tube pumice, with
densely packed vesicles (Table 2). Figure 3 illustrates some of
the varied pumice textures observed within the study sample.
Figure 2. Highly iridescent or chatoyant pumice, Miami
Circle (MDC.1.370.12).
Inclusions
A Bausch & Lomb 20x Hastings loupe was used to examine
the surface of each specimen for inclusions. Several types of
inclusion were noted in the sample, including phenocrysts,
autoliths, and xenoliths. Phenocrysts are crystals visible in the
pumice groundmass. Phenocrysts typically noted in the pumice
specimens were large, milky white amphibole and clear
plagioclase crystals, and black or brown platy crystals of
biotite. Phenocrysts were very common in the assemblage
(Table 3). Other inclusions include autoliths, which are pieces
of volcanic glass that have reformed within the pumice
groundmass, and xenoliths, which are inclusions of other
volcanic rock fragments. The autoliths in the pumice analyzed
were typically melted shards of volcanic glass, surrounded by
a void and slightly white area of the pumice groundmass.
Xenoliths noted were typically black rock fragments, but
examples of gray and brown rock fragments also were noted.
Surface Cortex
In most cases the pumice artifacts considered here had
weathered exterior cortices, often a darker color than the
material exposed at fresh broken surfaces (Figure 4). The
weathered surface usually was chalky or matte in appearance
and texture. Some examples did not have weathered surfaces,
but were bright and lustrous.
Maximum Dimensions
Measurements along three axes (length, width, and height)
were made in centimeters with vernier calipers. Specimens
ranged in size considerably, from small clasts around 1.2 cm
long to large examples around 13.7 cm long (Figure 5). The
Site Name Site No. Description Catalog No.
Key Largo I M025 plummet-form FLMNH 91-84-3
Key Largo 1 M025 grooved plummet FLMNH 91-84-5
Bear Lake 1 MO33 rounded clast FLMNH 98576
Ft. Center GL13 dome-shaped w/ one FLMNH A16334
flat surface
Ft. Center GL13 one smoothed, convex FLMNH 098767
surface
Ft. Center GL13 large loaf-shaped clast FLMNH A16093
Ft. Center GL13 two depressions on FLMNH A16262a
opposite faces
Ft. Center GL13 portion of cup-shaped FLMNH A16262b
depression
Ft. Center GL13 dome-shaped w/ one FLMNH A16262c
flat surface
Ft. Center GL13 one cup-shaped FLMNH A16262d
depression ___
Ft. Center GL13 sub-rounded, discoidal FLMNH A16262e
clast
W2
THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
WHEELER PUMICE ARTIFACTS FROM ThE MIAMI CIRCLE
cm cm
cm N J Cm
Figure 3. Examples of different pumice textures: a) tube pumice (MDC.1.471.6b); b) grainy (MDC.1.222.1 la); c) popcorn-like
(HMSF 427.5, Customs House site 8DA1064); d) spongy (MDC.1.222.11c); e) dense and chalky (FBAR 72.20.534.1, Indian
Key 8MO15). All from Miami Circle, except c and e.
average length is 4.2 cm. The distribution of pumice size
classes (by length) follows a bell curve, with the most clasts in
the 2.1-4.0 and 4.1-6.0 cm categories. Most of the pumice
artifacts from other sites have similar dimensions, though one
example fromFt. Center(8GL13)is quite large, measuring 16.7
cm in length.
Weight
All specimens were weighed, in grams, with an electronic
balance. Overall, 1,789 g of pumice was recovered from the
Miami Circle excavations. Specimens ranged in weight from
0.4 g to 116.3 g, averaging 14.8 g (Figure 6). The weight
distribution of pumice clasts is skewed toward to lower weight
classes. One exceptional example from Ft. Center (8GL13),
mentioned above, is 484.9 g.
Shape: Roundness
Standards of roundness and sphericity are presented by
WHEELER
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
THE FLORIDA ANTHROPOLOGIST
Table 2. Surface textures of Miami Circle pumice specimens.
Surface texture Count Percent
tube pumice 82 68
grainy 16 13
spongy 12 10
dense & chalky 10 8
popcorn-like 1 <1
Table 3. Inclusions in Miami Circle pumice specimens.
Inclusions Count Percent
of Total
phenocrysts 111 92
autoliths 14 12
xenoliths 17 14
none observed 8 7
Table 4. Roundness of Miami Circle pumice specimens.
Roundness Count Percent
N/A
well-rounded 5 4
rounded 67 55
sub-rounded 10 8
sub-angular 1 <1
angular 33 27
not determinable 5 4
Powers (1989) and are a means to accurately describe the shape
of a geological specimen (also see Pettijohn 1975, and Skinner
2000). Pieces were usually described as well-rounded, round-
ed, sub-rounded, sub-angular, or angular (Table 4). Pumice
deposited after eruption is typically angular, though most intact
examples within the study set were rounded. It is unclear if this
is due to transportation via water or from use. Guppy
(1917:164) notes that examples ofpumice washed ashore in the
Turks Islands are usually rounded cobbles, suggesting that this
may be the typical form encountered in specimens recovered
from beach drift.
Shape: Sphericity
The generalized sphericity of each pumice clast was
described. Sphericity standards are from Power (1989, also see
Pettijohn 1975, and Skinner 2000). Terms used to describe
sphericity include discoidal, sub-discoidal, spherical, sub-
prismoidal, and prismoidal (Table 5). Often times, fragmentary
specimens could not be described for sphericity. Most of the
specimens from the Miami Circle were described as sub-
discoidal or spherical. Powers (1989) states that geological
particles are usually discoidal, rodlike (prismoidal), or spheroi-
dal when they initially break out of or weather from parent
materials. Weathering and abrasion tend to produce more
spherical shapes. It is unclear if the pumice clasts examined
during this study reached their rounded, spherical forms from
use or from wave action and weathering.
Modification
Evidence for artifactual use, like grooves, flat sur-
faces, cup-shaped depressions, facets, or perforations
were noted (Figures 7 and 8). Forty of the pumice clasts
from the Miami Circle exhibit one or more of these forms
of modification. Examples from other sites had similar
modifications (see Table 1). These artifacts with clear
evidence of use are described below in more detail.
Shallow Grooves (n=7). Shallow grooves are linear,
often fusiform, depressions that are about as wide as they
are deep. Diameters of these grooves range from 0.2 to
1.0 cm. Shallow grooves may result from use of the
pumice clasts in polishing bone artifacts. The shape and
size of the shallow grooves corresponds favorably with
many of the mammal long bone implements from the
Miami Circle (Wheeler 2004a). Shallow grooves often
occur with other forms of modification, like cup-shaped
depressions. In some cases only one groove is present,
while multiple grooves (2 to 3) are common. These
grooves usually are parallel to one another and suggest
these artifacts were used like hones to polish or sharpen
bone artifacts. Pumice artifacts from other sites also
exhibit shallow grooves. Two clasts of pumice from
excavations at the Brickell Point site (8DA98), adjoining
the Miami Circle property to the south, have combina-
tions of shallow grooves and cup-shaped depressions. A
discoidal clast from the Granada site (8DAll) also
exhibits a single shallow groove.
Deep Scoring (n=8). Deep score marks are similar to shallow
grooves, but tend to be narrower and deeper. Score marks
range from 0.1 to 0.3 cm in width, but may be as deep as 0.4 to
0.5 cm. These score marks may have resulted from abrasion
with cordage. Cords may have been tied around some clasts,
causing the deep scoring, or cordage may have been used to
saw large clasts into smaller pieces.
Cup-shaped Depressions (n=5). Cup-shaped depressions,
usually occupying an entire surface of a rounded, spherical
clast, may have been the result of polishing rounded wood
surfaces. Cup-shaped depressions were noted on clasts from
other sites as well, including the Higgs site (8IR24), Brickell
Point (8DA98), and the Honey Hill/Stadium site (8DA411).
The example from the Honey Hill/Stadium site (8DA411) has
the remnants of a perforation at one end (Figure 9).
Flat Surfaces (n=16). Flat surfaces are often found on dome-
shaped clasts. These dome-shaped pieces with one flattened
surface are suggestive of sanding blocks. It is possible that
these pieces were used in finer sanding and polishing of wood.
Historically, pumice has been used in polishing and refinishing
of lacquered wood surfaces, frequently in the technique known
as "French polishing" (Hayward 1975; Matthew 1955). Pumice
also has been used in Europe for polishing ivory and some
metals (Timothy Burke, personal communication, 2002; also
2006 VOL. 59(3-4)
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
Figure 4. Weathered exterior of pumice clast, Sutton site (8DA1034) (HMSF 1874.1).
see discussion of pumice as an abrasive for cleaning armor
during the sixteenth century in Ffoulkes 1912:80).
Faceting (n=5). Faceting occurs when two or more, often
rather small, flat surfaces are adjacent to one another. Faceting
appears to result from use in polishing or abrading, like the
dome-shaped specimens with one larger flat surface described
above. In this case, however, the objects being polished may be
smaller or more intricate, requiring the user to rub with only a
small part of the clast. Multiple facets probably occur as the
clast is rotated to best fit into tight spaces. Pumice clasts with
faceting also may have served in polishing wood, especially
more elaborate carvings with recessed areas.
Ring or Crescent-shaped Forms (n=2). No complete ring-
shaped clasts (circular with a central perforation) were found,
but there are a number of crescent-shaped clasts that may be
broken rings. These are the most highly modified pumice
artifacts observed. It is not clear ifthese were utilitarian objects
or were ornamental.
Cube-shapedForms (n=3). Cube-shaped specimens have 6 flat
sides and resemble a contemporary gaming die. It is possible
that these cubes were used in abrading or polishing, though they
also may have been made into cubes for some other purpose.
Plummet Forms (n=2). Two specimens, both from Key Largo
1 (8M025) (see Table 1 and Figure 10), are pumice clasts
shaped like the plummet-form artifacts most typically made
from limestone or shell (Reiger 1990:227). The first example
is nicely made, almost circular at its widest point, with a
grooved knob at one end. The second specimen is flat and
rectangular with rounded corners, and a shallow, irregular
groove at one end.
Distribution of Pumice at Brickell Point
Pumice appears to be abundant throughout the Brickell
WHEELER
THE FLORIDA ANTHROPOLOGIST 2006 VOL 59(3-4)
Table 5. Sphericity of Miami Circle pumice
specimens.
Sphericity Count Percent
N/A 34 28
discoidal 9 7
sub-discoidal 17 14
spherical 29 24
sub-prismoidal 6 5
prismoidal 4 3
not determinable 22 18
Point site. Excavations associated with the 2.2 acre Miami
Circle property produced numerous examples from virtually
every area investigated. Excavations conducted in the early
1980s at the neighboring Sheraton Hotel property (referred to
here as 8DA98) also produced examples of pumice. The
vertical and horizontal distribution of pumice artifacts at
Brickell Point is discussed below for each phase of work.
Miami Circle Excavations
The joint Miami-Dade County and Archaeological &
Historical Conservancy excavations investigated three areas of
the site. Area 1 refers to the excavation of the Miami Circle
feature and is comprised of 57 5 x 5 foot square units; Area 2
consists of 8 units excavated to the northwest of the Miami
Circle feature near the historic edge of the Miami River; and
Area 3 is one unit excavated 107 feet (33 m) to the west of Area
1.
Thirty-four of the 57 excavation units associated with Area
1 and the Miami Circle feature produced pumice artifacts.
Examples of pumice also were found in the eastern extension
trench associated with Area 1 (n=2), and there are miscella-
50
2.1-4 cm
45 -
40----
35--
30 -
25
20
15 i
10 1
neous examples associated with the original Circle trench
(n=4). Pumice was found in 4 of the 8 units in Area 2, and Unit
17 (Area 3) produced 3 pumice artifacts. Interestingly, of the
39 units that produced pumice artifacts, 21 have more than one
piece, and several units have five or more pumice artifacts.
The distribution of pumice artifacts within Area 1-the
Miami Circle feature is heavily weighted toward the northwest
quarter of the Circle (n=33). The other quarters have half or
fewer as many pumice objects when compared to the northwest
quarter. This horizontal patterning suggests that while pumice
is found throughout the site, there may be some areas with
greater concentrations (Table 6). A weighted formula was
developed in order to account for different volumes of soil
excavated within each quarter (see Wheeler 2004b on shell
tools from the Miami Circle). Units in the NW of the Circle
feature produced the most pumice artifacts and had one of the
highest densities for pumice artifacts as well. This trend is
similar to that noted for shell tools, and may suggest activity
areas or areas with differential deposition of midden. Interest-
ingly, the pattern observed for shell tools indicates similar
densities of tools within areas 1 and 2, while the pumice
distribution is weighted more toward Area 1 (the Miami Circle
feature). This suggests that finer wood and bone working may
have been done in and around the area where the Miami Circle
is, while rougher woodworking tasks may have been concen-
trated near the original shore of the Miami River.
Vertical distribution ofpumice also can be demonstrated for
the deposits. Twelve units have pumice artifacts from two or
more adjacent levels, while 5 units have pumice from non-
adjacent units. If the pumice all originated from one or two
major volcanic events, the vertical and horizontal distribution
suggests the midden accumulated rather quickly over an
extensive area. Alternatively, if the pumice represents material
from a number of temporally distinct eruptions, there may be
other cultural factors explaining its ubiquitous presence in the
midden.
M<1 cm
I 1-2 cm
*2.1-4 cm
M4.1-6 cm
86.1-8 cm
*8.1-10 cm
*>10.1cm
<1 cm
Figure 5. Histogram showing size distribution of pumice clasts.
1-2 cm
6.1-8cm
8.1-10 cm
O >10.1cm
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
WHEELER PuMiCE ARTIFAUrS FROM THE MIAMI CIRCLE
11-20a
210 31-40g 41-50g
61-70 >70g
51-60g mM6170
Figure 6. Histogram showing weight distribution of pumice clasts.
State of Florida Investigations
Pumice artifacts were recovered from 19 proveniences
associated with the State of Florida investigation of the Miami
Circle at Brickell Point site (8DA12). Thirty-seven pieces
representing 27 clasts were found. Proveniences include
excavations around a septic tank, located 17 m west of the
Miami Circle feature, and features in the N34 or "Valley of the
Holes" area. The presence of pumice artifacts in these areas
helps confirm that this material is distributed across the
property. Vertical distribution is similar to that noted for the
Miami Circle excavation units. One meter square units
excavated adjacent to a septic tank produced 9 pumice clasts
from multiple levels. Within the N34 or "Valley of the Holes"
area 18 pumice clasts were recovered from midden deposits
within 10 of the holes or basins carved into the limestone
bedrock
Other Site Areas
Three clasts of pumice were recovered from the University
of Houston fieldschool in 2000, located between the Miami
Circle feature and the N34 or "Valley of the Holes" area. Other
proveniences across the site include Trench 1, one of the
demolition trenches dug to remove the footings of the buildings
comprising the Brickell Point Apartments.
8DA98
Salvage excavations on the Sheraton Hotel property
(referred to here as 8DA98), directly to the south of the lot
where the Miami Circle was found, also produced pumice
artifacts. Both lots are components of the larger Brickell Point
site, which extended west and south of the point where the
Miami River meets Biscayne Bay. Bob Carr directed work at
the Sheraton component in 1980-1981 prior to the construction
of the hotel (initially a Holiday Inn) (Carr and Ricisak
2000:266-267). Excavations included a north-south trench and
other tests, which uncovered holes cut into the limestone,
similar to those that make up the Miami Circle feature. Six
pumice artifacts recovered during the excavations are curated
at the Historical Museum of Southern Florida. These artifacts
help confirm that the deposits at the Sheraton component are
likely of similar age and nature as those encountered at the
Miami Circle component.
Discussion
Technological Place of Pumice Artifacts
John Goggin (n.d.:615-616) suggests that the flat surfaces
and facets observed on some pumice artifacts are produced by
grinding and smoothing tasks related to woodworking. Some
of the additional modified pumice artifacts, like those with cup-
shaped surfaces, also were likely used in polishing wood. A
role in woodworking makes sense, since many of the shell
artifact forms and some of the bone artifact types identified
from the Miami Circle site are best described as woodworking
tools (Wheeler 2004a, 2004b). Direct evidence of a sophisti-
cated tradition of woodworking is found at the Key Marco site,
1-10g
50 -
40 ---
20
10
0
B<1g
El-lOg
H1-10g
E11-20g
121-30g
M 31-40g
M41-50g
M 51-60g
S61-70g
SB>70g
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
WHEELER
200 THE FLORIDA ANTHROPOLOGIST
2006 VOL. 59(3-4)
Table 6. Pumice distribution for Excavation Areas 1, 2 and 3.
where hundreds of examples of finely carved, incised and
painted wooden artifacts were found (Cushing 1897; Gilliland
1975). Wooden artifact forms range from utilitarian objects
like bowls, pestles, and canoe paddles to ritual and decorative
pieces like animal and human effigies, masks, and ceremonial
tablets (Gilliland 1975; also see Fewkes 1928; Willey 1949:53-
59; Wheeler 1996:148-211). Bone and shell tools with
evidence of use in woodworking date back to at least the
Middle to Lake Archaic Mount Taylor culture (Wheeler and
McGee 1994). Analogy with European and American
woodworking traditions indicate that materials like pumice and
rottenstone (degraded limestone) are used in the finer polishing
and finishing of wood.
Pumice clasts with shallow grooves were likely used as
hones or abraders in forming and polishing bone artifacts,
especially those manufactured from the long bones of the deer.
Goggin (1951:61) notes the occurrence of limestone and sherd
hones at some sites as well. These have the same distinctive
shallow grooves likely produced from shaping and polishing
bone tools. Five-hundred and fifty-four bone artifacts were
recovered from the Miami Circle site, and many of these were
long bone implements (Wheeler 2004a). Debitage from bone
tool making helps confirm that many of these items might have
been made on site. Considering the importance of wood and
bone working among the native peoples of southern Florida, it
seems likely that pumice may have been a rare and valuable
item.
A short experiment with a hand-sized pumice clast and a
replica of a deer metapodial tool demonstrated that within ten
minutes a broad, shallow groove would be worn in the surface
of the pumice if used to polish the bone tool. The pumice
proved to be very durable and most of the residual dust ap-
peared to be from the bone implement. Polishing with the
pumice clast produced an even, moderate luster on the bone
tool. The pumice tended to obliterate other traces of manufac-
ture, like striations from a shark tooth knife. Wear patterns,
beyond the distinctive groove, were difficult to discern on the
pumice with a hand lens, largely because of the numerous air
cavities and vesicles characteristic of the volcanic glass,
Pumice in Regional Exchange
Pumice from Florida sites has received little consideration
Area Count Excavated Pumice clasts per
Volume (cubic cubic foot
feet
[NE “Circle [1023450 002
SW %Circle LIS 8S 600 0.03
[Exc.Area2 0 {8 SSO OT
[TOTALS OO 8h 37 0.03
as a raw material for tool making, let alone as an item with
potential social and economic significance. Prior to the
discovery of a large number of pumice clasts and fragments at
the Miami Circle site, it would have been logical to assume that
pumice was occasionally found on coastal beaches and retained
for use by the finders. In general, most sites that have produced
pumice are located along the southeastern Florida coast,
between the Indian River region and the Florida Keys (Figure
11). This distribution is consistent with recent “drifter†or drift
buoy experiments, which indicate that floating materials subject
to the Caribbean and Gulf Stream currents will come close to
shore or strand in southeastern Florida (see Leaman 1998).
Drift buoys generally do not strand on the southwestern Florida
Gulf Coast. There are, however, a number of interior sites that
also have produced pumice. Many of these are Everglades tree
island sites in Dade and Broward counties whose inhabitants
may have frequented the sea coast. There also are a number of
sites with pumice farther from the coast (e.g., Belle Glade, Fort
Center) or on the southwestern Gulf Coast (e.g., Wightman).
This suggests that pumice may have been involved in the
regional exchange system, along with other coastal products
like shark teeth and marine shells (see Kozuch 1993:32-34;
Luer 1989:112-114, 116-121). Admittedly, pumice was an
unpredictable commodity, but was likely collected when
available. It is possible that much of the pumice found in
Florida archaeological sites has origins in one or two major
volcanic events. If large rafts of pumice drifted ashore near the
Miami Circle the material could have been readily collected and
introduced into the regional exchange system. Whitebelt 1,
Belle Glade, and Fort Center, the three sites located around
Lake Okeechobee that have pumice, also have other items
indicating participation in regional and long-distance exchange,
including shark teeth and marine shells; groundstone axes and
other objects; chipped stone artifacts; galena; ceramic platform
pipes; and ceramics from neighboring areas (Sears 1982;
Wheeler 2000b; Willey 1949). Interestingly, these items are
characteristic of the Miami Circle site as well.
Excavators working at the Miami Circle at Brickell Point
were initially impressed by the diverse exotic artifacts recov-
ered from the site (Carr and Ricisak 2000). One small copper
bead, a galena bead and another small piece of galena suggested
long-distance exchange networks involving other parts of the
Southeast and Midwest. A sourcing study of basaltic stone axes
WHEELER PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
C
a"
I.-,
"~1 1
cup-shaped
depression
/'facets
( / ,
cI 27
i'i' 2 4 6
Figure 7. Forms of pumice modification: a) dome-shaped with one flat surface (MDC.1.254.25c); b) shallow groove
(MDC.1.471.6); c) shallow groove (MDC.1.166.4); d) deep scoring (MDC.1.579.12); e) deep scoring (MDC.1.427.9); f) cup-
shaped depression, side and reverse reviews (HMSF 492, Brickell Point site); g) facets on obverse, cup-shaped depression on
reverse (MDC.1.569.12); h) crescent-shaped (MDC.1.628.9); i) crescent-shaped (MDC.1.629.13); j) cube-shaped
(MDC.1.254.25a). AU from Miami Circle, except f.
helped confirm exchange networks reaching to the Macon,
Georgia area (Dixon et al. 2000). Studies of American Indian
exchange in Florida are rare, and southern Florida has always
been viewed as somewhat parochial. However, studies of
pumice and chipped stone artifacts from the Miami Circle at
Brickell Point site have provided a significant look at patterns
of exchange in southern Florida (see Austin 2004).
Austin's (2004) study of chipped stone material from the
Miami Circle complements the pumice study, since it points to
a connection with the Fort Center site located on Fisheating
Creek to the west of Lake Okeechobee; both sites exhibit
similar cobble technologies of lithic reduction and Austin
(2004) suggests that Fort Center may have served as a center
for redistribution of chert cobbles initially procured in the
Tampa Bay area. Likewise, the inhabitants of the Miami Circle
at Brickell Point may have controlled distribution of pumice.
Austin (2004:113, 117) suggests that down-the-line or
redistributional forms of exchange couldhave characterized the
system in place in southern Florida. He explains that in down-
the-line exchange, the chert cobbles would have moved from
the source area to Fort Center and then on to the Miami Circle,
while in the more formal redistributional system, one commu-
nity would serve as a redistribution center for other communi-
ties within the exchange network. Austin's (2004:118, 127-
128) analysis leads him to conclude that a redistributional
exchange system was in operation, with Fort Center serving as
a center for redistribution of cobbles and other cherts. The
Miami Circle site may well have had a similar role in the
redistribution of pumice. Prior to these studies our ideas about
patterns and systems of exchange in southern Florida were
limited.
Conclusion
A large number of pumice clasts (n=121) recovered from
excavations at the Miami Circle afforded an opportunity to
,CS
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
WHEELER
rD
Tm~FLRIA NTROOLGIT 00 VL.9(-4
cm 1 cm
Figure 8. Photograph of modified pumice clasts: a) shallow groove (FBAR 99.65.7.1, Whitebelt 1 site 8PB220); b) deep
scoring (MDC.1.579.12); c) dome-shaped with one flat surface (MDC.1.254.25c); d) crescent-shaped (MDC.1.629.13); e)
cube-shaped (MDC.1.254.25a); f) obverse, side, and reverse views showing shallow grooves and cup-shaped depression
(HMSF 492, Brickell Point site 8DA98). All from Miami Circle, except a and f.
2006 VOL. 59(3-4)
Tim noRinA ANTHROPOLOGIST
WHEELER PUMICE ARTIFACTS FROM THE MIAMI CIRCLE 203
Figure 9. Clast with cup-shaped depression and perforation, Honey Hill/Stadium site (8DA411) (HMSF 2471).
Figure 10. Plummet-form examples made from pumice, Key Largo 1 (8M025) (FLMNH 91-84).
TirE FLORIDA ANTHROPOLOGIST 2006 VOL.59(34)
Figure 11. Distribution of pumice artifacts in Florida.
study the role of this unusual, non-local material in native
Florida. Ocean currents are likely responsible for carrying
pumice to Florida from volcanoes in the Lesser Antilles or
Mexico (see Kish 2006, this issue).
Macroscopic examination of the pumice clasts revealed
some variation in color, texture, and inclusions, though the
majority of the specimens were within the range of variability
for materials produced by a single eruption.
Forty of the pumice clasts from the Miami Circle had
evidence of one or more form of modification. Evidence for
artifactual use includes shallow grooves, deep scoring, cup-
shaped depressions, flat surfaces, faceting, ring or crescent-
shaped forms, and cube-shaped forms. Similar modifications
were noted on examples from museum collections, indicating
a broader tradition of making and using pumice artifacts in
southern Florida. Forms and wear patterns indicate use in wood
polishing and bone tool manufacturing.
The vertical and horizontal distribution ofpumice within the
Brickell Point site, including the Miami Circle component, may
be evidence that the midden was deposited very rapidly over a
large area. Some areas with greater densities of pumice may
indicate specialized activity areas or differential deposition of
refuse and discarded tools. Alternatively, inhabitants of the
Brickell Point site may have been involved in the long-term
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
WHEE PuIEAimAr RO H isiCRL
collection of pumice produced by temporal distinct volcanic
eruptions. Pumice may have been a traditional item collected
by residents of the site, either because it was needed for
manufacture of wood and bone artifacts, or because it was an
important commodity in the regional exchange system.
In terms of the regional distribution of pumice, sites with
pumice artifacts are concentrated alongBiscayneBay and in the
Everglades of Dade and Broward counties. Other examples are
known from coastal sites stretching from Key West to the Cape
Canaveral area in Brevard County. This distribution is consis-
tent with studies of drift buoy tracks influenced by the Carib-
bean and Gulf Stream currents. There also are, however, sites
located around Lake Okeechobee and on the southwestern Gulf
Coast that have examples of pumice. Considering the other
exchange items found at the Miami Circle, including ground-
stone axes, galena, and chipped stone artifacts, it seems likely
that pumice was introduced into the regional exchange network.
Pumice was probably valued as an abrasive by the peoples of
southern Florida who had strong traditions of bone and wood
carving.
References Cited
Austin, Robert J.
2004 Chipped Stone Artifacts from the Miami Circle Excava-
tions at Brickell Point. The Florida Anthropologist 57(1-
2):85-131.
Bryan, W. B.
1972 Cummingtonite-Dacite Drift Pumice Stranded at West
Palm Beach, Florida. Geological Society ofAmerica
Bulletin 83(12):3745-3746.
Carr, Robert S., and John Ricisak
2000 Preliminary Report on Salvage Archaeological Investiga-
tions of the Brickell Point Site (8DA12), Including the
Miami Circle. The Florida Anthropologist 53(4):260-284.
Cushing, Frank Hamilton
1897 Exploration of Ancient Key Dwellers' Remains on the
Gulf Coast of Florida. Proceedings of the American
Philosophical Society 25(153).
Dickel, David N.
1988 Test Excavations at the Singer Island Site (8PB214),
MacArthur Beach State Park, Palm Beach County,
Florida. Florida Archaeological Reports No. 3. Bureau of
Archaeological Research, Florida Division of Historical
Resources, Tallahassee.
Dixon, Jacqueline Eaby, Kyla Simons, Loretta Leist, Christopher
Eck, John Ricisak, John Gifford, and JeffRyan
2000 Provenance of Stone Celts from the Miami Circle Archae-
ological Site. The Florida Anthropologist 53(4):328-341.
Donovan, Stephen K.
1999 Pumice and Pseudoplankton: Geological and
Paleontological Implications of an Example from the
Caribbean. Caribbean Journal ofScience 35(3-4):323-
324.
Dove, John
1745 A Surprising Shoal of Pumice-Stones found Floating on
the Sea. In Memoirs of the Royal Society; or a new
abridgment of the Philosophical Transactions, 2"' ed.,
Vol. 8, pp. 318-319. Baddam's, London.
Ewen, Charles R., and John H. Hann
1998 Hernando De Soto among the Apalachee: The Archaeol-
ogy of the First Winter Encampment. University Press of
Florida, Gainesville.
Eyster, Irving R.
1986 Archaeological and Historical Site Survey of Proposed
Village at Key West Resort Prepared for Norman B.
Wood and Bernie C. Papy, Jr., Key West.
Fewkes, J. Walter
1928 Aboriginal Wooden Objects from Southern Florida.
Smithsonian Miscellaneous Collections Vol. 80, No. 9.
Smithsonian Institution, Washington, D.C.
Ffoulkes, Charles
1912 The Armourer and His Craft,from the X to the XVIth
Century. Methuen and Co., London.
Gilliland, Marion S.
1975 The Material Culture ofKey Marco, Florida. University
of Florida Press, Gainesville.
Goggin, John M.
n.d. Archeology of the Glades Area, Southern Florida (ca.
1949, with later additions). Typescript on file, P.K.
Yonge Library of Florida History, University of Florida,
Gainesville.
1950 Stratigraphic Tests in the Everglades National Park.
American Antiquity 15(3):228-246.
1951 Archeological Notes on Lower Fisheating Creek. The
Florida Anthropologist 4(3-4):50-66.
Goggin, John M., and Frank H. Sommer III
1949 Excavations on Upper Matecumbe Key, Florida. Publica-
tions in Anthropology No. 41. Yale University, New
Haven.
Griffin, John W.
1982 Lithic Artifacts from the Granada Site. In Excavations at
the Granada Site: Archaeology and History of the
Granada Site, Vol. 1, by John W. Griffin, Sue B. Rich-
ardson, Mary Pohl, Carl D. McMurray, C. Margaret
Scarry, Suzanne K. Fish, Elizabeth S. Wing, L. Jill
Loucks, and Marcia K. Welch, pp. 65-70. Prepared for
City of Miami by Florida Division of Archives, History
and Records Management, Tallahassee.
1988 The Archeology ofEverglades National Park: A Synthe-
sis. Contract CX 5000-5-0049. Prepared for Southeast
Archaeological Center, National Park Service, Tallahas-
see.
Guppy, H. B.
1917 Plants, Seeds, and Currents in the West Indies and
Azores. Williams and Norgate, London.
Hayward, Charles
1975 Staining and Polishing. Evans Brothers, London.
WHEELER
PtmicCE ARTwFAcTs FROM THE MMIvmI CutcLE
TalE FLORIDA ANTHROPOLOGIST 2006 VOL 59(34)
Jaffee, Howard
1976 Preliminary Report on a Midden Mound and Burial
Mound of the Boynton Mound Complex. The Florida
Anthropologist 29(4):145-152.
Jordan, Douglas F.
1963 The Jungerman Site. In Papers on the Jungerman and
Goodman Sites, Florida, pp. 1-23. Contributions of the
Florida State Museum, Social Sciences No. 10. Gaines-
ville.
Kish, Stephen A.
2006 Geochemical and Petrologic Characterization of Pumice
Artifacts from the Miami Circle-Brickell Point Archaeo-
logical Site Plus other Sites in Florida-Potential Prove-
nance Locations. The Florida Anthropologist 59(3-
4):209-240 (this issue).
Kozuch, Laura
1993 Sharks and Shark Products in Prehistoric South Florida.
Monograph No. 2. Institute of Archaeology and
Paleoenvironmental Studies, University of Florida,
Gainesville.
Laxson, Dan D.
1959a Excavations in Dade County during 1957. The Florida
Anthropologist 12(1):1-8.
1959b Three Salvaged Tequesta Sites in Dade County, Florida.
The Florida Anthropologist 12(3):57-64.
1962 Excavations in Dade, Broward Counties, 1959-1961. The
Florida Anthropologist 15(1): 1-10.
1970 Seven Sawgrass Middens in Dade and Broward Counties,
Florida. The Florida Anthropologist 23:151-158.
Leaman, Kevin
1998 Drifters in the Southwestern Caribbean and Status of
Drifters in the southwestern Caribbean as of August 17,
1998. Available at http://drifters.doe.gov/results.
Luer, George M.
1989 Calusa Canals in Southwestern Florida: Routes of Tribute
and Exchange. The Florida Anthropologist 42(2):89-130.
Matthew, W. P.
1955 French Polishing. Burke, New York.
Miami-West India Archaeological Society and Broward County
Archaeological Society
1975 The Arch Creek Site, Dade County. The Florida Anthro-
pologist 28(1):1-13.
Mottana, Annibale, Rodolfo Crespi, and Giuseppe Liborio
1978 Guide to Rocks and Minerals. Simon and Schuster, New
York.
Pettijohn, F. J.
1975 Sedimentary Rocks. Harper & Row, New York.
Powers, Maurice C.
1989 Comparison Chart for Estimating Roundness and Spheric-
ity (AGI Data Sheet 30.1). InAGI Data Sheets for Geol-
ogy in the Field, Laboratory, and Office, edited by J.T.
Dutro, Jr., R.V. Dietrich, and R.M. Foose. American
Geological Institute, Alexandria.
Reiger, John F.
1990 "Plummets"-An Analysis of a Mysterious Florida Arti-
fact. The Florida Anthropologist 43(4):227-239.
Sears, William H.
1958 The Grant Site-Br-56. The Florida Anthropologist
11(4):114-124.
Simkin, Tom (editor)
1981 Volcanoes ofthe World: A Regional Directory, Gazetteer,
and Chronology of Volcanism during the Last 10,000
Years. Hutchinson Ross, Stroudsburg.
Skinner, Craig E.
2000 Descriptive Terminology for Obsidian. Northwest Re-
search Obsidian Studies Laboratory, Corvallis. Available
at http://www.obsidianlab.com/megascop.html.
Smith, Hale G.
1956 The European and the Indian: European-Indian Contacts
in Georgia and Florida. Publications No. 4. Florida
Anthropological Society, Gainesville.
Steinen, Karl T.
1982 Other Nonceramic Artifacts. In Fort Center: An Archaeo-
logical Site in the Lake Okeechobee Basin, by William H.
Sears, pp. 68-110. University Presses of Florida, Gaines-
ville.
Wheeler, Ryan J.
1996 Ancient Art of the Florida Peninsula: 500 B.C. to A.D.
1763. Ph.D. dissertation, Department of Anthropology,
University of Florida, Gainesville.
2000a The Archaeology of Brickell Point and the Miami Circle.
The Florida Anthropologist 53(4):294-322.
2000b Archaeological Exploration ofDupuis Reserve Part One:
Cultural Resource Assessment ofFour Archaeological
Sites at Dupuis Reserve, Palm Beach County. Bureau of
Archaeological Research, Florida Division of Historical
Resources, Tallahassee.
2004a Bone Artifacts from the Miami Circle at Brickell Point
(8DA12). The Florida Anthropologist 57(1-2):133-158.
2004b Shell Artifacts from the Miami Circle at Brickell Point
(8DA12). The Florida Anthropologist 57(1-2):159-186.
2003 Archaeological Survey and Evaluation of Additional
Portions of the Pavilion Site (8SL1173), St. Lucie
County, Florida. Prepared for Robert T. and JoLynn
Hansen by Panamerican Consultants, Inc., Tampa.
Wheeler, Ryan J., and Ray M. McGee
1994a Technology of Mount Taylor Period Occupation, Groves'
Orange Midden (8V02601), Volusia County, Florida.
The Florida Anthropologist 47(4):350-379.
1994b Wooden Artifacts from Groves' Orange Midden. The
Florida Anthropologist 47(4):380-389.
THE FLORIDA ANTHIROPOLOGIST
2006 VOL. 59(3-4)
WHEELER PUMICE ARTIFACrS FROM THE MIAMI CIRCLE
Willey, Gordon R.
1949 Excavations in Southeast Florida. Publications in An-
thropology No. 42. Yale University, New Haven.
Williams, Wilma B., and Bert Mowers
1977 Markham Park Mound No. 2, Broward County, Florida.
The Florida Anthropologist 30(2):56-78.
APPENDIX A: Pumice from archaeological sites in Florida.
Site No. Site Name Description Count Context Reference or Collection
8BD20 Broward No. 2 abrader 2 Glades II-III Laxson (1962:9)
8BD59 Broward No. 6 2+ Glades I-III Laxson (1970:154)
8BD183 Markham Park 2 5 Williams and Mowers (1977:68,72)
8BR56 Grant teardrop-shap- 1 St. Johns lib Sears 1958; FLMNH 94137
ed w/ one flat
side, grayish
pink color
8BR136 Jungerman small frag. 1 St. Johns I-II Jordan 1963:16; FLMNH 96335
8IR24 Higgs 1+ A.D. 1715-1716 Smith (1956:88); FBAR 75.62.10.1;
FBAR 75.62.102.1
Sebastian Inlet 2 FLMNH A7464
Midden
8DA23 Arch Creek 8 Miami-West India Archaeological Soci-
ety 1975:Table 2, Figure 6)
8DA93 Lehigh abrader 1 Laxson (1959a:6)
8DA48 OpaLocka 1 6 Glades I-II Goggin(n.d.:518); Willey (1949:86)
8DA9 Snapper Creek Goggin(n.d.:518)
8DA21 Surfside perforated 1 Goggin (n.d.:518); Willey (1949:82)
8GL13 Mound A, Mound 8 Glades I Steinen (1982:Figure 6.2, 98); FLMNH
B, A-B pond cor- collections
plex, Fort Center
80B29 Fulford Goggin (n.d.:518)
8LE853B Governor Martin hand-sized 1 A.D. 1539-1540 Ewen and Hann (1998:71-72)
pumice, smooth
on 2 surfaces
8M017 Upper Matecumbe facets from 3 Goggin (n.d.:518); Goggin and Sommer
abrading __ (1949:67-68; Plate 8f)
8M033 Bear Lake perforated 2+ Glades I-II Goggin (n.d.:518, 1950:243); Griffin
_(1988:211); FLMNH collections
8M025 Key Largo 1 plummet forms 2+ Glades lia Goggin (n.d.:518); FLMNH collections
8M038 Cape Sable 2 Goggin (n.d.:518)
8PB40 Belle Glade mid- rubbing or 1 Willey (1949:37)
den smoothing wear
8PB214 Singer Island 1 Glades III Dickel (1988:45)
8PB220 Whitebelt Circle- abrader 1 Glades I Wheeler (2000b:7, Fig. 8); FBAR
Ditch 99.65.7.1
8PB100 Boynton Mounds 2+ Glades I-III Jaffee 1976:146
8LL54 Wightman private collections, Jerald T. Milanich,
personal communication, 2003
8MO15 Indian Key American terri- FBAR 72.20.534.1; FBAR 72.20.555.1;
trial FBAR 72.20.888.1; FBAR 72.20.832.1
8M02 Stock Island 2+ Glades I-II FBAR 74.2.10.8; Eyster 1986
8BR82 De Soto Grove B St. Johns II FBAR 94.123.135.1
8DA6519 Deering Estate 3 Glades IIIc FBAR 00.14.39.8; FBAR 00.14.47.1
Midden
8SL37 Shark Butchery 1 Glades I-II FBAR 76.209.1.1
8SL1146 Herman's Bay several clasts 3 Glades II-III Driscoll and Wheeler 2004:44
8SL1173 Pavilion small frag. 1 Glades II Wheeler 2003:27
WHEELER
PUMICE ARTIFACTS FROM THE MIAMI CIRCLE
Site No. Site Name Description Count Context Reference or Collection
8DA411 Honey 2 Glades HII HMSF 2471; HMSF 2942.1
Hill/Stadium
8DA1034 Sutton 1 Glades II-HI HMSF 1874.1
8DA1064 Custom House 1 Glades I late HMSF 427.5
8DA11 Granada 4 Griffin 1982:69
8DA11 Granada HMSF 1018.18
8DA98 Brickell Point HMSF 865.3; HMSF 582.1; HMSF
(Sheraton compo- 559.1; HMSF 687.1; HMSF 492; HMSF
nent) 896
8DA12 Brickell Point abrader 1 Laxson (1959b:62)
8DA12 Brickell Point (Mi- 121 Glades I this report
ami Circle compo-
nent)
8HG678 Blueberry small ball I KVAHC
80K239 Wynnhaven Beach clasts with 26 Santa Rosa- Panamerican Consultants, Inc. mitigation
modifications, Swift Creek,
including cup- Weeden Island
shaped depres-
sions, grooves,
and flat
surfaces
8CR1 Chokoloskee Key effigy of human 1 Moore 1907:463
face/head (?)
8SJ48 North Beach __St. Johns II Ashley and Rolland 1997:25
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
GEOCHEMICAL AND PETROLOGIC CHARACTERIZATION OF PUMICE ARTIFACTS FROM
THE MIAMI CIRCLE-BRICKELL POINT ARCHAEOLOGICAL SITE PLUS OTHER SITES IN
FLORIDA POTENTIAL PROVENANCE LOCATIONS
STEPHEN A. KISH
Department of Geosciences, Florida State University, Tallahassee, FL 32306-4100
E-Mdil: kish@gly.fsu.edu
Introduction
Pumice is a light-colored, highly vesicular igneous rockthat
normally has sufficient buoyancy to float on water. The
formation of pumice is associated with the violent eruption of
volcanoes, features that are not present in Florida. Wheeler
(2006, this issue) reported that at least 121 pumice artifacts
have been recovered from excavations at the Miami Circle-
Brickell Point archaeological site (8DA 12) and there are at least
30 other sites in southern Florida that contain a limited number
of pumice artifacts (Figure 1). Several pumice artifacts (-3-4)
also were discovered on the northwestern Gulf Coast ofFlorida
at the Wynnhaven Beach site (80K239) during recent site
studies (Ryan J. Wheeler, personal communication). A single
location, the Governor Martin site (8LE853B), in the inland
portion of the Florida Panhandle, near Tallahassee, Florida also
produced a single pumice fragment. A majority of the pumice
artifacts appear to be associated with the Glades I period
(Wheeler 2006), however, material from younger sites has
been discovered, including early European contact period
locations such as Governor Martin site (DeSoto winter encamp-
ment); the early eighteenth century Higgs site and the early
nineteenth century Indian Key site (Wheeler 2006). Radiocar-
bon (4C) dating of charcoal from the Miami Circle and
adjacent Brickell Point Sheraton site, yield calibrated ages in
the range of BC 500 AD 400 (Carr and Ricisak 2000). A
similar range of 14C ages has been obtained from Lucina shells
taken from different stratigraphic intervals at the site (Widmer
2004). At the Miami Circle site no other types of volcanic
material, such as obsidian, dense rhyolite, or basalt, have been
discovered. Ground stone celts found at the Brickell Point
excavations are composed of dark, aphanitic igneous rock that
has a basaltic composition, however, the material most likely
came from Mesozoic intrusive (diabase) dikes exposed in the
Piedmont region of Georgia (Dixon et al. 2000).
As a working hypothesis, it is assumed that the pumaceous
material found at archeological sites in Florida is derived from
oceanic drift material that washed ashore on beaches and may
have been locally redistributed to inland communities. In recent
times (1950s present) a limited number of pumice clasts have
washed ashore along the Atlantic coastline of Florida (Bryan
1972; Dennis 2000), however, the presence of such a large
number of pieces of this exotic geologic material at the Miami
Circle site may indicate that a relatively massive, short-term
influx of pumice occurred along this portion of the shoreline of
southeast Florida approximately 2000 years ago, or perhaps
there was a long-term interest by native inhabitants in collecting
sporadic occurrences of this material for use as tools or for
ornamental purposes.
The purpose of this study is to provide information on the
physical and chemical characteristics of selected pumice
samples for the Miami Circle site and other selected sites in
Florida. Specific objectives ofthe study are to: (1) characterize
the petrology (rock type) of the pumice artifacts; (2) use the
mineralogy and chemistry of the samples to test for the homo-
geneity of samples at one site and petrologic diversity between
sites, and (3) test for the ultimate source of the pumice by
comparing the observed petrologic and chemical characteristics
of pumice artifacts with pumice samples collected from areas of
recent volcanic activity in the circum-Atlantic, Caribbean and
material from areas adjoining the western Gulf of Mexico.
Sample Selection
Twenty-two samples were selected for petrographic and
geochemical studies. Five samples were selected from the
Miami Circle site (8DA12). Pumice artifacts from sites in close
proximity to the Miami Circle also were analyzed. These
included two samples from the Brickell Point site (8DA98) and
one from the Granada site (8DAll). An additional eleven
pumice samples were analyzed for other sites in southern
Florida and the Keys. Three pumice samples from the
WynnhavenBeach (80K239) site, located on the Gulf Coast of
northwestern Florida, also were analyzed. The locations of
these sites, plus the locations of all other archeological sites
known to contain pumice artifacts are shown in Figure 1. The
specific site names and individual sample repository numbers
are listed in Table 1.
Sample Preparation
Individual pumice samples were cut into slabs with a
diamond saw to prepare the material for petrographic and
geochemical analyses. The fresh surfaces of the pumice were
observed using a lowpower, binocular microscope andphysical
characteristics (color, vesiculation and presence ofpheoncrysts)
were noted. Three samples were selected for scanning electron
microscopy (SEM) studies. A small chip from each of these
samples was ultrasonically cleaned in acetone and then pro-
cessed using standard SEM procedures. Petrographic thin
sections for a selected subset of samples were prepared by
vacuum impregnation ofthe porous rock with a blue dye epoxy,
VOL 59(3-4) THE FLORIDA ANTHROPOLOGIST SEPTEMBER-DECEMBER 2006
VOL. 59(3-4)
THE FLORMDA ANTHTFROPOLOU;ST
SEPTEmBER-DEcFMmER 2006
THE FLORIDA ANTHROPOLOGIST 2006 VOL. 59(34)
Figure 1. Location of pumice artifacts selected for petrographic and geochemical analyses. The Miami Circle, Brickell Point,
and Granada location symbols overlap at the scale of this map. Pumice location map modified from Wheeler (2006). Symbols -
Brickell Point BP, Bear Lake 1 -BL, Custom House CH, Dearing Estate DE, Fort Center FC, Granada GR, Higgs Site -
HG, Indian Key IK, Miami Circle MC, Shark Butchery SB, Sutton Site ST, Whitebelt 1 WB, Wynnhaven Beach WH.
then mounting the impregnated chip onto a glass thin section
blank, then grinding and polishing the thin section to a standard
thickness of 30 pnm. Five to ten gram sample slabs were
crushed and split into two fractions. One fraction was sieved,
and the 185-210 pn size fraction was ultrasonically cleaned in
distilled water, air dried and then used for index of refraction
measurements. The other portion of the crushed sample was
pulverized to <300 mesh in preparation for geochemical
analyses.
Samples were prepared for geochemical analyses byheating
a mixture ofpulverized whole rock and lithium metaborate flux
in a graphite crucible at 1050' C for 15 minuets. The fluxing
procedure insures dissolution of refractory minerals such as
zircon. The flux bead was then dissolved into 100 ml of five
percent high purity nitric acid. The dissolved sample was split
into two solutions. One solution was diluted for major element
analyses. The second undiluted solution was used for trace
element analyses. Major element analyses were performed
using a Jarrel Ash Atom Comp Model 975 emission spectro-
graph. The accuracy of internal laboratory standard solutions,
used for calibration ofthe emission spectrography, was checked
against solutions prepared from United States Geological
Survey whole rock standards. The estimated precision for
major element analyses is one percent. Accuracy for major
Explanation
Pumice Artifact Sites
Analyzed Sites
Other Sites
0 50 100 200 Kilometers
0 25 50 100 150 Miles
0 25 50 100 150 Miles
2006 VOL. 59(3-4)
THE FLORIDA ANTHROPOLOGIST
GEOCHEMICAL STUDIES OF PUMICE
Table 1. Physical Characteristics of Florida Pumice Artifacts.
Sample Location Site Munsell Vesiculation Alteration
Number Density' Properties2 Color'
Southern Florida
MDC 1.567.6b Miami Circle 8DA12 <1.0 M N9 ME
MDC 1.166.4 Miami Circle 8DA12 <1.0 VW 10YR 8/1 VFE
MDC 1.222.11a Miami Circle 8DA12 <1.0 M 10YR9/2 ME <1mm
MDC 1.472.4 Miami Circle 8DA12 <1.0 M 10YR 9/2 VCE
MDC 1.893.1 Miami Circle 8DA12 <1.0 M N9 VFT <1mm
HMSF 896 Brickell Point 8DA98 <1.0 M 10YR 9/2 VFT 3mm
FBAR 75.62.10.1 Higgs Site 8IR24 <1.0 VW 10YR 9/2 ME Imm
FBAR 72.20.534 Indian Key 8MO15 <1.0 M 10YR 9/1 VFT
FLMN A16262a Fort Center 8GL13 <1.0 VW N9 FT <1mm
FLMN 98576 Bear Lake 1 8M033 <1.0 M 10YR 3/1 ME 3mm
HMSF 582.1 Brickell Point 8DA98 >1.0 M 10YR 5/1 CT
HMSF 1018.18 Granada 8DAll <1.0 M N9 FT
HMSF 2942.1 Honey Hill 8DA411 <1.0 M 10YR 9/1 FT 1-2mm
HMSF 427.5 Custom House 8DA106 1.0 S 10YR 2/1 S <1mm
FBAR 74.2.10.8 Stock Island 8M02 <1.0 VW 10YR 8/1 VFT 1-3mm
FBAR 00.14.47.1 Deering Estate 8DA651 <1.0 M 10YR 7/3 MT >15mm
FBAR 76.209.1.1 Shark Butchery 8SL37 <1.0 W 10YR 8/1 FT
FBAR 99.65.7.1 Whitebelt 1 8PB220 <1.0 M 10YR 8/4 MT >15mm
HMSF 1874.1b Sutton Site 8DA103 <1.0 S 10YR 6/4 MT 5-10mm
Northwestern Florida
EGLIN 056 Wynnhaven 80K239 <1.0 W 5YR4/1 CT <1mm
EGLIN 198 Wynnhaven 80K239 <1.0 M 7.5 YR 6/1 ME
EGLIN 742 Wynnhaven 80K239 <1.0 M 10YR 7/3 FT 5-10mm
1. Density reported relative to water (1.0 gm/cm3).
2. Relative magnetic susceptibility as measured by the attraction of a rare earth magnet to individual specimens. Strong (S) attraction
takes place at > 5mm; Moderate (M) attraction takes place at 3-5 mm; Weak (W)- attraction takes place at less than 3 mm; Very weak
(VW) magnet must touch sample to be attracted.
3. Munsell Color color of fresh surface as viewed under indoor fluorescence lighting.
4. Vesicle size (VC) very coarse 3-5 mm; (C) coarse 2-3 mm; (F) fine 0.5-1mm; (VF) very fine (<0.5 mm).
(S) Scoria coarsely vesicular basaltic rock. E equidimensional, T tubular.
element analyses is estimated to be approximately 1-3 percent.
The water content of the samples was estimated by measuring
the weight loss of a sample when heated to approximately
9500C for 90 minutes. This value is reported as "loss on
ignition" (LOI) in Table 4. Trace element analyses were
performed using a Perkin-Elmer Elan 600 ICP mass spectrome-
ter using standardized laboratory solutions. Estimates of the
accuracy and precision for trace element analyses range
between 5-10 percent depending upon the absolute concentra-
tion of the trace element in each sample.
The index of refraction for pumice glass was measured as a
means of estimating the SiO2 concentration in the glass compo-
nent of each sample. Very fine, pulverized pumice (185-210
[im) was mounted on a microscope slide and immersed in
liquids with known refractive indices (Cargille Certified
Refractive Index Liquids, Series A). Measurements were made
using plane polarized light at a wavelength of 589 nm. All
measurements were made at 230C. The index of refraction for
the pumice glass was first roughly matched to index liquids
using the "relief' of the sample versus the index liquid and then
more precisely matched using the Becke line method. The
refractive index of each sample was matched to the nearest
0.001 based on the available range of the refraction liquids.
Samples that contained visible phenocrysts, as well as a
limited number of purely vitric (glassy) samples also were
characterized by petrographic thin section analysis. The
volume percentage of different components in the thin section
sample ("mode") was measure by standard point counting
procedures. Approximately 200 points were counted for each
sample. Modal analysis provided quantitative values for the
percent void space, glass, and mineral percentage of the
analyzed samples.
KISHi
THE FLORIDA ANTHROPOLOGIST 2006 VOL 59(34)
A E4EE
1 cm M cm5
wlI in h
In in
CM CC
EIUTnJ
*^Jcm ^ii cm
in In
.INI.4
D
Figure 2. Pumice textures: a) very fine, uniform vesicular texture (sample FBAR 72.20.534.1); b) fine vesicular texture with
abundant 1-3 mm plagioclase phenocrysts (sample HMSF 896); c) coarse, vesicular texture, with abundant mafic phenocrysts
(sample MDC 1.222.11a); d) Very coarse, vesicular texture (sample MDC 1.893.1).
Physical Characteristics
Macroscopic and SEM Studies
Wheeler (2006) has provided a detailed analysis covering
the external appearance of pumice material, including parame-
ters such as physical size, shape and coloration. The outside
surfaces of all pumice samples are stained with organic mate-
rial, producing a dark, brownish- gray color. This color is not
characteristic of a fresh surface. The interior, fresh surface of
most pumice is nearly pure white (Figure 2). Igneous rocks that
are very lieht in color are said to have a "felsic" color and
composition, while dark-gray to black igneous materials are
classified as having a "mafic" color and composition. The
limited number of mafic samples available for this study were
black or variegated, dark greenish-gray and purple. Thin(<0.5-
3 mm) weathering/hydration rims are present on the surface of
all samples, but the interiors of the pumice clasts appear very
fresh and are indistinguishable from modem samples.
The size of gas cavities or vesicles varies between samples;
two distinct textural types appear to be present. Most of the
samples that were studied have very smooth surfaces composed
of very small vesicles plus a few larger vesicles (Figure 2a).
The remaining samples have a coarse, sponge-like, irregular
THE FLORIDA ANTHROPOLOGIST
2006 Vot. 59(3-4)
Table 2. Selective Petrographic Modes of Florida Pumice Artifacts.
Sample Location Void Glass Mt. Plag. Qtz. Bt. Cpx. Glom. Other Count' Density2
Southern Florida
MDC 1.567.6b Miami Circle 68.4 25.0 0.5 1.4 0.0 1.4 0.0 2.4 0.5 212 0.75
MDC 1.166.4 Miami Circle 71.1 20.1 Tr. 0.5 2.5 0.5 0.0 5.4 0.0 204 0.68
MDC 1.222.11a Miami Circle 77.8 7.4 Tr. 7.4 2.0 1.0 0.0 0.0 2.0 203 0.55
MDC 1.472.4 Miami Circle 74.5 20.2 Tr. 0.5 0.0 0.5 0.0 3.8 0.0 208 0.60
MDC 1.893.1 Miami Circle 70.8 27.4 Tr. 1.4 0.0 0.5 0.0 0.0 0.0 219 0.69
HMSF 896 BrickellPoint 75.1 16.6 Tr. 0.6 1.1 0.0 0.0 6.1 0.0 181 0.59
FBAR 75.62.10.1 Higgs Site 66.4 29.9 Tr. 1.4 0.0 0.0 0.0 2.4 0.0 211 0.80
FBAR 72.20.534.1 Indian Key 45.0 24.5 Tr. 0.0 0.0 0.0 0.0 0.0 0.0 216 0.59
HMSF 427.5 Custom House 45.5 16.5 Tr. 6.5 0.0 0.0 1.5 0.0 0.0 711 0.68
Northwestern Florida
EGLIN 056 Wynnhaven Beach 58.6 14.4 Tr. Tr. 0.0 0.0 0.0 0.0 0.0 309 1.15
Explanation for table labeling. Mt. = magnetite; Plag.= plagioclase; Qtz. = quartz; Amp. = amphibole; Bt. = biotite; Cpx. = clinopyroxene; Glom. = glomerocrysts; Tr. = trace amounts, observed
but not counted. Modes are reported in volume percent.
1. Total number of points used to calculate petrographic modes.
2. Calculated density based upon the ration of void space to glass and minerals using the following density information from Daley et al. (1966): rhyolitic glass (2.37), andesitic glass
(2.47), basaltic glass (2.77). Compositional information for each sample was obtained from Figure 5.
THE FLRD NmoooIsI 200IVO.II(3I
Figure 3. Scanning electron microscope images of
Miami Circle pumice artifacts: a) highly elongate or
"tubular" vesicles; b) variable vesicle size associated
with coarsely vesicular pumice; c) enhedral biotite
phenocrysts within vesicular pumice glass; d) high
magnification of a finely vesicular pumice sample; e)
Low magnification of the same sample as in Figure 3d.
A scale bar is provided in the lower right-hand portion
of each image.
surface texture, produced by large (3-5 mm) vesicles (Figure
2d). Vesicles within individual samples are often highly
elongate or tubular in shape (Figure 3a). Other samples,
particularly those with coarser vesicles, exhibit more
equidimensional form (Figure 3b).
Fine- (1 mm) to coarse-grained (5 mm) mineral pheoncrysts
are present in most samples. Large plagioclase pheoncrysts
were noted in several samples (Table 2; Figure 2b). Biotite
phenocrysts (Figure 3c) are present in all samples, however,
their overall abundance is less than two percent. Samples that
have an intermediate composition andesitee) contain relatively
abundant (10-15 %) mafic phenocrysts that include biotite,
amphibole, and altered pyroxene (?) (Figure 2c). Nearly all the
pumice is moderately magnetic. This indicates the presence of
small grains of magnetite within the sample (Table 1).
Scanning electron microscopy imaging of pumice clearly
indicates the two textural types of vesicle shapes. Approxi-
mately fifty percent of the analyzed samples have highly
elongated, "tubular," vesicles (Figure 3a). The remaining
samples have equidimensional vesicles the may vary in size
(Figure 3 b,d, e). Biotite phenocrysts (Figure 3c) are present as
euhedral grains in most samples.
Modal thin section analyses of a selected set of pumice
samples are reported in Table 2. All the analyzed samples have
a relatively high volume of void space, ranging from 59 to 78
volume percent. Using the appropriate composition for the
samples (Figure 5), the bulk densities for these samples are
calculated to range from 0.40 to 0.70 am/cm3 (Table 2). One
sample has a calculated density greater than 1.0, but the
samples does float in water.
The glass of most pumice samples is normally colorless and
clear. The glassy vesicle walls are very thin, averaging less
than 1-10 pm thick. (Figure 3d). Large, aggregate grains of
plagioclase are common in many samples. Biotite (Figure 3c)
is common in both rhyolite and andesite samples, while
hornblende and cloudy, medium-green aggregates or
glomorocrysts of altered pyroxene (?) or hornblende are present
in low silica rhyolite and andesite. Both biotite and hornblende
are clear and do not exhibit evidence of alteration on the rims
of grains. Mafic samples have either a scoriaceous texture or
have flattened vesicles. The mafic sample from the Custom
House site (8DA1064) is characterized by large, evenly
distributed vesicles that are surrounded by a dark-brown, clear
glass. In addition to glass, the Custom House samples contains
coarse-grained, light brown, subhedral clinopyroxene pheno-
crysts and very fine-grained, euhedral plagioclase. The mafic
pumice sample from the Wynnhaven Beach site (80K239) is
purely vitric, and is characterized by 2-5 mm bands with
alternating dark-green, medium-green and brownish-purple
coloration.
The observed mineral assemblage for most felsic samples is
relatively simple, consisting of plagioclase + biotite amphi-
bole quartz + magnetite (Table 2). This mineral assemblage,
along with the absence of mineral phases associated with
peralkaline felsic igneous rocks and the relatively high silica
content (see next section) of the samples are consistent with
classification of a major portion of the samples as being calc-
alkaline rhyolites.
Index of Refraction Pumice Glass
The optical index of refraction of natural glass is a function
of bulk chemical composition and the post-cooling hydration.
Geologists have used fused rock powder as a means for rapid
categorization of the approximate silica content of fine-grained
volcanic rocks (McKee 1968). While geologists have noted a
high correlation between the SiO2 content ofnatural glasses and
the index of refraction (Tilley 1922), direct use of index of
refraction for the precise determination of silica content is not
possible due to variation in index of refraction by factors other
than SiO2 content. These factors include variation of index of
refraction with varying iron and alkali contents for glass having
the same SiO, concentration and post-cooling hydration of
glass, which can produce a higher index ofrefraction relative to
the same material in an unhydrated state (Ross and Smith
1955). However, the index of refraction method can be readily
used as an aid in "fingerprinting" the provenance of natural
glass samples and in comparing the similarity of material
between different archeological sites. This approach, combined
with geochemical information, has been used as an aid in
obsidian provenance studies. The index of refraction of a
volcanic glass also can be used to test for chemical alteration
and hydration of material following eruption. Departure of a
predicted SiO2 content, derived from index of refraction
measurements, relative to chemically determined values, is an
indication of hydration alteration. Index of refraction measure-
ments on Florida pumice can be used for both provenance
THE FnoniA ANTintOPOUo~isT
2006 VOL. 59(3-4)
1(15K GEOCHEMICAL STUIMES OF PUMICE
Table 3. Index of Refraction Measurements of Florida Pumice Artifacts.
Sample Location Refractive Predicted Predicted Measured Residual
Index SiO2 2 R.I. 3 Si02 4
Southern Florida
MDC 1.567.6b Miami Circle 1.504 71.8 1.509 71.1 0.7
MDC 1.166.4 Miami Circle 1.498 73.5 1.493 75.1 -1.5
MDC 1.222.11a Miami Circle 1.505 71.5 1.557 62.5 9.0
MDC 1.472.4 Miami Circle 1.499 73.2 1.493 75.8 -2.5
MDC 1.893.1 Miami Circle 1.501 72.7 1.490 76.0 -3.3
HMSF 896 Brickell Point 1.499 73.2 1.488 75.2 -2.0
FBAR 75.62.10.1 Higgs Site 1.489 76.2 1.489 77.0 -0.8
FBAR 72.20.534.1 Indian Key 1.499 73.2 1.491 76.0 -2.7
FLMNH A16262a Fort Center 1.496 74.1 1.487 76.7 -2.6
FLMNH 98576 Bear Lake 1 1.498 73.5 1.497 74.9 -1.4
HMSF 582.1 Brickell Point 1.496 74.1 1.501 72.2 1.9
HMSF 1018.18 Granada 1.500 72.9 1.505 71.0 2.0
HMSF 2942.1 Honey Hill 1.503 72.1 1.499 72.8 -0.8
HMSF 427.5 Custom House 1.619 46.9 1.639 47.0 -0.1
FBAR 74.2.10.8 Stock Island 1.500 72.9 1.491 75.1 -2.2
FBAR 00.14.47.1 Deering Estate 1.497 73.8 1.491 75.0 -1.1
FBAR 76.209.1.1 Shark Butchery 1.502 72.4 1.487 75.6 -3.2
FBAR 99.65.7.1 Whitebelt 1 1.497 73.8 1.475 62.4 11.4
HMSF 1874.1b Sutton Site 1.501 72.7 1.554 60.3 12.3
Northwestern Florida
EGLIN 056 Wynnhaven 1.514 69.0 1.549 64.6 4.4
EGLIN 198 Wynnhaven 1.498 73.5 1.597 51.6 21.9
EGLIN 742 Wynnhaven 1.498 73.5 1.519 71.7 1.81
1. Index of refraction measured at a wavelength of 589 nm using an Edmund Optics narrow bandpass interference filter and Cargille Laboratories
refractive index liquids (A-X series) at a temperature of 230C. The refractive index of individual samples is reported to the nearest 0.001.
Samples with an odd number in the thousandth place (e.g. 1.501) are based upon bracketing values between higher and lower refractive index
liquids.
2. SiO2 weight percent values based on the predicted median value derived from SiO2" refractive index curves reported by Fisher and Schmincke
(1984).
3. Theoretical value for an anhydrous glass with the same major element composition as the sample (Table 4). Theoretical refractive index
values are based upon the empirically derived coefficients reported by Church and Johnson (1980).
4. Whole rock SiO2weight percent (Table 4), recalculated to an anhydrous whole rock equivalent.
5. Residual values represent the difference between the predicted SiO2 versus the chemically measured SiO2 content of a sample (in weight
percent).
fingerprinting and to evaluate the possibility that these samples
may be chemically altered.
Results
The glass from light-colored pumice has relatively low
indices of refraction, ranging from 1.489-1.505 (Table 3).
Predicted SiO2 concentrations, based on refractive index, range
from 69.7-76.2 weight percent SiO2and are within the predicted
index of refraction range for rhyolite (Church and Johnson
1980). The SiO, values derived from refraction measurements
(Table 3) are very close to the measured SiO2 values for the
same samples. This indicates that the glass within the pumice
has not been significantly altered by hydration. Three samples
KISH
GEOCHEMICAL STUDIES OF PUMICE
Table 4. Whole Rock Major Element Composition of Florida Pumice Artifacts.
Sample Location SiO2 A1203 Fe2,Ot MgO CaO Na20 K0O TiO2 P20, MnO LOI SUM"
Southern Florida
MDC 1.567.6b Miami Circle 67.80 13.70 2.54 0.58 2.37 4.01 3.33 0.39 0.45 0.08 4.60 99.97
MDC 1.166.4 Miami Circle 69.28 11.58 1.27 0.20 1.88 3.80 3.52 0.11 0.50 0.06 7.50 99.81
MDC 1.222.11a Miami Circle 60.74 15.92 6.51 2.51 6.45 2.97 1.27 0.48 0.21 0.14 2.60 99.83
MDC 1.472.4 Miami Circle 72.72 12.72 1.13 0.18 1.56 3.49 3.70 0.12 0.16 0.07 3.90 99.86
MDC 1.893.1 Miami Circle 73.00 12.73 0.88 0.16 1.15 3.58 4.07 0.12 0.23 0.06 3.80 99.88
HMSF 896 Brickell Point 71.42 12.72 0.94 0.15 1.36 3.45 3.75 0.11 0.88 0.07 4.90 99.87
FBAR 75.62.10.1 Higgs Site 74.78 12.23 0.92 0.26 0.90 3.40 4.20 0.14 0.10 0.06 2.80 99.89
FBAR 72.20.534.1 Indian Key 72.20 12.37 0.82 0.20 1.65 3.57 3.83 0.10 0.13 0.06 4.80 99.84
FLMNH A16262a Fort Center 72.77 11.83 0.59 0.11 0.97 3.61 3.97 0.10 0.07 0.06 5.10 99.29
FLMNH 98576 Bear Lake 1 72.86 12.63 1.33 0.45 2.00 3.93 2.78 0.20 0.15 0.05 2.70 99.19
HMSF 582.1 Brickell Point 70.99 12.86 1.95 0.29 2.30 3.82 4.10 0.28 0.36 0.03 1.70 98.77
HMSF 1018.18 Granada 67.83 14.57 1.60 0.43 2.92 4.94 2.05 0.25 0.26 0.05 4.40 99.39
HMSF 2942.1 Honey Hill 67.56 13.07 1.61 0.39 2.47 3.73 2.32 0.22 0.78 0.04 7.20 99.49
HMSF427.5 CustomHouse 46.40 12.71 12.81 6.12 11.03 2.21 1.03 *4.45 *1.72 0.16 1.30 99.99
FBAR 74.2.10.8 Stock Island 71.60 11.92 0.74 0.36 1.95 3.33 4.00 0.13 0.25 0.06 4.60 99.05
FBAR 00.14.47.1 Deering Estate 71.92 13.32 0.88 0.25 1.06 3.30 3.82 0.13 0.16 0.08 4.00 99.03
FBAR 76.209.1.1 Shark Butchery 73.16 12.54 0.62 0.14 1.05 3.56 4.24 0.10 0.18 0.06 3.20 98.25
FBAR 99.65.7.1 Whitebelt 1 55.63 14.39 1.61 0.27 3.72 2.95 2.65 0.16 ** 0.05 10.8 **
HMSF 1874.1b Sutton Site 59.30 17.76 6.36 2.20 6.50 3.33 0.94 0.46 0.79 0.16 1.70 99.52
Northwestern Florida
EGLIN056 Wynnhaven Beach 64.89 17.57 8012 2.58 1.04 1.76 3.51 0.58 0.25 0.09 <0.1 100.48
EGLIN 198 Wynnhaven Beach 50.47 16.88 11.18 2.51 6.11 4.5 2.43 2.36 1.14 0.22 1.9 99.88
EGLIN 742 Wynnhaven Beach 69.54 12.31 5.8 0.23 1.26 3.36 3.44 0.16 0.75 .12 3.4 100.37
All major element analyses are reported in weight percent. LOI = weight loss on ignition.
t Total iron as Fe2O.
tt Sum includes LOI and trace element barium (Table 7) recalculate as weight percent BaO.
* Replicate values by ICP-MS as Florida State University; TiO2 = 3.60 wt. %, P205 = 1.92 wt. %.
** Value omitted due to analytical problems.
1(15K GEOCHEMICAL STUDIES OF PUMICE
50
45
40--
1.480
1.540
1.600
1.660
Refractive index [n]
Figure 4. Plot of the measured refractive index of glass in pumice samples versus the calculated anhydrous SiO, concentration
for the bulk rock in the same sample. The curve shown in the figure represents the observed refractive index values for a wide
range of volcanic glass samples (Fisher and Schmincke 1984). The samples that fall outside the boundaries of the curve have
an andesitic composition and are characterized by having moderately high (>10 %) concentrations of mafic and plagioclase
phenocrysts. Samples for the Whitebelt and Wynnhaven sites are not shown due to possible alteration of the sample material.
have predicted SiO2 contents that are much higher than their
observed bulk-rock silica composition (Figure 6). Petrographic
examination of one of these samples (MDC. 1.222.11 a) reveals
the presence of>15 percent mafic mineral phenocrysts, which
include both biotite and amphibole. These mineral phases have
low SiO2 contents. A mixture of high silica glass and low silica
mafic minerals will produce the observed bulk-rock SiO,
composition. The index of refraction for the glass from the
mafic sample (EGLIN 056) taken from the Wynnhaven Beach
site is far lower than would be indicated by it bulk chemical
composition. The glass in this sample appears very cloudy and
contains a very high volume ofmicrolites (microcrystals). The
presence ofthe microlites may explain the discrepancy between
observed and predicted SiO, content for the sample.
Geochemical Characterization
Major Element Chemistry
Since most volcanic rocks are either very fine grained or are
composed of a large percentage of glass, a complete, conven-
tional lithologic classification and characterization of volcanic
material based upon mineralogy is not possible. Most studies
of volcanic rocks make use of the chemical composition of
individual samples or a suite of cogenetic samples to provide a
more exact characterization ofthe volcanic material. Petrologic
classification methods that utilize the major element chemistry
can relate the major element chemical composition of the
volcanic rock to a hypothetical mineral assemblage know as
"normative minerals" (Irvine and Baragar 1971). Volcanic
KISH
GEOCHEMICAL STUDIES OF PUMICE
TEE FLORIDA ANTHROPOLOGIST 2006 VOL. 59(3-4)
50 55 60 65 70 75
Si02 [wt%]
Figure 5. Total alkali-silica (TAS) classification for Florida pumice artifacts. Explanation for abbreviations: T Trachyte;
TA Trachyandesite; BT Basaltic trachyandesite; TBA Trachybasalt; PH Phonlite; TPH Tephriphonolite; PTA -
Phonotephrite and FD --Foidite. Classification diagram from LeMaitre et al 2002. Heavy line represents the boundary be
subalkaline (SA) and alkaline (A) rocks (Irvine and Baragar 1971). Open square symbols represent pumice samples from
southern Florida. Filled circles represent pumice samples from the Wynnhaven Beach site located in northwestern Florida.
samples also can be classified by their position within
compositional fields plotted on bivariate oxide-oxide major
element diagram (Le Maitre 2002). For most geochemical
classification schemes, the total major element composition of
a sample is recalculated to an anhydrous total of 100 percent,
without the inclusion of volatile compounds such as water and
carbon dioxide. This adjustment places all samples in the same
relative comparison frame, eliminating the dilutional effects of
the volatile components (Irvine and Baragar 1971). A. small,
moderately weathered sample from the Whitebelt 1 site (FBAR
99.65.7.1) had an unacceptably high LOI value (10.8 wt. %)
and was not included for classification.
For this study, the total alkali-silica diagram (TAS) of Le
Bas et al. (1986) was used as the primary means for chemical
classification of Florida pumice artifacts. The samp es also
were characterized using the K2O versus SiO2 diagram of
Peccerillo and Taylor (1976). This classification has additional
subdivisions of the calc-alkaline basalt-rhyolite series -- a low-
K or tholeiite series, a medium-K calc-alkaline series and a
high-K calc-alkaline series. This classification scheme can be
used to infer some information about the tectonic setting of the
volcanic material. High-K calc-alkaline volcanic activity
normally takes place in a more mature volcanic arc setting
relative to tholeiitic or medium K calc-alkaline volcanics.
Major element, whole rock analyses for Florida pumice
artifacts are reported in Table 4 and the calculated normative
minerals for the samples are reported in Table 5. All felsic
pumice samples have high normative quartz, orthoclase, and
albite, and the total normative mafic components (diopside +
hypersthene + magnetite + ilmenite) make up less than 10
weight percent of the normative composition.
Results Major Element Classification
The TAS classification diagram for the Florida pumice
samples is shown in Figure 5. Fifteen out of nineteen pumice
samples from southern Florida fall within the field of
subalkaline rhyolites. Eleven of the 15 samples also have
uniform, high SiO, concentrations (75-76 wt. %) and are
classified using the informal term "high-silica rhyolites." Four
samples have SiO2 concentrations in the range 71-72 weight
percent and are classified using the informal term "low-silica
14
12
10
8
O
6
0
z 4
2
0 -
45
THE FLOREDA ANrl[IROPOLOGIST
2006 VOL. 59(3-4)
Table 5. CIPW Normative Mineral Composition of Florida Pumice Artifacts.
Sample Location Qz Co Or Ab An Di Hy 01 Mt II Ap D.I.*
Southern Florida
MDC 1.567.6b Miami Circle 28.2 0.3 20.7 35.7 9.3 0.0 2.1 0.0 1.9 0.8 1.1 85.5
MDC 1.166.4 Miami Circle 33.6 0.0 22.6 34.9 4.5 1.7 0.2 0.0 1.0 0.2 1.3 91.1
MDC 1.222.11a Miami Circle 21.5 0.0 7.8 26.0 27.2 3.6 8.6 0.0 3.9 0.9 0.5 56.0
MDC 1.472.4 Miami Circle 36.5 0.6 22.8 30.8 7.0 0.0 0.9 0.0 0.9 0.3 0.4 90.1
MDC 1.893.1 Miami Circle 35.9 0.9 25.1 31.6 4.4 0.0 0.7 0.0 0.7 0.3 0.6 92.5
HMSF896 Brickell Point 38.2 2.8 23.4 30.8 1.0 0.0 0.7 0.0 0.7 0.2 2.2 93.4
FBAR75.62.10.1 Higgs Site 37.9 0.7 25.6 29.7 4.0 0.0 1.0 0.0 0.7 0.3 0.2 93.2
FBAR72.20.534.1 Indian Key 35.2 0.0 23.8 31.8 6.8 0.8 0.4 0.0 0.6 0.2 0.3 90.9
FLMNHA16262a Fort Center 36.6 0.0 25.0 32.5 4.6 0.0 0.5 0.0 0.5 0.2 0.2 94.1
FLMNH98576 Bear Lake 1 36.1 0.0 17.0 34.5 8.9 0.3 1.4 0.0 1.0 0.4 0.4 87.7
HMSF582.1 Brickell Point 30.2 0.0 25.0 33.4 6.0 2.4 0.0 0.0 1.5 0.6 0.9 88.6
HMSF1018.18 Granada 26.6 0.0 12.8 44.1 12.1 1.1 1.0 0.0 1.2 0.5 0.6 83.5
HMSF2942.1 Honey Hill 36.0 2.0 14.9 34.3 7.8 0.0 1.5 0.0 1.3 0.5 2.0 85.2
HMSF427.5 Custom House 3.0 0.0 6.2 19.2 22.3 18.1 14.8 0.0 3.8 8.7 4.2 28.3
FBAR74.2.10.8 Stock Island 35.3 0.0 25.1 29.9 6.1 1.9 0.2 0.0 0.6 0.3 0.6 90.3
FBAR00.14.47.1 Deering Estate 37.7 2.3 23.8 29.5 4.5 0.0 1.0 0.0 0.7 0.3 0.4 90.9
FBAR76.209.1.1 Shark Butchery 35.8 0.7 26.2 31.5 4.2 0.0 0.6 0.0 0.5 0.2 0.5 93.4
FBAR99.65.7.1 Whitebelt 1 ** ** ** ** ** ** ** ** ** ** ** **
HMSF1874.1b Sutton Site 20.2 1.4 5.7 28.9 27.8 0.0 9.4 0.0 3.8 0.9 1.9 54.8
Northwestern Florida
EGLIN056 Wynnhaven Beach 32.0 9.6 20.8 14.9 3.5 0.0 15.1 0.0 2.4 1.1 0.6 67.7
EGLIN198 Wynnhaven Beach 0.0 0.0 14.8 39.3 19.3 3.6 6.2 6.1 3.4 4.6 2.8 54.1
EGLIN742 Wynnhaven Beach 36.0 2.7 21.0 29.4 1.4 0.0 3.0 0.0 4.4 0.3 1.8 86.4
Explanation of abbreviations: Qz- quartz; Co corundum; Or orthoclase; Ab albite; An anorthoite; Di diopside; Hy hypersthene; 01 olivine; Il- ilmenite; Ap apatite. Normative
mineral compositions are based upon the Fe2O0/FeO ratios recommended by Middlemost (1989) for volcanic rocks, using the TAS subdivisions shown in Figure 8
* D.I. Differentiation Index (Thornton and Tuttle 1960). The sum of all salic normative minerals quartz+albite+orthoclase +nephlene+leucite+kalsilite).
** Normative mineral composition was not calculated due to a high LOI value.
TIlE FLOlUDA AMHhIOPOLOGIsT 2006 VOL.59(3-4)
6.0
5.0-
4.0
3.0
0
2.0-
/
7
0/
/
/
Medium K
0
LJGR
MC O
ST
Low K
60.0
70.0
SiO2 [wt%]
Figure 6. K,O versus silica classification diagram for Florida pumice artifacts. The symbol notation is the same as that used
in Figure 8. The explanation for the letter code for medium-K samples is location on Figure 1. Classification diagram
modified from Peccerillo and Taylor (1976).
rhyolites." Two samples, one from the Miami Circle site
(1.222.11a) and one from the Sutton Site (1874.1b) have
andesitic compositions. A single sample from southern Florida,
Custom House (HMSF 427.5), has a mafic, near-alkaline,
basaltic composition. Selected pumice samples from the
Wynnhaven Beach site in northwestern Florida appear to have
a much more heterogeneous compositional range, including low
silica rhyolite, dacite, and basaltic trachyandesite (Figure 5).
The TAS classification for these sample is substantiated using
the normative mineral classification of Irvine and Baragar
(1971) all samples classified as rhyolites using the TAS
diagram have less than 10 percent normative mafic minerals
and less than 20 percent normative anorthoite (Table 5) and a
high differentiation index (the sum of all "felsic" normative
mineral), which ranges above 90 percent for most samples.
The chemical compositional characteristics of pumice
samples can be refined using the K20 versus SiO2 classification
of Peccerrillo and Taylor (1976). All rhyolitic samples from
the Miami Circle site have a high-K composition (Figure 6).
The rhvolite samples form the Granada (8DA11). Honev Hill
(8DA411), and Bear Lake 1 (8M033) sites have a medium-K
composition. The andesite samples from the Miami Circle and
Sutton sites also have medium-K compositions. Mafic samples
have moderate to distinctly alkaline signatures using TAS
classification. The Custom House basalt scoria (HMSF 427.5)
has normative olivine (Table 5) and a very high concentration
of TiO2(4.45 wt. %) and P20, (1.72 wt. %). These values are
exceptionally high, so they were checked with additional
replicate analyses (see footnote in Table 4). Somewhat lower
concentrations of TiO2 (2.36 wt. %) and P20O (1.14 wt. %) are
present in the pumaceous basalt (EGLIN 198) from the
Wynnhaven Beach site along the northwestern GulfCoast. The
elevated TiO2 and P20, concentrations observed in these
samples are normally associated with alkaline mafic rocks. The
grouping of pumice material that is present on TAS and SiO2
versus KO2 diagrams also is reflected in distinct grouping ofthe
same samples on rare earth element diagrams, which is dis-
cussed in a later section. The chemical classification names for
all the samples are listed in Table 6. High-silica, high-K,
rhyolitic pumice is the most widely distributed type ofmaterial.
/
/
//
/ High K
1.0
11 9
] BL
HH
0.0
I
40.0
50.0
80.0
THE ftoitiuA ANTiEROPOLOGIST
2006 VOL. 59(3-4)
KISR GEOCREMICAL STUDIES OF PuMICE
Table 6. Volcanic Rock Names for Florida Pumice Artifacts.
Location TAS Name' K Alkalinity2
Sample
Southern Florida
MDC 1.567.6b
MDC 1.66.4
MDC 1.222.11a
MDC 1.472.4
MDC 1.893.1
HMSF 896
FBAR 75.62.10.1
FBAR 72.20.534.1
FLMNH A16262a
FLMNH 98576
HMSF 582.1
HMSF 1018.18
HMSF2942.1
HMSF 427.5
FBAR 74.2.10.8
FBAR 00.14.47.1
FBAR 76.209.1
FBAR 99.65.7.1
HMSF 1874.1B
Northwestern Florida
EGLIN 056
EGLIN 198
EGLIN 742
Miami Circle
Miami Circle
Miami Circle
Miami Circle
Miami Circle
Brickell Point
Higgs Site
Indian Key
Fort Center
Bear Lake 1
Brickell Point
Granada
Honey Hill
Custom House
Stock Island
Deering Estate
Shark Butchery
Whitebelt 1
Sutton Site
Wynnhaven Beach
Wynnhaven Beach
Wynnhaven Beach
1 Classification of samples based upon TAS (total alkalies-silica) diagram of La Bas et al. (1986)
2 Potassium alkalinity classification based upon the diagram of Peccerillo and Taylor (1976)
All High-K rhyolite samples are also classified as being "high-silica" except for the samples
marked with an asterisk.
It is present along most of the archeological sites on the
Atlantic coastline of central and southern Florida, the adjacent
Keys and inland near Lake Okeechobee. The limited number
of analyzed, medium-K rhyolitic pumice samples appear to
come from a restricted region of sites that are located a few
kilometers north of the Miami Circle and at one site (Bear Lake
1) at Cape Sable on Florida Bay.
Additional Major Element Characteristics
Igneous rocks that have a granitic-rhyolitc composition fall
within a narrow field when plotted on a triangular diagram of
the normative minerals albite-quartz-orthoclase. Tuttle and
Bowen (1958) found that for a large number of analyzed
samples (1269) that were classified as either granites or
rhyolites most of the samples had approximately equal propor-
tions of the three normative minerals (Figure 7). Chemical
Rhyolite*
Rhyolite
Andesite
Rhyolite
Rhyolite
Rhyolite
Rhyolite
Rhyolite
Rhyolite
Rhyolite*
Rhyolite
Rhyolite
Rhyolite
Basalt
Rhyolite
Rhyolite
Rhyolite
Rhyolite
Andesite
Andesite
Basaltic
Trachyandesite
Rhyolite*
High K
High K
Medium K
High K
High K
High K
High K
High K
High K
Medium K
High K
Medium K
Medium K
Medium K
High K
High K
High K
Medium K
Medium K
High K
Shoshonite
High K
KISH
GEOCREMICAL STUDIES OF PUMICE
THE FWmDA ANTRRO1'OLOGIST 2006 VOL 59(34)
Qz
Ab Or
Figure 7. Triangular plot of the normative minerals quartz (QZ), albite (Ab) and orthoclase (Or) for Florida pumice samples
that have a rhyolitic composition. Open squars represent high siliica (>74% SiO2) samples; filled squares are low silica (70-75
% SiOz) samples; open circle is a pumice sample from the Wynnhaven Beach site in northwestern Florida. The outlined field
represent the dominant composition field for 1269 chemically analyzed rhyolite and granite samples (Tuttle and Bowen
1958).
alteration, such as silicification, potassic or sodic metasomatism
of granitic-rhyolitic rocks, will produce samples that plot
outside of this central field. The high-silica, rhyolitic pumice
samples from this study all plot within the center of Tuttle and
Bowen's "average granite" field (Figure 7). This relationship
suggests the observed chemical composition of the felsic
pumice reflects a primary, rhyolitic signature that has not been
affected by alteration.
A second measure of the chemical character and possible
alteration ofvolcanic rocks is the relative abundance of alumina
(A1203) relative to the alkali content of the rock. Shand (1927)
established that igneous rocks may be subdivided into four
categories based upon alumina saturation: alumina saturated
(A1203>Na2O+K20+CaO) or peraluminous corundumm norma-
tive); metaluminous and subaluminous (no normative corun-
dum or acmite in norm) and peralkaline rocks in which
A1,23
appear in the norm.
Unaltered, felsic calc-alkaline igneous rocks are normally
metaluminousorweaklyperaluminous. Igneous rocks thathave
been hydrothemally altered, with clays replacing the primary
silicates such as feldspars, will be highly peraluminous. The 15
felsic pumice samples analyzed for this study are either
metaluminous or weakly peraluminous (Figure 8). A single
sample from the Deering Estate site (FBAR 00.14.47.1) has a
moderate peraluminous value of 1.19. The overall alumina
saturation values of all samples are within the normal range of
calc-alkaline rocks. This also is reflected by the low values of
normative corundum (Table 5), which is associated with a
peraluminous composition.
Trace Element Chemistry
Trace element studies can be combined with the major
element information for a suite of cogenetic igneous rocks to
refine the model for magma generation and evolution or the
trace element composition of individual samples canbe used as
a "fingerprint," to infer the potential tectonic setting of the
sample. For the felsic pumice samples selected for this study
the variation in major element chemical composition is limited,
suggesting that most of the materials originated and evolved
under similar conditions. In this case, trace element character-
ization of the pumice samples can be useful for "tectonic
discrimination" (characterizingthe geological setting associated
with the volcanic source of the pumice) and as a means of
looking for additional geochemicalvariations between samples
that may indicate separate sources. A major problem in using
trace elements studies for SiO2-rich, volcanic material such as
rhyolite, is that relatively large-scale trace element variations
can be present within a single eruptive unit that has a relatively
2006 VOL. 59(3-4)
THE FLOREDA ANTHROPOL4DIST
KIsa GEOCIIEMICAL STUDIES OF Pt MICE
3.0
2.5
2.0
1.5
1.0 -
F
0.5 -4
0.5
1.0 1.5
2.0
A1203/(CaO + Na20 + K20)
Figure 8. Shand's alumina saturation classification of south Floirda pumice samples that have a rhyolitic composition. A single
sample from the Wynnhaven Beach site in northwestern Florida is shown as an open circle.
limited range in major element composition (Hildreth 1981).
The principle trace element compositions for the Florida
pumice samples are reported in Table 7. Barium (Ba) a trace
element that is selectively removed by the fractional crystalliza-
tion of potassium feldspar, is uniformly high (850-1100 ppm)
in all felsic pumice samples. The elevated concentration of Ba
indicates that potassium feldspar was not a residual phase in the
source material for the pumice and was not a crystallizing phase
during magma evolution. Strontium (Sr) exhibits a moderate,
negative correlation with silica concentration and a positive
correlation with the sodium content of felsic samples. High
silica, high-K pumice has Sr concentrations that range from
approximately 100-200 ppm. Low silica samples and medium-
K samples have Sr concentrations in the range 250-380 ppm.
Zirconium (Zr) concentration exhibits significant variation for
different samples (55-245 ppm), however, the variation of Zr
concentration does not appear to have any significant correla-
tion with the major element composition and may be a product
of inhomogeneous distribution of the trace mineral zircon
between different samples.
Rare Earth Elements
The 14 rare earth elements (REE) are often studied sepa-
rately from other trace elements due to their high, internal
chemical coherence and resistance to mobility or modification
associated with hydrothermal alteration or surface weathering.
The interrelationship between the different REE is normally
represented by a graphical plot of the rare earth elements
arranged in increasing atomic number on the x axis, versus
REE concentration logarithmically normalized to chondritic
meteorite reference values presented on the y axis (Rollinson
1993). The REE concentrations for Florida pumice samples are
reported in Table 8. The average chondritic values reported by
Boynton (1984) were used to calculate the normalized values
shown in Figures 9 and 10. There is a strong correlation
between the major element chemistry and observed REE
distribution pattern (Figures 9 and 10). High-K, high-silica
pumice (12 samples) has "lazy L" shaped REE patterns (Figure
9) that are associated with strong enrichment in the light REE,
with Lanthanum (La) concentrations approximately 70-90 times
chondritic abundances, and Lac/Ybc ratios ranging from
KISH
GEOCHEMICAL STUDIES OF PUMICE
Table 7. Principle Trace Element Composition of Florida Pumice Artifacts.
Sample Location Ba Ni Co Cs Ga Hf Nb Rb Sr Ta Th U V Zr Y
Southern Florida
MDC 1.567.6b Miami Circle 1110 <20 3.4 3.4 16.2 7.3 6.6 115.6 242.6 0.6 8.7 3.2 22.0 245.5 21.9
MDC 1.166.4 Miami Circle 942 5.5 2.1 4.8 12.4 2.2 5.0 110.9 176.7 <0.1 12.6 4.5 7.0 57.8 11.4
MDC 1.222.11a Miami Circle 234 4.2 15.7 2.1 21.4 3.3 2.7 40.2 315.5 0.4 3.2 1.2 135.0 102.8 22.5
MDC 1.472.4 Miami Circle 996 1.9 1.5 6.3 12.4 2.4 5.2 119.0 130.5 0.7 11.0 3.9 5.0 61.6 11.9
MDC 1.893.1 Miami Circle 887 1.4 1.0 6.0 13.4 2.4 6.3 120.2 134.8 0.9 13.3 4.2 <5 68.7 12.2
HMSF 896 Brickell Point 1055 1.7 1.5 5.5 12.1 2.4 5.6 111.5 235.3 0.9 11.3 4.2 <5 62.8 12.2
FBAR 75.62.10.1 Higgs Site 863 1.4 0.8 5.7 12.2 3.8 6.4 117.2 104.7 0.7 12.4 5.1 <5 139.8 12.2
FBAR 72.20.534.1 IndianKey 1022 0.5 0.5 5.8 13.9 1.9 6.1 119.5 165.1 0.8 10.4 4.7 <5 55.0 12.1
FLMNHA16262a Fort Center 996 0.5 <.5 5.2 12.9 2.8 5.7 120.1 81.7 0.6 12.4 4.3 <5 83.9 11.3
FLMNH 98576 Bear Lake 1 962 0.6 1.4 2.5 14.6 3.6 3.8 74.4 239.7 0.3 12.0 2.3 7.0 108.1 8.8
HMSF 582.1 BrickellPoint 788 0.8 1.6 8.8 17.0 5.9 8.3 147.3 159.5 1.0 16.5 5.5 11.0 200.5 23.6
HMSF 1018.18 Granada 827 0.9 1.4 2.3 16.0 4.7 3.8 59.1 379.9 0.3 7.1 2.1 12.0 165.3 8.2
HMSF 2942.1 Honey Hill 900 0.7 1.5 2.3 14.9 3.8 4.0 64.2 361.1 0.3 12.1 2.8 12.0 142.1 9.6
HMSF 427.5 Custom House 491 70.3 46.2 0.2 21.4 5.2 48.9 20.4 890.6 3.5 4.3 1.1 323.0 205.5 29.7
FBAR 74.2.10.8 Stock Island 966 0.3 0.6 5.3 12.6 2.2 5.6 119.2 202.1 0.7 12.1 4.8 <5 47.1 10.8
FBAR 00.14.47.1 Deering Estate 962 1.2 0.6 5.1 13.8 2.7 5.9 111.5 109.5 0.8 14.6 4.2 <5 63.3 15.6
FBAR 76.209.1.1 Shark Butchery 994 0.3 <.5 5.3 13.8 2.1 5.3 122.0 106.1 0.6 12.6 4.5 <5 50.8 12.0
FBAR 99.65.7.1 Whitebelt 1 2163 <.1 1.3 2.9 14.7 2.8 4.3 71.0 790.3 0.6 15.1 3.1 <5 74.0 10.8
HMSF 1874.1b Sutton Site 195 0.4 11.9 1.1 20.2 2.8 2.6 26.0 366.5 0.3 2.8 0.9 68.0 92.1 21.3
Northwestern Florida
EGLIN 056 Wynnhaven 777 75 29.3 1.1 22.9 3.4 10.7 142.1 179.5 0.8 12.1 6.5 136 106.1 22.4
EGLIN 198 Wynnhaven 1573 35 13.3 0.7 20.8 9.1 137.3 52.1 1385.6 7.9 12.1 2.8 102 404.2 44.8
EGLIN 742 Wynnhaven 967 <20 4.8 5.2 11.2 2.7 2.7 106.7 161.3 0.7 15.0 4.7 5.0 79.0 11.5
Trace element concentrations reported in parts per million (ppm)
100
E Low silica rhyolite
High silica rhyolite
O Medium silica rhyolite NW Florida
C:
0
.-C
O
a, 0
-1
E 10
I1I I I I I I I I I I
La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu
Rare Earth Element
Figure 9. Chondrite-normalized rare earth element abundances in Florida pumice samples that are classified as high-K, rhyolites. Thirteen samples are shown on
the diagram, including eleven high-silica rhyolites and two low-silica rhyolites. The low silica rhyolite samples include Miami Circle (MD 1.222.11a) and Brickell Point
(HMSF 582.1).
1000
100
10
Medium K rhyolite
Andesite
* Dacite NW Florida
* Basalt
O Basaltic trachyandesite
. S I I I
La Ce Pr Nd
Sm Eu Gd Tb Dy Ho
Er Tm Yb Lu
Rare Earth Elements
Figure 10. Chondrite-normalized rare earth element abundances in Florida pumice samples that are classified as medium -K rhyolites, dacites and andesites, plus
pumice and scoria with basaltic compositions.
4-L
Co
0
I I
" I
|
