Membership Information
 Table of Contents
 Editor's Page
 The Pigeon Creek Site San Salvador,...
 Lucayan Fishing Practices: An Experimental...
 Biological Archeology in the West...
 The Olsen COllection from Ile A...
 Investigation of Preceramic Sites...
 An Effigy Ceramic Bottle From Green...
 Columbus' Landfall an dthe Indian...
 The St. Joseph and Mayo Collections...
 The Loyalist Plantations on San...
 Highland House, Barbuda: An 18th...
 Resource Management Strategies...
 Archaeology on the Galways...
 Back Matter
 Back Cover

Group Title: Florida anthropologist
Title: The Florida anthropologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00027829/00024
 Material Information
Title: The Florida anthropologist
Abbreviated Title: Fla. anthropol.
Physical Description: v. : ill. ; 24 cm.
Language: English
Creator: Florida Anthropological Society
Conference on Historic Site Archaeology
Publisher: Florida Anthropological Society.
Place of Publication: Gainesville
Frequency: quarterly[]
two no. a year[ former 1948-]
Subject: Indians of North America -- Antiquities -- Periodicals -- Florida   ( lcsh )
Antiquities -- Periodicals -- Florida   ( lcsh )
Genre: periodical   ( marcgt )
Summary: Contains papers of the Annual Conference on Historic Site Archeology.
Dates or Sequential Designation: v. 1- May 1948-
 Record Information
Bibliographic ID: UF00027829
Volume ID: VID00024
Source Institution: University of Florida
Holding Location: Department of Special Collections and Area Studies, George A. Smathers Libraries, University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 01569447
lccn - 56028409
issn - 0015-3893

Table of Contents
    Membership Information
        Unnumbered ( 3 )
    Table of Contents
        Table of Contents
    Editor's Page
        Unnumbered ( 5 )
    The Pigeon Creek Site San Salvador, Bahamas
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
    Lucayan Fishing Practices: An Experimental Approach
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
    Biological Archeology in the West Indies
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
    The Olsen COllection from Ile A Vache, Haiti
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
    Investigation of Preceramic Sites on Ile A Vache, Haiti
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
    An Effigy Ceramic Bottle From Green Turtle Cay, Abaco
        Page 200
        Page 201
        Page 202
    Columbus' Landfall an dthe Indian Settlements of San Salvador
        Page 203
        Page 204
        Page 205
        Page 206
        Page 207
    The St. Joseph and Mayo Collections from Trinidad, West Indies
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
    The Loyalist Plantations on San Salvador Island, Bahamas
        Page 216
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
        Page 222
    Highland House, Barbuda: An 18th Century Retreat
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
        Page 233
        Page 234
        Page 235
        Page 236
        Page 237
        Page 238
        Page 239
        Page 240
        Page 241
        Page 242
    Resource Management Strategies on an Eighteentht Century Caribbean Sugar Plantation: Interpreting the Archaeological and Archival Records
        Page 243
        Page 244
        Page 245
        Page 246
        Page 247
        Page 248
        Page 249
        Page 250
    Archaeology on the Galways Plantation
        Page 251
        Page 252
        Page 253
        Page 254
        Page 255
        Page 256
        Page 257
        Page 258
    Back Matter
        Page 259
    Back Cover
        Page 260
Full Text


2000 Florida Anthropological Society Inc.

The Florida Anthropological Society Inc. holds
source text of the Florida Anthropologist
considered the copyright holder for the text
these publications.

all rights to the
and shall be
and images of

The Florida Anthropological Society has made this publication
available to the University of Florida, for purposes of
digitization and Internet distribution.

The Florida Anthropological Society reserves all rights to this
publication. All uses, excluding those made under "fair use"
provisions of U.S. Code, Title 17, Section 107 are restricted.

Contact the Florida Anthropological Society for additional
information and permissions.


S "V




THE FLORIDA ANTHROPOLOGIST is published quarterly in March, June, September, and
December by the Florida Anthropological Society, Inc., Geoarcheological Research
Center, Department of Geology, University of Miami, Coral Gables, Fl. 33124.
Subscription is by membership in the Society for individuals and institutions
interested in the aims of the Society. Annual dues are $10.00 ($12.00 after
March 1). Requests for membership and general inquiries should be addressed to
the Secretary; dues, changes of address to the Membership Secretary; manuscripts
for publication and orders for back issues to the Editor; and newsletter items
to the President. Address changes should be made at least 30 days prior to the
mailing of the next issue. Second class postage paid at Miami, Florida 33124.


President: Marion M. Almy
5321 Avenida del Mare
Sarasota, FL 33581

1st Vice President: John Beriault
3550 Bolero Way
Naples, FL 33942

2nd Vice President: Claudine Payne
1009 Pine Street
Tallahassee, FL 32303

Secretary: Ray Williams
Department of Anthropology
University of South Florida
Tampa, FL 33620

Membership Secretary: Joan Deming
1839 Pine Cone Circle
Clearwater, FL 33520

Treasurer and Resident Agent:
Ralph L. Struever
5350 Woodland Lakes Drive, Apt. 112
Palm Beach Gardens, FL 33410


Three years: Mary Lou Watson
P. 0. Box 9566
Panama City Beach, FL 32407

Two years: Jerry Hyde
4233 Oristano Road
Jacksonville, FL 32210

One year: Karen Malesky
5312 Bay State Road
Palmetto, FL 33561


Editor: Robert S. Carr
Geoarcheological Research Center
Department of Geology
University of Miami
Coral Gables, FL 33124

Editorial Board:
Morgan Crook
West Georgia College
Carrollton, GA 30118

Kathleen A. Deagan
Historic St. Augustine Preservation
St. Augustine, FL 32084
Yasar Iscan
Department of Anthropology
Florida Atlantic University
Boca Raton, FL 33431

Assistant Editor:
George M. Luer
Sarasota, FL

James J. Miller
Cultural Resource Management, Inc.
Tallahassee, FL 32303

George Percy
Division of Archives, History and
Records Management
Tallahassee, FL 32304

(USPS 200880)

COVER: Blackbeard's Tower, New Providence, Bahamas. Courtesy of the
Historical Museum of Southern Florida.







Editors' Page . . . . . .

The Pigeon Creek Site, San Salvador, Bahamas
by Richard Rose . . . . .
Lucayan Fishing Practices: An Experimental Approach
by William F. Keegan . . . .
Biological Archeology in the West Indies
by Storrs L. Olson . . . .
The Olsen Collection from Ile a Vache, Haiti
by Irving Rouse . . . . .
Investigation of Preceramic Sites on Ile a Vache, Haiti
by Clark Moore . . . . .
An Effigy Ceramic Bottle from Green Turtle Cay, Abaco
by Robert S. Carr and Sandra Riley . .
Columbus' Landfall and the Indian Settlements of
San Salvador
by Ruth G. Durlacher-Wolper . . .

The St. Joseph and Mayo Collections from Trinidad,
West Indies
by Stephen D. Glazier . . .

. 128

. 129

. 146

. 162

. 169

. 186

. 200

S. 203


Three Loyalist Plantations on San Salvador Island, Bahamas
by Kathy Gerace . . . . .

Highland House, Barbuda: An 18th Century Retreat
by David R. Watters and Desmond V. Nicholson . .

Resource Management Strategies on an Eighteenth Century
Sugar Plantation: Interpreting the Archaeological
and Archival Records
by Lydia Mihelic Pulsipher. . . .
Archaeology on the Galways Plantation
by Conrad M. Goodwin . . . . .









The Florida Anthropologist is pleased to present the papers of the
Third Bahamas Archaeological Conference. This hefty issue was made possible
by co-publication with the College Center of the Finger Lakes'Bahamian Field
Station on San Salvador. The issue includes all the papers presented except
for that of Shaun Sullivan which may be published in a future issue of the

I would like to clarify some confusion regarding the article by Wilma
Williams and the late Bert Mowers published in the previous issue of The
Florida Anthropologist (September 1982). Ms. Williams' current affiliation
with the Everglades Archeological Society, as indicated at the end of that
article, suggests that it was that society that conducted the excavation
at the Rolling Oaks II site. In fact, the entire project was conducted by
the Broward Archeological Society while Ms. Williams was still a member there.

Robert S. Carr

The Third Bahamas Conference on Archaeology was held from March 18-21,
1982, on the island of San Salvador, Commonwealth of the Bahamas. The Con-
ference included individuals who presented information on archeological
resources from throughout the Caribbean. The scope of the conference was
expanded to the entire Caribbean Basin because the increased economic develop-
ment there is contributing to the destruction of much of the cultural heritage
of the Caribbean. The role of the archaeologist is becoming increasing
important if these countries are to preserve and adequately record their
cultural heritage.

The Conference was hosted by the College Center of the Finger Lakes'
Bahamian Field Station at Graham's Harbour. The facilities provided the 26
participants with an opportunity to interface with one another,during and
after the working sessions. In addition to the sessions, the participants
were shown several of the prehistoric and historic sites that are located on
San Salvador, as well as a tour of the New World Museum. As with most small
conferences, much valuable information was exchanged between the participants.

I would like to thank Dr. Donald and Kathy Gerace for hosting the Con-
ference and for arranging all transportation to, from, and on San Salvador.
Thanks also goes to Robert Carr, editor of The Florida Anthropologist, for
again providing the Conference with the opportunity to publish its proceedings,
and to Ruth Durlacher-Wolper for making the New World Museum available to the
Conference. Last, but not least, I would like to thank the people of San
Salvador for their hospitality and friendship. If all goes well, the 4th
Bahamas Conference on Archaeology will be held on either Pine Cay in the
Turks and Caicos, or at New Providence in 1986.
John H. Winter
Assistant Professor
Conference Coordinator
Molloy College
Rockville Centre, New York


Richard Rose

The Bahama Islands, which are made up of more than 30 major islands and
over 600 small cays located off the southeastern coast of Florida, have re-
ceived relatively little attention from archeologists working in the Americas.
Although a number of surveys have been conducted in the Bahamian archipelago,
which includes the Bahamas, Turks and Caicos Islands, few sites have been
excavated. Consequently, the basic archeological determinations of chronology
and prehistoric settlement in the region have yet to be clearly established
and up-to-date studies of cultural adaptations to the Bahamian island ecosystem
have yet to be done.

During the past decade, one prehistoric site in particular has attracted
the attention of archeologists working in the Bahamas. The Pigeon Creek site
on San Salvador Island has been the focus of archeological interest beginning
in 1973 (Pratt 1974a) and culminating with the present writer's ongoing exca-
vation since 1978 (Rose 1980,1981). As a result of our excavations it is
possible to state with some degree of certainty that the Pigeon Creek site
represents one of the most important archeological resources in the Bahamas.
The Pigeon Creek settlement appears to have been the largest on San Salvador,
and, with its 12 or more acres, the site may indeed have been the largest
prehistoric settlement in the entire Bahamian archipelago. The size and com-
plexity of the site are thought to be significant, particularly in their
potential for providing information on native settlement systems as well as
local and regional adaptations to a limestone and coral environmental setting.

Moreover, the island of San Salvador has been of considerable ethno-
historic interest as the gateway for the Columbian voyages of discovery.
Columbus' eyewitness description of the indigenous Lucayan Arawak inhabitants
of San Salvador represents a valuable resource for archeological as well as
ethnohistoric studies in the northern Caribbean region.

San Salvador

The island of San Salvador is located along the eastern bank of the
Bahamian archipelago at 240 N Lat., 74030' W Long. It is about 370 miles
ESE of Miami, Florida, and 210 miles to the NE of Cuba. The closest islands
are Rum Cay, 22 miles to the SW, and Cat Island which is 45 miles due west
(Fig. 1).

San Salvador is approximately 12 miles long and 6 miles wide. Its topo-
graphy is typical of the central Bahamas with limestone and coral surface
features and generally low elevations. A series of large brackish lakes form
the interior of the island while expansive white sand beaches with phytokarst
and limestone outcrops constitute the coastline. Off-shore cays and coral
reefs surround the island. The vegetation is generally scrub-land character-
ized by coppice and coastal thicket secondary growth. Rainfall is approx-
imately 900 mm annually.


Bahama Little Abaco

t>z ^l Great

1Thenis ,Berry IslaZ
Bimin-is ..

300 km
* fI



.Cay Sal

\ Cat Island
0 San Salvador
Conception Is.o R
Exuma ay
Long Island

Crooked Is.c P
Fortune Is.0 ^ QPan
Ragged- ocklinsc=Mayaguana
Is.% Is. North Caicos
Providencialescpc Middle Caicos
o 'East Caicos
Sg Grand Turk
Grea t<


Figure 1. Major Islands of the Bahamian Archipelago
(after Granberry 1980)



At the time of Columbus' landing in 1492 the island was called Guanahani
by the Lucayan natives. Columbus renamed it San Salvador which the British
changed to Watlings Island. In 1926 the island was officially renamed San

Archeological research on San Salvador includes surveys and test exca-
vations by De Booy (1912,1913), Rainey (1940), Krieger (1937), Goggin (see
Hoffman 1967), Granberry (1955,1956), Hoffman (1967,1970), Pratt(1974a,1974b)
and Winter (1980,1981). In 1934 Rainey spent five days on San Salvador looking
for sites. In 1936-1937 Krieger found pottery, wood, shell and stone artifacts
in caves on the island. Goggin conducted excavations at four sites in 1960
and in 1965 Hoffman's excavation of the Palmetto Grove site on the northern
end of the island was the first in-depth excavation of an archeological site
in the Bahamas. In 1973 a'nd 1974 the Pratts conducted excavations at the
Pigeon Creek site, located in the southern part of the island and in 1980
Winter excavated the Ward site at the north end of San Salvador. Winter has
since completed a site survey of San Salvador (Winter 1980, 1981).

Until recently no more than five prehistoric sites were thought to exist
on San Salvador. The discovery of six new sites brings the total to eleven
with the possibility of additional discoveries in the future (Winter 1980,
1981). Sites on off-shore cays, represented by small surface accumulations
of pottery sherds, are also known.

The Pigeon Creek Site

The Pigeon Creek site is located along the leeward slope of a dune ridge
on the southeastern, or windward, side of San Salvador (Fig. 2). The site
is situated adjacent to a protected cove on the northeastern shore of Pigeon
Creek, which is in fact not a creek as its name implies but a large lagoon.
At present the site is covered by coppice and coastal thicket vegetation
which is almost impenetrable in places. Wherever vegetation has been removed,
however, a commanding view of the Pigeon Creek lagoon is revealed.

Archeological research at Pigeon Creek was initiated by the writer in
1978 following earlier investigations by the Pratts (Pratt 1974a,1974b) and
sporadic collecting by Wolper (1964). During 1978 and 1979 a limited test
excavation and site survey program was undertaken for the purpose of designing
a research project to recover data on the site's chronology and subsistence/
settlement systems with particular attention to determining the adaptive
strategies of the site's prehistoric inhabitants. The program was co-sponsored
by the CCFL Bahamian Field Station located on San Salvador, Hartwick College
in Oneonta, New York and the Rochester Museum and Science Center in Rochester,
New York. Archeological investigations at Pigeon Creek have continued from
1978 to 1982 with four field sessions averaging one month duration each.

A site survey employing surface collecting, shovel test pits and meter-
square excavation units has indicated that the site proper measures 200 x
240 meters, or 4.8 hectares (approximately 12 acres), making it the largest
of the eleven known prehistoric sites on San Salvador and possibly the largest
site in the Bahamian archipelago.

Test excavations at Pigeon Creek have revealed the presence of a multi-
component occupation characterized by shell and pottery refuse middens and
specialized work/activity areas with cultural deposits up to 50 cm in depth.



.0-1. POW


... *, --Pigeon Creek Site

San Salvador Island
the Bahamas
(Based on research
a by Sandra Riley)


o u6


Ceramics, predominantly of the Palmetto Series (cf. Hoffman 1967; Sears and
Sullivan 1978), as well as coral, shell and limestone tools have been re-
covered as has been an abundance of shell and bone food remains.

The preliminary results of the ongoing excavation program at Pigeon
Creek are summarized here under the broad topics of chronology, ceramics,
non-ceramic artifacts, settlement and subsistence.

The chronological position of Pigeon Creek falls generally within the
framework for Lucayan occupation of the Bahamas, Turks and Caicos Islands
as established by Sears and Sullivan (1978:22). They have suggested that
the Caicos Islands in the southern Bahamas were first settled between A.D.
800 and 1000 by Arawaks of the Meillacoid sub-tradition. Subsequently, the
central Bahamas were settled at least by A.D. 1000 based on the presence
of pottery of the Palmetto Series. Although Sears and Sullivan have sug-
gested that settlement proceeded from the Caicos to the islands of the central
Bahamas there is reason to believe that early settlement did not necessarily
follow a strictly linear route (i.e., Greater Antilles to Caicos to central
Bahamas). Migrations directly to the central Bahamas from the Greater
Antilles may well have occurred considering the proximity of Cuba and the
presence at Pigeon Creek of trade sherds which appear to be related to the
ceramics of the Ban{ culture of Cuba (John Winter, personal communication).

A series of five radiocarbon samples obtained during the 1981 field
season and processed by Dr. J. Stipp of the University of Miami Radiocarbon
Laboratory has indicated a wide range of dates from A.D. 596 to A.D. 1765
(Table 1). The A.D. 596 date seems too early and should be regarded with
caution even though the dated sample was collected from the lowest levels
(40-60 cm) of the occupation zone. The two dates at the upper range (A.D.
1765 & A.D. 1675) are also difficult to interpret and may be the result of
agricultural burning activities during the post-contact Loyalist Period.

The remaining dates (A.D. 1000 & A.D. 1360) fit well within the estab-
lished time frame for the central Bahamas suggesting that Pigeon Creek was
either contemporaneous with or a bit later than Palmetto Grove (cf. Hoffman
1967), and may have been occupied when Columbus landed in 1492. Radiocarbon
samples obtained during the 1982 field season may help to narrow the excessive
range of dates thus far recorded for Pigeon Creek.

Table 1. Radiocarbon Dates--Pigeon Creek Site
(by Geoarcheological Research Center, Depart-
ment of Geology, University of Miami)
Lab # Material Wt. Level C14 Age Calendar Age
UM-2271 Charcoal 10.0 g 30-40 cm 305 75 A.D. 1675
UM-2272 Charcoal 31.0 g 10-20 cm 215 60 A.D. 1765
UM-2273 Charcoal 7.0 g 30-40 cm 580 90 A.D. 1400
UM-2274 Charcoal 15.5 g 26 cm 620 t 70 A.D. 1360

Bone (fish) 31.0 g 40-50 cm


1384 65 A.D. 596



A total of 3226 pottery sherds has been catalogued for Pigeon Creek
since 1978. Close to 2% of the sherds have been identified as trade ware
represented by thin-walled, quartz-tempered pottery. The remaining 98% of
sherds collected at the site are characterized by the thick-walled, shell-
tempered pottery of the Palmetto series as identified by Hoffman for Palmetto
Grove and described by Sears and Sullivan for the central Bahamas (Hoffman
1967,1970; Sears and Sullivan 1978). Table 2 shows the relative frequencies
of the ceramic samples collected at Pigeon Creek. Approximately 84% of the
collection is represented by Palmetto Plain (Fig. 3a-c), 14% by Palmetto
Mat Marked (Fig. 3d,e), 0.6% by Palmetto Applique (Fig. 4a) and 0.2% by
Palmetto Incised (Fig. 4b,c).

Examples of ceramics suggestive of Ostionoid and Meillacoid Series
pottery have been recovered at Pigeon Creek. Sherds exhibiting the rim peaks
characteristic of the Ostionoid sub-tradition (Fig. 4e) and examples of
Meillacoid-type designs represented by raised rim bands with cross-hatched
incisions, sigmoid appliques and effigy decorations are known (Fig. 4a,f,g).

As to the occurrence of pottery vessel forms, only a preliminary in-
ventory is available at the present time. The majority of sherds are from
hemispherical bowls. These vessels were probably used for water collection
and storage as indicated by the unslipped, roughened surface texture which
would have facilitated handling when wet (Fig. 3a,4d). Such bowls were
probably also used for food storage and preparation.

A number of flat or slightly concave pottery griddles are represented
by mat-marked sherds which comprise 14% of the ceramic inventory (Fig. 3d,e).
These griddles appear to have been rounded in shape and approximately 30 cm
(12 inches) in diameter, though larger griddles may also be represented.

Table 2. Pottery Frequencies--Pigeon Creek Site

Total Plain Mat-marked Incised Appliqued Trade
Year No. # % # % # % # % # %

1978 681 509 74.74 143 21.00 1 0.15 13 1.91 15 2.20
1979 223 196 87.89 24 10.76 2 0.90 1 0.45 0 0.00
1981 1042 862 82.73 148 14.20 1 0.10 3 0.29 28 2.69
1982 1280 1124 87.81 136 10.63 2 0.16 1 0.08 17 1,33

3226 2691 83.42 451 13.98 6 0.19 18 0.56 60 1.86

Non-Ceramic Artifacts
Non-ceramic artifacts excavated at Pigeon Creek include stone, coral
and shell objects. Stone specimens are represented primarily by indigenous
limestone objects and secondarily by exotic igneous stone implements. Coral
objects include tools fashioned from Acropora palmata and A. cervicornis
specimens. Shell artifacts from Pigeon Creek are relatively rare with
conch (Strombus gigas) representing the predominant material used for tools.
Ornaments of shell and quartz are also known.



b. Palmetto Plain

a. Palmetto Plain (x2)

c. Palmetto Plain

d. Palmetto Mat Marked

Figure 3
Pigeon Creek Ceramics

e. Palmetto Mat Marked (x2)



c. Palmetto Incised

a. Palmetto Applique b. Palmetto


d. Palmetto Applique

e. Palmetto Plain w/ Rim Points

g. Meillacoid-type
Sherd (xl)

f. Meillacoid-type
Sherd (xl)

Figure 4
Pigeon Creek Ceramics



Typical of the limestone objects are a number of flat-topped slabs and
a fragment of a large limestone bowl or trough. Similar objects of limestone,
both slabs and troughs, have been reported from the Palmetto Grove site
(Hoffman 1967). Two limestone slabs were found at Pigeon Creek in 1981 in
association with a refuse midden of clam shells (Fig. 5). The intended func-
tion of these slabs is not known though it is possible they may have served
as surfaces upon which clam shells were opened. Close to half of the clam
shells collected in the illustrated feature show evidence of having been
smashed or otherwise broken along the umbo suggesting that clams were opened
in this manner. It is also possible that the flat-topped limestone slabs
were used for processing manioc flour though their association with clam
shells suggests otherwise in this particular instance.

A number of small tools manufactured from calcified limestone have been
recovered at Pigeon Creek. Small, knife-like tools, when held between the
thumb and forefinger, would have made effective implements for peeling manioc
roots and cutting leaves for basket or mat plaiting (Fig. 6a). Two limestone
burin-like tools excavated during the 1978 season may also have been used
for manioc processing (Fig. 6b). Close examination of these objects suggests
that they may have been set into a handle with only their worked tips exposed.
The resulting implement would have been effective for grating manioc roots
in the preparation of cassava flour.

Other stone tools found at Pigeon Creek include two pestles carved from
aragonite, an exceptionally well-crafted, blunt-ended grinding tool of polish-
ed greenstone and a fragment of a jade celt ( Fig. 6d,e). Aragonite is known
to occur on Abaco Island in the northern Bahamas. The closest source of green-
stone to San Salvador is the Dominican Republic in the Greater Antilles and
jade is known only from northern Venezuela and Guatemala. X-ray diffraction
tests conducted by Kent Johnson at SUNY Binghamton have indicated the Motagua
Valley of Guatemala as a probable source for the jade fragment (Johnson n.d.).

Coral tools are commonly found at Pigeon Creek. Pieces of staghorn coral
(Acropora cervicornis) (Fig. 6c) lack the normally sharp polyps found on fresh
coral probably as the result of wear. Coral tools would have functioned well
as abrasive rasps used for woodworking or as grinding implements. Comparison
of the archeological specimens with fresh coral clearly shows the difference
in surface texture.

The occurrence of shell tools at Pigeon Creek is relatively infrequent.
One exceptionally well-made conch scraper (lig. 6i) was found during the
1982 excavation. To date we have been unable to identify tools made from
clam shells even though these would have made useful scrapers ana could have
been used for other purposes as well. It is recognized, however, that such
tools could have been fashioned on the spot for specific tasks and then dis-
carded leaving little evidence of use on the shell edge.

Shell and stone ornaments also occur at Pigeon Creek. Shell beads
(Fig. 6h), bead blanks and pendents (Fig. 6g) have been collected. Beads
fashioned from quartz (Fig. 6f) and aragonite are also known.

There has been little research on Lucayan settlement systems in the
Bahamas. Details of Lucayan villages, house types and inter-community re-
lationships are practically nonexistent with the possible exception of



Figure 5. Clam Shell Feature. Pigeon Creek


14?'"OC' i


Sullivan's (1980) discovery of a Lucayan residential zone on Middle Caicos

Ethnohistoric descriptions have given us a brief view of Lucayan settle-
ments. Columbus reported seeing two or three villages on San Salvador and
on Long Island he reported houses which resembled Moorish tents. Of the many
villages seen by Columbus none contained more than from twelve to fifteen
houses (Columbus 1960).

We have no idea at present as to the number of houses at Pigeon Creek.
Traces of a post mold were identified by the Pratts in 1974, but little is
known concerning the location and arrangement of houses in the settlement.
A large flat area bordered by two occupation ridges suggests the presence
of a plaza or batey near the center of the site. This area will be investi-
gated in more detail in the future.

The Pigeon Creek site appears to have been located in an ecologically
advantageous and politically strategic area. Built along a ridge overlooking
a protected cove on the Pigeon Creek lagoon, the village's location would
have provided residents with ready access to the marine resources of the
lagoon as well as the nearby coastal resource zone. The settlement's location
inland from the coast surely gave a certain amount of protection from ocean
storms. Its location atop a dune ridge would have given the residents a
sweeping view of the lagoon, thereby offering "early warning" protection
against enemy attacks by canoe. Finally, the village's proximity to the
extensive inland lake system of San Salvador would have provided an effective
communications and exchange link with other settlements on the island. It
is possible, moreover, that this large inland lake system may have been a
major stimulus for settlement on San Salvador. Any discussion of the settle-
ment patterns of San Salvador must take into consideration that Pigeon Creek,
which may have been the largest prehistoric settlement in the Bahamas, is
located on one of the largest protected lagoons in the archipelago. The prime
location of the settlement coupled with the close availability of abundant
marine resources and the presence of relatively rich soil for agricultural
purposes strongly suggest that these factors were instrumental in the settle-
ment and growth of the prehistoric village of Pigeon Creek.

A varied subsistence economy based on marine resources and root crop
agriculture was practiced at Pigeon Creek. Shellfish constitutes the largest
percentage of actual food remains recovered in the excavations. Clam shells
(Codakia obicularis) and the shells of land snails (Cerion sp.) are found
in abundance, though it is questionable whether the latter were utilized as
a food resource (J. Rose, personal communication). Conch (Strombus gigas)
and sea snails (Cittarium pica) have also been recovered archeologically.
There are also a fair number of chiton, nerita and other small mollusca
represented. Fish bones, particularly grouper (Epinephelus striatus) and
parrot fish (Sparisoma sp.), are also present. It is clear from the arche-
ological evidence that the prehistoric inhabitants of Pigeon Creek subsisted
largely on seafood with clams representing a major portion of their diet.

The people of Pigeon Creek also practiced an agricultural economy as
suggested by the large number of griddle sherds recovered from the refuse
middens. Flat round griddles are known to have been used by the Arawaks
of the Greater Antilles and northern South America for baking manioc flour



a. Limestone Tool

c. Coral Rasp (xl)

b. Limestone
Tool (xl)



ii -~

d. Greenstone
Pestle (xl.25)

f. Quartz Bead

e. Aragonite
Pestle (x1.25)

. a,

g. Shell Pendent (x0.5)

i. Conch Scrapper (xl)

h. Shell Bead

Figure 6
Non-Ceramic Artifacts
Pigeon Creek Site


into cassava cakes, a staple food in the Caribbean region. In addition to
griddle sherds, non-ceramic artifacts which possibly had agricultural uses
have been identified at Pigeon Creek. The small knife-like tools of calcified
limestone and pieces of staghorn coral would have functioned well in the pro-
cessing of manioc roots, perhaps as peeling and grating devices.

Two physical features of the Pigeon Creek site are significant in terms
of the subsistence system. First, the site is located along a shallow cove
of a large tidal lagoon which serves today as a natural breeding area for
deep water ocean fish. The lagoon ecosystem would have provided the ancient
residents of Pigeon Creek with an abundant supply of fish and other seafood,
especially clams which could have been dug from the shallow shoreline adja-
cent to the site.

A second feature which must be considered in reconstructing the aborig-
inal subsistence system is the agricultural potential of the Pigeon Creek
region. In contrast to the rest of San Salvador with its generally poor
soils, the Pigeon Creek area is known for its good quality soil. Thus, the
site's location on a richly stocked lagoon in an area of good agricultural
potential suggests that the people of Pigeon Creek were able to take advan-
tage of ample subsistence resources close at hand.

In addition to marine resources and manioc agriculture, it is likely
that other aquatic and terrestrial foods were utilized. Wild plant gathering
was surely practiced and the hunting of small animals such as the iguana,
hutia and dog as well as the sea turtle and various kinds of birds probably
provided important supplements to the diet.

Columbus reported seeing "no beast of any kind (on San Salvador), except
parrots" in his journal entry for October 12th, the day of his arrival.
Subsequently, on Rum Cay he saw parrots and lizards (presumably iguanas) and
on Long Island he mentioned seeing small dogs that didn't bark (Columbus 1963).
It is certain that iguanas were also present on San Salvador. The Lucayan
name for the island was Guanahini, meaning Place of the Iguana and iguanas
have been reported in historic times. As to dogs, there has been no arche-
clogical evidence of their presence on San Salvador but they are known to
have been present on nearby islands.

The extent to which land animals, whether wild or domesticated, may have
been utilized as food by the Lucayans must await further research. Remains
of hutia, a small rodent-like animal which was once widely distributed in
the Bahamas, have been found in archeological contexts, and it is known that
dogs were eaten in Cuba as well as in Mexico during Pre-Columbian times
(Wing 1978). Presumably, iguanas were also eaten. In an area with few land
animals these additional food resources would have provided an important
protein supplement to the plant and marine diet of the Lucayans. We should
be cautious, however, not to place too much emphasis on the possible exploita-
tion of land animal resources by the Lucayans. Research by James Nason (1975)
on Etal Island in the Pacific Islands indicates that land animals such as
pigs and chickens are of minor consequence in the local diet and the pre-
dominant protein intake of the islanders is from fish and other marine animals.
On the other hand, Neitschmann's study of modern Moskito fisherman along the
Caribbean coast of Nicaragua shows that a fisherman is unable to produce
sufficient calories from fish alone to support the consumption requirements
of himself and one other adult (Stark and Voorhies 1978:281).




Current archeological research in the Bahamian archipelago reflects the
attempts by archeologists to solve the basic and long-standing problems of
the region's prehistory. A quarter of a century ago Granberry attempted to
delineate the origins of prehistoric culture in the Bahamas (Granberry 1955,
1956). Today the question of origins is still with us. Although it is gen-
erally agreed that the native populations of the Bahamas were predominantly
Arawakan-speakers with a possible later Carib migration into the region,
there is a lack of consensus among archeologists as to the place or places
of origin of the Bahamian Arawaks, or Lucayans as they are called today.
Granberry (1956) has argued for a northern Haitian origin whereas Sears and
Sullivan (1978) and Winter (personal communication) have suggested that Cuba
may have been the homeland. There is reason to believe, moreover, that both
Haiti and Cuba may have contributed to early settlement in the Bahamas as the
ceramic sequence at Pigeon Creek suggests.

As to the problems of chronology, Sears and Sullivan have argued that
the Bahamas were first settled between A.D. 800 and A.D. 1000. Hoffman, based
on his work at Palmetto Grove, has indicated A.D. 850 for the beginnings of
settlement at that site (Hoffman 1967). To date, earlier occupations have
not been identified, although radiocarbon dates from Pigeon Creek suggest the
possibility of a pre-A.D. 800 settlement on San Salvador. For the time being,
at least, we should not disregard the possibility of earlier sites in the
Bahamas even though their presence appears unlikely.

Even though the beginnings of settlement in the Bahamas is presently
uncertain, there can be little doubt as to the terminal date of major Lucayan
occupation in the region. Columbus reported seeing numerous village settle-
ments in 1492 and population estimates for the Bahama Islands range from
20,000 to 40,000 at the time of European contact. Yet, by 1513, a mere
twenty-one years later, the Bahama Islands were found to have been depop-
ulated when Juan Ponce de Leon raided them on a slave hunting expedition
(Sauer 1969).

Of particular interest in Bahamian archeology has been the study of
prehistoric adaptations to the coral and limestone ecology of the islands.
Early settlers, presumably from the Greater Antilles, would have had to adapt
their lifestyle to an ecosystem with severe environmental limitations. The
coral islands of the Bahamas with their generally poor soils and scrub-like
vegetation differ significantly from the volcanic islands of Cuba, Hispaniola
and Puerto Rico with their abundant supply of hard stone for toolmaking and
good soils for agriculture. Consequently, immigrants to the Bahamas would
have had to adjust to substantially different food resources, an absence of
igneous stone for proper tools, and the probable scarcity of fresh water
for drinking. At Pigeon Creek there is evidence that these restraints were
met with positive results. Tools manufactured from limestone, coral and
shell, particularly conch, would have offered a workable alternative for
tasks of food preparation, woodworking and the like.

Another adaptive mechanism which is evident both at Pigeon Creek and
the Bahamas in general is the use of crushed shell for tempering pottery.
In the absence of locally available quartz grit for pottery making, ground
bits of shell were used by the Lucayan potters. Although structurally weak
in comparison with the thin-walled, quartz-tempered pottery of the Greater


Antilles, the shell-tempered Palmetto ware made an acceptable substitute
for water and storage jars as well as the flat griddles used to bake cassava

Evidence of trade at Pigeon Creek is supported by the presence of igneous
stone implements and quartz-tempered pottery derived from other locations in
the Caribbean region. Whether there existed an organized trading network or
these items were brought to the islands by the original inhabitants is not
clear. It is known, however, that the Lucayans were seafaring people.
Columbus mentioned seeing canoes of many different sizes, from small, one-
person models to crafts large enough to hold forty persons. During his
journey through the Bahamas, Columbus encountered canoes traveling from
island to island which suggests that a lively inter-island trading network

The study of cultural adaptations to the coral and limestone environment
of the Bahamas is still at its beginning stages. It is important to realize,
however, that the Pigeon Creek site offers the potential for in-depth studies
of environmental restraints on an immigrant population. Such studies may,
in turn, enable us to understand better the processes of settlement in the
Bahama Islands. For example, it is suggested that the Pigeon Creek settle-
ment may have been in response to the ecological limitations faced by the
early Lucayan immigrants. The site's prime location, large size and cultural
complexity must be regarded as significant factors in the growth and develop-
ment of the Pigeon Creek settlement.

References Cited

Columbus, C.
1960 Journal. Edited by Cecil Jane; rev. by L. A. Vigneras.
New York: C. N. Potter Publ.

1963 Journals and Other Documents on the Life and Voyages of
Christopher Columbus. Trans. and edited by Samuel Eliot
Morison. New York: The Heritage Press.

De Booy, T.
1912 Lucayan Remains on the Caicos Islands. American Anthropol-
ogist 14:81-105.

1913 Lucayan Artifacts from the Bahamas. American Anthropologist

Granberry, J.
1955 A Survey of Bahamian Archaeology. MS, Masters Thesis,
Dept. of Anthropology, University of Florida, Gainesville.

1956 The Cultural Position of the Bahamas in Caribbean Archaeology.
American Antiquity 22:128-134.

1980 A Brief History of Bahamian Archaeology. The Florida
Anthropologist 33:83-93.



Hoffman, C. A., Jr.
1967 Bahamian Prehistory: Cultural Adaptations to an Island
Environment. Ph. D. dissertation, University of Arizona.

1970 The Palmetto Grove Site on San Salvador, Bahamas.
Contrib. of the Florida State Museum, Social Sciences,
No. 16.

Johnson, K. R.
n.d. MS. Results of Tests to Determine Mineralogy of an
Artifact from San Salvador, Bahamas.

Krieger, H. W.

Nason, J. D.

Pratt, M. K.

The Bahama Islands and Their Prehistoric Population.
Explorations and Field Work of the Smithsonian Institution
in 1936, pp. 93-98.

The Effects of Social Change on Marine Technology on a
Pacific Atoll Community. In Casteel & Quimby, eds., Marine
Adaptations of the Pacific, Moulton Publ., pp, 5-38.

Preliminary Report 1973, Prehistoric Archaeology of San
Salvador, Bahamas. Island Environmental Studies, Corning,
New York: CCFL.

1974b Preliminary Report 1974, Prehistoric Archaeology of San
Salvador, Bahamas. Island Environmental Studies, Corning,
New York: CCFL.

Rainey, F. G.

Sauer, C. O.

Porto Rican Archaeology. Scientific survey of Porto Rico
and the Virgin Islands, New York Academy of Sciences
18 (pt. 1).

The Early Spanish Main. Berkeley: University of California

Sears, W. H. and S. D. Sullivan
1978 Bahamas Prehistory. American Antiquity 43:3-25.

Stark, B. L.

and B. Voorhies
Future Research Directions. In Stark & Voorhies, eds.,
Prehistoric Coastal Adaptations: The Economy and Ecology
of Maritime Middle America, New York: Academic Press.

Sullivan, S. D.
1980 An Overview of the 1976 to 1978 Archaeological Investi-
gations in the Caicos Islands. The Florida Anthropologist


Winter, J.


Wolper, R. G.


A Preliminary Archaeological Survey of San Salvador,
Bahamas. Bahamas Archaeology Project, Reports and
Papers. CCFL Bahamian Field Station.

1981 Archaeological Site Reconnaissance, San Salvador,
Cat Island, and Rum Cay. Bahamas Archaeology Project,
Reports and Papers, CCFL Bahamian Field Station.

A New Theory Identifying the Locale of Columbus' Light,
Landfall and Landing. Smithsonian Miscellaneous
Collections 148(1):1-41.

Richard Rose
Curator of Anthropology
The Rochester Museum
and Science Center
Rochester, New York 14603



William F. Keegan

This paper is a preliminary report of ongoing research directed toward
explaining the subsistence behavior of the prehistoric Lucayan/Arawaks who
occupied the Bahamas archipelago from about 800-1500 A.D. (Sears and Sullivan
1978). The research reported is one component of a research program struc-
tured by the concept of culture as an adaptive system (Kirch 1980), with more
specific models drawn from evolutionary ecology (Pianka 1974; Winterhalder
and Smith 1981) and formalist economics (Earle and Christenson 1980). This
report is the result of what Jochim (1979) has called "breaking down the
system," and it is organized to reflect this. The successive levels of re-
fining the problem from general theory are presented first. The second half
of the paper discusses Lucayan subsistence and the Pine Cay fishing project.

Research Program

Briefly, my approach can be illustrated by two figures. The first simply
indicates that culture is viewed as the means of adapting a population to its
environment, with environment subsuming both physical and social milieu (Kirch

The second figure illustrates this interaction from the perspective of
food procurement only, in which a group's behavior and technology are directed
toward obtaining (satisfying) the group's food requirements through inter-
action with the physical environment. The second figure obviously severs
connections with other cultural components that may have an impact on food-
getting behavior (e.g., exchange, ritual). By so doing the predicted output
may not completely correspond with the actual output. However, cultural
systems are too complex to be studied holistically except at a very general
level of resolution. The system must be broken down into manageable subsystems
or components if specific studies are to be conducted (Hill 1977). When the
results of the studies of specific components become available they then can
be integrated to restructure the system as a whole. At that time additional
inputs through specific variables can be recognized and these variables can
be modified accordingly.

For example, our study of subsistence may predict that x number of
molluscs should have been collected to satisfy a group's nutritional needs.
If these same molluscs also were used as a medium for exchange, then they
would have been collected at a greater than expected frequency. Once the
system has been reintegrated, with social transactions added to subsistence
practices, the variable "mollusc collecting" can be modified to reflect the
impact of exchange (the effects of this modification on other variables can
also be determined).

Optimal foraging and cost-minimization models provide methods for pre-
dicting the optimal solution to resource use (Winterhalder and Smith 1981;
Earle and Christenson 1980). However, before these models can be used the





After Kirch 1980:105

Figure 1.

Representation of culture as a system that links human
populations with their environmental matrix,





Figure 2. Operational model of human subsistence behavior.



investigator minimally must know something about group size and density,
technology and behavior, and the environment that was exploited. In many
cases the archaeological record has not preserved the evidence needed to
employ these models (Keene 1981). The author is optimistic that the archae-
ological data base in the Bahamas is adequate to permit the use of these models
once prerequisite data have been collected.

Modeling the Environment

The first task is the definition of resource availability. In most cases
anthropologists can rely on the ecological studies of biologists, or they can
collect their own data on the effective environment of the population under
study (Hawkes et al. 1982). Archaeologists often can employ those studies,
but their reconstruction of the prehistoric populations effective environ-
ment is complicated by the problems of preservation and they must demonstrate
that present environmental conditions adequately reflect those of the past
(Butzer 1964).

Catchment analysis is one method used to indicate the resource points on
which a prehistoric population probably focused (Sullivan 1982). This method
is based on the assumption that settlements are located to maximize the avail-
ability of specific resources (Krebs 1978) or to minimize the costs of trans-
porting those resources. In keeping with the present discussion we may sug-
gest that catchment analysis directs attention to the prehistoric group's
effective environment. The habitats included in a given radius around a site
should be those that were most frequently exploited. Attention can then be
focused on defining the resources available in these habitats, and a possi-
bilistic reconstruction of environmental exploitation can be developed.

Lucayan Subsistence

The term subsistence is used to indicate the satisfaction of the human
body's physiological requirements through the procurement of food sources
(Wing and Brown 1979); it is represented by the term output in Fig. 2.

Certain of those food items are represented at archaeological sites by
identifiable shells and bones, and plant foods may be represented (by pollen,
phytoliths, seeds, etc.) although collection methods have not been employed
for their recovery. There is also ethnohistoric documentation of agricultural
practices (manioc and maize cultivation; see Sears and Sullivan 1978; Gran-
berry 1980), and we should assume that a variety of wild plant foods were
gathered when they were available.

From that evidence, Lucayan subsistence can be viewed as the sum of the
contributing procurement strategies:
Diet = GP + DP + MC + H + F

where GP is gathered plants, DP is domesticated plants, MC is mollusc col-
lecting, H is hunting of terrestrial fauna (e.g., hutia and lizards),and F
is fishing/fish. The relative contribution of each strategy is not known,
although several independent lines of investigation are being followed.


With the possible exception of sites associated with Armstrong Pond on
Middle Caicos (Sullivan 1980), all of the known Lucayan sites share a common
attribute when any size catchment area is drawn: a significant portion of
the catchment is comprised of marine habitats. While factors other than food
resource availability probably contributed to the choice of these site loca-
tions (e.g., inter-island communication, freshwater availability beneath
coastal sand, etc.) the discovery of mollusc shells and fish bones at these
sites indicates that marine habitats were used.

Site catchments can suggest where those marine resources were procured.
However, we cannot understand their exploitation until their availability in
the environment has been explicated. Unfortunately, the limited study of
marine habitats and the behavior of their occupants makes it difficult to
reconstruct the availability of most marine organisms. Most references are
limited to typological descriptions which provide very general habitat criteria,
and specific studies are aimed at resources of current economic importance
and/or those organisms whose behavior relates to ecological problems. In
most cases those studies are of little relevance to the archaeological study
of human subsistence behavior. The Pine Cay fishing project was designed to
provide information about the availability of fish species in the vicinity
of the Lucayan site on Pine Cay, Turks and Caicos Islands.

Lucayan Fishing Practices

Before reviewing the Pine Cay fishing project two other sources of evi-
dence concerning fishing practices will be reviewed. Those are ethnohistoric
reports and fish bone analysis.

Rouse (1963:524), citing ethnohistoric reports from the Spanish chron-
iclers, indicates that cotton or fiber nets, poison, many pronged spears,
weirs, and basketry traps were being used by the Arawaks of Hispaniola at
the time of Spanish contact. Since the Arawaks who colonized the Bahamas
probably came from Hispaniola (Sears and Sullivan 1978; Sullivan 1981), it
is possible that any or all of those techniques were used by the Lucayans
(see Cherry 1981:61-62, for a discussion of how the operation of the "founder
principle" may result in the loss of techniques represented in the total
cultural inventory). Such possibilism does not help to explain Lucayan fish-
ing practices, but it can be used to support the reconstruction of subsistence
strategies in the absence of artifactual evidence. Of course, such recon-
structions require additional supporting evidence. Rouse (1963:544) also
reports that the bow and arrow reportedly was used by the Lucayans, and two
wooden fishhooks have been recovered from a Lucayan cave site on Crooked
Island (Rainey 1934; Granberry 1978:36).

Fish bones from a number of Lucayan sites have been identified at the
zooarchaeology laboratory at the Florida State Museum (Wing 1969, personal
communication). The results of the identification of specimens from the Pine
Cay Site are summarized in Table 1.

The faunal sample from the Pine Cay Site was not collected with the
fishing project in mind; the sample was the result of an attempt to define
midden concentrations on portions of the site. Therefore, the representa-
tiveness of the sample cannot be statistically defined (Keegan 1981). In
addition, the earth was sifted through 5 mm2 mesh screen, which is probably
not adequate for the collection of fish bones (Wing and Brown 1979:7).



Table 1 should therefore be interpreted with caution since sample size may
have a significant effect on derived measures (Grayson 1981). At this point
in time it should not be necessary to state that meaningful samples cannot
be obtained without considering the data to be collected and the sampling
theory and sample significance required by a study (Thomas 1978; Mueller 1979).
This problem will be rectified in future excavations at the Pine Cay Site.

Pine Cay Fishing Project

Fishing can be viewed as a predator-prey interaction. Organisms are
non-randomly distributed in the environment, and ecological studies have
demonstrated that predator behavior falls somewhere between random and
directional (Krebs 1978). Those studies have not considered the effects of
learning and information sharing on predator behavior so human predators
may act in an even more conscious, directional way. To capture a fish the
humans) must behave in a manner that articulates with the fish's behavior.
From that simple premise it should be apparent that the fish captured pro-
vides evidence of the human behavior that led to capture. By focusing on
the fish's availability (e.g., habitat, density, etc.) and behavior (e.g.,
feeding strategy, predator avoidance, etc.) a variety of information about
human behavior becomes available. In some cases the method of capture is
very prey specific (e.g., sturgeon hooks; cf. Rohan-Csermak 1963), but in any
case some indication of the method of capture and the habitat ("patch") ex-
ploited will be revealed.

With that in mind an experimental fishing project was conducted on Pine
Cay from November 25, 1980 through July 20, 1981. The project was monitored
by a technician with the Foundation for the Protection of Reefs and Islands
from Degradation and Exploitation (P.R.I.D.E.), a not-for-profit scientific
and educational research foundation based on Pine Cay. Haitian-style basketry
fish traps were set in the two principle marine environments: the tidal flats
and the reef (Fig. 3). Conch (Strombus gigas) and/or crawfish (Panulirus
argus) were used for bait. The traps were emptied at one week intervals,
weather permitting, and captured fish were recorded by Genus species, length,
and weight (Tables 2 and 3).

The purpose of the project was to determine what types of fish were
available in each habitat, the approximate frequency of their availability,
whether their availability varied during the year, the problems associated
with fishing in these environments (e.g., weather, water turbidity, wave
action, the activities of other predators, trap deterioration, etc.), and
to collect a controlled sample of fish species captured in basketry traps.


Basketry traps were selected for the experiment for several reasons.
First, they are one of the known technologies employed in the prehistoric
Caribbean. Second, these facilities require the least amount of human mani-
pulation during use. Hooks, spears, and the bow and arrow require a level
of skill and enculturation (e.g., prey selection) that cannot adequately be
replicated during a short-term study. Other facilities, such as weirs and
nets, require a large initial investment and group cooperation during use.



Barrier Reef -
1-5fathams "

Patch Reef

Patch Reef


\John Brown's
Hole (5fathoms)

Tidal Flats

F; mrn Pivue (Thy and rnurroufldifla waters,

scale o loom
o Trap Locations1
o Trap Locations

Barrier Reef




6%" 0i''U


A third factor was the representativeness of the sample to be collected.
Any technique will yield a non-random sample of the fish population in the
environment. Basketry traps seemed to provide the best available sampling
technique. In their study of vertebrate remains from Antigua, West Indies,
Wing et al. (1968:134) report: "With few exceptions the kinds of fishes
represented are those most frequently caught in traps today, suggesting that
they were caught in the same way in aboriginal times." Wing (personal com-
munication) has since suggested that the evidence does not adequately support
trapping as the principle prehistoric fishing technique. It is perhaps more
likely that traps capture the most common fish species in a habitat, and that
prehistoric techniques resulted in a similar sample.

The basketry traps were set at specific locations within the two main
environmental zones to which fish species are assigned: the tidal flats and
the reef (Fig. 3). These general environments subsume a variety of habitats
and patches which may be defined by the availability of resources within them.
These patches can then be located by conducting marine surveys. This specifi-
cation of marine patches will be undertaken in subsequent research. The pre-
sent study was primarily concerned with distinguishing between the two general

The need for that distinction was suggested by an apparent discontinuity
indicated by a site located adjacent to one environment (the tidal flats) and
fish remains identified as coming from the other (the reef). The fringing
barrier reef is over one mile from the site, and, even though patch reefs do
occur closer to the site, the shifting coastline is redepositing the site
on the tidal flats. Unless fishing was of little significance in the choice
of site location there should be some other explanation for this apparent

The results of the trapping experiment indicate that the fish identified
from the Pine Cay Site were probably available in the adjacent tidal flat
environment. While sampling biases do not at present permit the calculation
of this probability, the evidence strongly supports this belief (cf. Tables
1 and 2). Part of the difficulty in calculating the probability derives
from the identification of fish remains at the site to the taxonomic level
of Genus, while fish captured during the experiment can be identified to the
species level. The association of many of these fish with reef environments
(e.g., Sparisoma, Scarus) and their occurrence in tidal flat environments
could result from congeneric behavioral variations and/or the very general
identification of their primary habitat. In other words, their habitat may
have been defined by where they were seen rather than why they were seen
there. Having apparently resolved the problem of fish remains being assigned
to reef environments, Elizabeth Wing (personal communication) has suggested
that several of the species identified from the Pine Cay Site could have
derived from either of the two main environments. However, studies on differ-
ences in fish availability in these environments had not previously been
conducted. More specific research on patches within the tidal flats envi-
ronment is warranted.

The fishing study also provided data on the use of basketry traps as a
fishing device. Traps set on the reef suffered a variety of indignities
caused by wave action, damage apparently caused by nurse sharks, loss or
theft, etc. Wave action and water turbidity caused problems in recovering
the traps. At one point a trap was left unrecovered for three weeks due



to weather and water conditions. Those factors contributed to the smaller
sample size for fish captured on the reef (Table 3). While this short exper-
iment is not sufficient to eliminate the possibility that basketry traps were
used for fishing on the reef, it does suggest that more efficient techniques
could be devised. The traps used on the tidal flats suffered from similar
destructive forces, but not to the extent of those set on the reef.

Differences in the types of fish captured on the reef and tidal flats
do not appear to be significant (cf. Tables 2 and 3). The only possible
exception is the much higher frequency of capture for blue striped grunt
(Haemulon sciurus). However, a factor that may have contributed to this
difference was the higher than average capture of these fish on a single
occasion (26 H. sciurus, almost 27 percent of the total, were captured on
January 20, 1980). Therefore, it would be difficult to distinguish the
environment exploited from site samples if a mixed strategy of reef and tidal
flat fishing was practiced (cf. Davenport 1960; Cordell 1981).

When the fishing experiment was originally designed it was intended to
last for a one year period. A variety of problems conspired to limit the
length of the project (including trap loss and deterioration). Therefore,
no evidence of seasonal changes in fish availability were noticeable.

The comparison of capture data from the traps and the site provides an
indication of the fishing practices of the Lucayans. The high incidence of
bonefish (Albula vulpes) at the site suggests a focus on the tidal flat
environment (Table 1). Bonefish travel in schools and feed on benthic
organisms in shallow tidal flat environments (La Monte 1946). They are well
known for their fighting behavior when hooked; they are easily disturbed and
respond by scattering (Grosvenor 1965:133). The most efficient methods for
capturing bonefish in significant numbers would be those which maintained
the school's integrity and limited the fishes' ability to escape (e.g., nets,
corrals, or weirs). The absence of these fish from trap samples is probably
a function of their behavior, the patches in which the traps were set, and
present exploitation by game fishermen.

A second difference between site and trap samples is the absence of
yellowtail goatfish (Mullodichthys martinicus) from the site and their rela-
tively frequent occurrence in trap catches. Yellowtail goatfish are bottom
dwellers who sift the substrate for food. Their occurrence in facilities
set on the substrate is not surprising, and their absence from site deposits
may indicate that prehistoric fishing techniques did not effectively exploit
this patch. It is interesting that two fish species that exploit very similar
patches are absent from the sample in which the other occurs. That seems to
indicate that some variable other than habitat or patch may have had more of
an impact on the fishing technology that was employed.


The study of Lucayan subsistence from an adaptive perspective is sug-
gested, and models derived from evolutionary ecology and formalist economics
are proposed as a means of analyzing this process. The Pine Cay fishing
project was designed and conducted to acquire some of the data needed to
apply those models to fishing strategies in particular and Lucayan subsist-
ence in general.



The project involved the setting of Haitian-style basketry fish traps
in the two principle marine environments, the tidal flats and the reef.
These traps were emptied at approximately one week intervals and the fish
captured were recorded by Genus species, length, and weight. The purpose
of the project was to determine the availability of fishes in each environ-
ment, the approximate frequency of their availability, the problems asso-
ciated with the use of basketry traps, and the collection of a controlled
sample of fish captured in these facilities.

Sampling biases preclude the application of statistical techniques to
the site and trap samples. Therefore, this report is descriptive and possi-
bilistic. Future studies will attempt to avoid these shortcomings.

The study does indicate that the Lucayan inhabitants of the Pine Cay
Site could have obtained all of the fish represented at the site from tidal
flat patches. A further implication is that basketry traps were not the
principle fishing technique (due to the absence of certain bottom-dwelling
fish), and that nets, corrals, or weirs may have been used.

Fishing, even within a circumscribed estuarine environment (Cordell 1981),
involves complex strategies. The definition of patches within the primary
environments and the further analysis of fish behavior and availability should
provide a better understanding of the Lucayan's fishing strategies. By com-
bining that study with studies of the other Lucayan subsistence strategies,
a clearer picture of Lucayan adaptation can be developed.

When Charles A. Hoffman, Jr. titled his dissertation "Bahamas Prehistory:
Cultural Adaptation to an Island Environment" (1967), a demonstration of use
was considered sufficient to demonstrate adaptation. Today, more comprehen-
sive theories and models are being introduced to examine the problems of
adaptation in greater detail. The archaeological data base for the Bahamas
is surprisingly rich for the study of adaptation and cultural processes.
While this preliminary report has maintained the descriptive and possibil-
istic approach of most previous studies, this is a function of the lack of
attention given to maritime aspects of Caribbean prehistory (Watters 1981).
Further studies that employ specific methods derived from comprehensive
theories should provide a better understanding of what cultural adaptation
to an island environment really means.


The author would like to express his gratitude to William and Virginia
Cowles of the Meridian Club, Pine Cay, Turks and Caicos Islands. Special
thanks are owed to Chuck Hesse, Gary and Bennett Gardiner of the P.R.I.D.E.
Foundation, Pine Cay, for their assistance in monitoring the fishing project.
The Pine Cay fishing project was made possible by a generous grant from the
P.R.I.D.E. Foundation. The author also wishes to thank William H. Sears for
his help in organizing the project, Elizabeth S. Wing and Helen Donney for
their analysis of fish bones from the Pine Cay Site, and Timothy K. Earle
for his comments on earlier drafts of this paper. Finally, the author appre-
ciates the editorial assistance of Robert Carr for accepting this expanded
version of a paper presented at the Third Bahamas Conference on Archaeology.
Any errors are, of course, solely my responsibility.

Table 1. Fish remains identified from the Pine Cay Site.

Rank by No. of Weight Estimated
MNI Genus Common name specimens MNI (gins) Biomass (kg)

Epinephe us


Note: The ranking of identified fish by Genus was
with trap samples. Ranking these remains by other
more meaningful information.

done to permit the comparison of the site sample
categories (e.g., estimated biomass) may provide




Table 2. Summary of fish species captured in the tidal flat environment.

Total number of fish
captured, by:
Rank Genus species Common name species Genus



3 Haemulon

4 Epinephelus

Mu loidichthys









redtail parrotfish
yellowtail parrotfish
stoplight parotfish
mutton snapper
gray snapper
blue striped grunt
french grunt
sailor's choice
nassau grouper
rock hind
yellowtail goatfish
gray angelfish
french angelfish
blue tang
queen parrotfish
blue parrotfish
rainbow parrotfish
midnight parrotfish
yellowtail snapper

Table 2 (cont.). Summary of fish species captured in the tidal flats environment,

Total number of fish
captured, by:
Rank Genus species Common name species Genus

11 Lactophrys polygonia honeycomb cowfish 2 2
Holacanthus Angelfishes 2
ciliaris queen angelfish 1
tricolor rock beauty 1
13 Calamus calamus saucer-eye porgy 1 1
Diodon hoLocanthus balloonfish 1 1
Ho ocentrus rufus squirrelfish 1 1
Caranx bartholomaei yellowtail jack 1 1

Sample Total: 279

Note: Fish are ranked by Genus to facilitate comparison with fish remains identified for the
Pine Cay Site.

Table 3. Summary of fish species captured in the reef environment.

Total number of fish
captured, by:
Rank Genus species Common name species Genus



Mu ZZoidichthys

Ho lacanthus


Lachno aimus
13 FZanmeo

flavo ineatum


ci iaris


po lygonia

blue striped grunt
french grunt
mutton snapper
redtail parrotfish
yellowtail goatfish
nassau grouper
red hind
queen angelfish
blue parrotfish
rainbow parrotfish
saucer-eye porgy
jolthead porgy
honeycomb cowfish
blue tang
longspine squirrelfish
grass scorpionfish
queen triggerfish
four-eye butterflyfish
gray angelfish
rainbow runner (wrasse)

Sample Total:







References Cited

Butzer, Karl W.
1964 Environment and Archaeology: An Introduction to Pleistocene
Geography. Chicago: Aldine.

Cherry, John F.
1981 Pattern and Process in the Earliest Colonization of the
Mediterranean Islands. Proceedings of the Prehistoric
Society 47:41-68.

Cordell, John
1978 Carrying Capacity Analysis of Fixed-Territorial Fishing.
Ethnology 17:1-24.

Davenport, William
1960 Jamaican Fishing a Game Theory Analysis. Yale University
Publications in Anthropology No. 59. New Haven, Yale
University Press.

Earle, Timothy K. and Andrew L. Christenson, eds.
1980 Modeling Change in Prehistoric Subsistence Economies.
New York: Academic Press.

Granberry, Julian
1978 The Gordon Hill Site, Crooked Island, Bahamas. Journal
of the Virgin Islands Archaeological Society 6:32-44.

1980 A Brief History of Bahamian Archaeology. The Florida
Anthropologist 33:83-93.

Grayson, Donald K.
1981 The Effects of Sample Size on Some Derived Measures in
Vertebrate Faunal Analysis. Journal of Archaeological
Science 8:77-88.

Grosvenor, Melville Bell, Editor-in-Chief
1965 Wondrous World of Fishes. Washington, D.C.: National
Geographic Society.

Hawkes, Kristen, Kim Hill and James F. O'Connell
1982 Why Hunters Gather: Optimal Foraging Theory and the Ache
of Eastern Paraguay. American Ethnologist 9:379-398.

Hill, James N., ed.
1977 Explanation of Prehistoric Change. Alburquerque: New
Mexico Press.

Hoffman, Charles A., Jr.
1967 Bahamas Prehistory: Cultural Adaptation to an Island
Environment. Ph.D. dissertation, University of Arizona,
University Microfilms, Ann Arbor.



Jochim, Michael A.
1979 Breaking Down the System: Recent Ecological Approaches
in Archaeology. IN Advances in Archaeological Method and
Theory (Vol. 2), Michael Schiffer (ed.), New York: Academic

Keegan, William F.
1981 Artifacts in Archaeology: A Caribbean Case Study. MA
thesis, Florida Atlantic University, University Microfilms,
Ann Arbor.

Keene, Arthur S.
1981 Optimal Foraging in a Nonmarginal Environment: A Model of
Prehistoric Subsistence in Michigan. IN Hunter-Gatherer
Foraging Strategies, Bruce Winterhalder and Eric Alden
Smith (eds.), Chicago: University of Chicago, pp. 171-193.

Kirch, Patrick V.
1980 The Archaeological Study of Adaptation: Theoretical and
Methodological Issues. IN Advances in Archaeological
Method and Theory (Vol. 3), Michael Schiffer (ed.), New York:
Academic Press, pp. 101-156.

Krebs, John R.

Optimal Foraging: Decision Rules for Predators. IN
Behavioral Ecology, J. R. Krebs and N. B. Davies (eds.),
Oxford: Blackwell Scientific Publications, pp. 23-63.

La Monte, Francesca
1946 North American Game Fishes. Garden City: Doubleday.

MacLaury, James C.
1968 Archaeological Investigations on Cat Island, Bahamas.
Unpublished MA thesis, Department of Anthropology, Florida
Atlantic University, Boca Raton.

Mueller, James W., ed.
1975 Sampling in Archaeology. Tucson: University of Arizona

Pianka, Eric F

Evolutionary Ecology. New York: Harper and Row.

Rainey, Froelich G.
1934 Diary Beginning January 22, 1934, Upon Arrival in
Port-au-Prince, Haiti. Manuscript on file at the Yale
Peabody Museum, New Haven.

Rohan-Csermak, Geza de
1963 Sturgeon Hooks of Eurasia. Viking Fund Publications in
Anthropology 35, Wenner-Gren Foundation.

Rouse, Irving

The West Indies, IN Handbook of South American Indians,
Julian H. Steward (ed.), Vol. 4, The Circum-Caribbean
Tribes, Pt. 3, Cooper Square, New York.



Sears, William H. and Shaun D. Sullivan
1978 Bahamas Prehistory. American Antiquity 43:3-25.

Sullivan, Shaun D.
1974 Archaeological Reconnaissance of Eleuthera, Bahamas.
Unpublished MA thesis, Department of Anthropology,
Florida Atlantic University, Boca Raton.

1976 Archaeological Reconnaissance of the Turks and Caicos
Islands, British West Indies. Report submitted to the
government of the Turks and Caicos Islands.

1980 An Overview of the 1976 to 1978 Archaeological Investigations
in the Caicos Islands. The Florida Anthropologist 33:120-142.

1982 Prehistoric Patterns of Exploitation and Colonization in
the Turks and Caicos Islands. Ph.D. dissertation, University
of Illinois at Champaign-Urbana, University Microfilms, Ann

Thomas, David Hurst
1978 The Awful Truth About Statistics in Archaeology. American
Antiquity, 43:231-244.

Watters, David R.
1981 Assessing the Ocean's Roles in Antillean Prehistory. Paper
presented at the 9th International Congress for the Study
of Pre-Columbian Cultures of the Lesser Antilles, Santo
Domingo, Dominican Republic.

Wing, Elizabeth S.
1966 Zooarchaeology Report. First Bahamas Conference on
Archaeology, San Salvador, Bahamas, Yale University
Department of Anthropology, New Haven, pp. 7-8.

1969 Vertebrate Remains Excavated from San Salvador Island,
Bahamas. Caribbean Journal of Science 9:25-29,

Wing, Elizabeth S. and Antoinette Brown
1979 Paleonutrition. New York: Academic Press.

Wing, Elizabeth S., Charles A. Hoffman, Jr. and Clayton E, Ray
1968 Vertebrate Remains from Indian Sites on Antigua., West
Indies. Caribbean Journal of Science 8:123-139.

Winterhalder, Bruce and Eric Alden Smith, eds.
1981 Hunter-Gatherer Foraging Strategies. Chicago: University
of Chicago Press.

William F. Keegan
c/o Department of Anthropology
University of California
Los Angeles, California 90024


Storrs L. Olson

It is time to recognize a new direction in archeology, particularly on
islands, that for want of another term I shall call "biological archeology."
Whereas zooarcheology, as usually practiced, consists mainly of providing
faunal lists of animal remains, biological archeology emphasizes the impact
that prehistoric (and in some cases, historic) man had on natural environments
and the biologically significant information that is retrievable from archeo-
logical sites.

Biological archeology is concerned to a large extent with the extinction
of organisms, whether extirpation of a local population or the total ex-
tinction of species. From such data the biologist hopes to learn of the
existence of previously unknown species, or of zoogeographic patterns that
are not apparent from the historic distribution of living forms. The anthro-
pologist can learn about which animals were used in the culture under study,
about possible instances of animal husbandry, and about the resources that may
have been available to the earliest colonizers of an island or archipelago
that were not necessarily available for following generations. Man-induced
alterations in environment for agricultural or other purposes have caused many
extinctions. Documenting the extent, nature, and timing of such alterations
is the concern both of the biologist and of the archeologist, each of whom
stands to benefit from broadening the scope of their investigations and from
developing new techniques.

One of the best recent illustrations of the impact of man on insular
faunas comes from combined paleontological and archeological investigations
in the Hawaiian Islands. It has recently been determined that well over half
of the species of birds that originally occurred in the archipelago were exter-
minated by man in the prehistoric period (Olson and James in press a,b,c).
Bones of a number of these species were found in direct archeological contexts,
thus proving their contemporaneity with Polynesians. Some species are as yet
known only from late Holocene fossil deposits that antedate the arrival of
man. As these species all survived any Pleistocene climatic changes that may
have influenced the islands, the most likely cause of their extinction is
prehistoric man. In this case, the purely paleontological deposits still
contain information pertinent to anthropologists, as they document potential
food resources for the earliest colonizers that were subsequently lost, such
as large flightless geese and certain seabirds. These doubtless would have
helped sustain the first human populations until a reliable agricultural
system could be established.

Extinction of a number of species of Hawaiian birds, particularly large
flightless cones, can be attributed to simple overexploitation. The extinction
of many species of small arboreal passerines, such as finches, cannot be ex-
plained in this manner, however. The paleontological evidence, in tandem with
that from botany, archeology, and other sources, points towards the virtual
annihilation of dry, lowland forest habitats as the primary cause of the dis-
appearance of large portions of the original fauna and flora of the Hawaiian




Islands. This destruction by fire and other means, made land available for
the extensive cultivation of taro and sweet potatoes, which in turn made
possible the large populations of Polynesians observed in the islands at the
time of western contact.

From the new paleontological data it is clear that few reliable zoo-
geographical or other biological inferences can be made about Hawaiian birds
without reference to the fossil record. It is also evident that a sound
knowledge of the prehistory of the Hawaiian archipelago would not be possible
without extensive interchange between biologists and archeologists.

In the West Indies, there is a fairly good fossil record of vertebrates,
at least from the Greater Antilles. From this, it has been possible to docu-
ment numerous extinctions and changes in distribution of reptiles, birds, and
mammals (Olson 1978; Pregill and Olson 1981). Many of these extinctions have
been attributed to Pleistocene changes in climate and sea level. It has been
hypothesized that during the last glaciation the West Indies not only com-
prised fewer and larger islands, but they were generally more arid, with
extensive areas of grassland and savanna. The postglacial reduction in land
area, combined with the expansion of more mesic environments are believed to
have led to the extinction or fragmentation in range of many species of
vertebrates (Pregill 1981; Pregill and Olson 1981; Olson 1982a,b; Olson and
McKitrick 1982).

Whereas some extinctions appear almost certain to be the result of
Pleistocene climatic changes (e.g. the disappearance of grassland-inhabiting
meadowlarks, Sturnella, from the Bahamas [Olson and Hilgartner 1982]), others
are known to be more recent. In the past, these recent extinctions have been
documented almost entirely through archeological excavations.

Several species of rodents are known from middens and fossil deposits but
nave never been taken in life by naturalists. The extinct rodents Brotomys
voratus, Isolobodon portoricensis, Plagiodontia hylaeum, and Quemisia gravis
were popular human food on Hispaniola and may be abundant in midden deposits
(Miller 1929). In Puerto Rico and the Virgin Islands, Isolobodon portori-
censis is known only from midden deposits and it has long been speculated
that this species was transported to those islands by man (Miller 1918;
Reynolds et al. 1953). My own paleontological work in Puerto Rico convinces
me that this was the case, as I never found Isolobodon outside of a cultural
context and it is invaribly absent in Pleistocene cave deposits (Olson and
Pregill 1982). I have also examined specimens of one of the endemic species
of Heteropsomys from a midden deposit in Puerto Rico, which indicates that
it, too, persisted well into the Holocene. A large, extinct genus and species
of rice rat, as yet undescribed, occurs abundantly in middens on Antigua, and
it, or a similar form, is known from midden material from Monserrat and
Guadeloupe as well (Wing et al. 1968). It also occurs as a fossil on Barbuda.

The cause of the extinction of these rodents has not been determined. A
decrease in abundance of bones of the extinct rice rat in the upper midden
layers on Antigua was interpreted as "possibly indicating marked reduction
of the rat population, perhaps as a result of human predation" (Wing et al.
1968:128), but habitat destruction cannot be ruled out either, as we shall

Transportation of endemic West Indian animals from one place to another
by aboriginal man has been documented in a number of instances and should be



an important consideration in attempting to reconstruct the natural dis-
tribution of various organisms. I have already mentioned the case of the
rodent Isolobodon. A similar one involves the transport of the Cuban rodent
Capromys pilorides, which has been found in a midden in the Dominican Republic
(Miller 1929; Rimoli 1974).

Transportation of endemic birds also occurred with the large, extinct,
flightless rail Nesotrochis debooyi, which has been found in fossil deposits
in Puerto Rico and in midden deposits in Puerto Rico and the Virgin Islands,
including St. Croix (Olson 1974). Because St. Croix is not on the Puerto
Rican Bank and was never connected with Puerto Rico or the other Virgin Is-
lands, it is highly unlikely that it would have had the same species of
flightless bird as those islands. Thus it seems probable that man was trans-
porting and very likely raising Nesotrochis in captivity (Olson and Pregill
1982). Another instance involves the macaw Ara authocthones, which was
described from a single bone from a midden on St. Croix (Wetmore 1937). This
specimen does appear to be from an extinct species, but its provenience is far
from certain. The only macaw definitely known from the West Indies is Ara
tricolor of Cuba, which became extinct in the 19th century. Several species
of macaws were attributed to other Antillean islands by early explorers and
were subsequently named despite the absence of specimens (Greenway 1958).
Given the fact that there was a lively trade in macaws between Indians in
tropical Mexico and those in the desert southwest,it seems probable that trade
in macaws would have been carried on in the West Indies as well, so that Ara
autocthones may well not have been autochthonous to St. Croix at all.

So far, I have mentioned only rather large animals that could possibly
have been exterminated directly, by overhunting. But we also have evidence
of habitat destruction causing the extinction of smaller vertebrates, as in
Hawaii. In 1980, G. K. Pregill, D. W. Steadman, and I excavated a fissure
fill in a limestone quarry on Antigua that contained numerous bones of small
vertebrates such as lizards, snakes, frogs, bats, birds, and rodents. In
the same deposit were a few pieces of worked flint and remains of marine
mollusks and crustacea that most likely represent midden material. The source
of the vertebrate bones, which are largely from species that are unlikely to
have been eaten, is as yet undetermined. This site has been radiocarbon
dated at about 3500 years B.P., which is well into the Holocene and a late
date for fossils but an early date for man in the Lesser Antilles. One of
the bats from this site is a species that is otherwise known only from the
Pleistocene of Puerto Rico and one of the lizards is known only from the
Pleistocene of Barbuda. There is an extinct snake and two or more taxa of
extinct rodents, as well as a bird (Cinclocerthia ruficauda) that is wide-
spread in the Lesser Antilles but that has never been recorded on Antigua.
Because these small animals could hardly have been hunted to extinction,
habitat destruction within the past 3500 years is the most likely cause of
their disappearance. We do not know whether these extinctions took place in
the pre-Columbian period of aboriginal cultivation, or in the colonial period
when agriculture was much more intense (Harris 1965). Thus there is biological
information potentially available even from colonial archeological sites.

The Antiguan site illustrates several important points. First, a
"fossil" site that might otherwise seem of little importance to paleontolo-
gists because of its late age, and of no significance to archeologists because
of its poverty of cultural evidence, may actually provide data of considerable
interest to both disciplines. Second, we see once again that zoogeographers
cannot rely on historical data for reliable information. And third, it is



absolutely essential to use fine-mesh screen when sampling sites that are
at all likely to yield vertebrate remains. Most of the specimens we re-
covered from Antigua would have been missed had we not used window screen
(1.5 mm). Such labor would not be repaid at most archeological sites but
samples of sediment could be checked periodically when larger bones are
regularly encountered. The 1/4 inch (7 mm) mesh that is usually used at
archeological sites is practically useless for making accurate faunal surveys,
whereas with 1/8 inch (3.5 mm) mesh, all but the smallest vertebrate bones
will be recovered.

As the focus of this conference is to some extent on the Bahamas, I will
point out some interesting problems in biological archeology in those islands.
So far, the main archeological site in the Bahamas with interesting vertebrate
remains is that excavated on Crooked Island in 1933-1934 by Froelich Rainey.
The mammals and birds from the Crooked Island site have been treated by Law-
rence (1934), Wetmore (1938), Olson and Pregill (1982), and Olson and Hil-
gartner (1982). No date other than "pre-Columbian" was suggested for these
deposits. Other significant vertebrate localities in the Bahamas are appar-
ently Pleistocene in age and are restricted to New Providence and Little
Exuma (Olson and Pregill 1982).

The Bahamian hutia (Geocapromys ingrahami) is known as a living animal
only from tiny East Plana Cay. It has been reported from middens on Crooked
Island (Lawrence 1934) and San Salvador (Wing 1969), and as a fossil or sub-
fossil from Abaco, New Providence, Andros, Great and Little Exuma, and the
Cat Islands (Olson and Pregill 1982). Three subspecies have been recognized
(Lawrence 1934), but the validity of these is questionable without more
extensive series of specimens from the various islands. If the hutia was
able to survive up to the present on East Plana Cay, why did it become extinct
elsewhere? And when did it become extinct on each of the various islands it
inhabited? Was it in the process of dying out because of post-Pleistocene
changes in habitat and was only helped along by man, or was its demise due
entirely to overexploitation and habitat destruction? If the Lucayans main-
tained the hutia in captivity, as the inhabitants of Puerto Rico and the Virgin
Islands must have done with Isolobodon, then they could well have transported
the animals to islands where they never occurred naturally or from which they
had already disappeared due to natural processes (Olson and Pregill 1982).
Thus the occurrence of Geocapromys in an archeological site does not indicate
that it occurred on that island in a natural state. Only with accurately
dated archeological and paleontological excavations on many islands will it
be possible to determine the natural distribution and time of extinction of
the various populations of Geocapromys in the Bahamas.

In the material from Crooked Island were remains of White Ibis (Eudocimus
albus), Cuban Parrot (Amazona leucocephala), and Cuban Crow (Corvus nasicus),
none of which have been recorded from the island in historic times. The crow
occurs in the Bahamas today only in the Caicos but is known from fossils from
New Providence and Little Exuma (Olson and Hilgartner 1982). Its occurrence
in midden on Crooked Island indicates that its disappearance from most of the
Bahamian archipelago either occurred during, or continued into, the Holocene.
The parrot, too, had a wider distribution in the Bahamas in the past and was
presumably extirpated from many islands by hunting for food and trapping for

Most puzzling, however, is the occurrence among the Crooked Island
material of two species of petrels (Olson and Hilgartner 1982)-- the Cahow


(Pterodroma cahow), known to breed only on Bermuda, and the Black-capped
Petrel (P. hasitata), which is known to breed only on Cuba, Hispaniola, and
certain of the Lesser Antilles. Except when they come to land to nest in
burrows or crevices in rocks, these birds are strictly pelagic and would have
been almost impossible to capture by any sort of aboriginal hunting methods.
On land, however, they are quite vulnerable to predation and make a ready
source of tasty protein. The Cahow, for example, was almost exterminated
very quickly by the early settlers of Bermuda and persists only in small
numbers. Thus, we are confronted with the possibility that one or both of
these petrels may have bred in the Bahamas, or elsewhere in the West Indies,
where neither species has been recorded. Intense human predation may have
exterminated petrels from much of their natural range early in the period of
settlement of the West Indies. Without the attention of archeologists and
paleontologists to this problem, we may never know what the true breeding
distribution of Pterodroma was in the Antilles.

It is just this sort of interdisciplinary exchange that I hope to
encourage in this essay. Biologists have great need of information derived
from archeological sites. The implications of their findings should have
direct application to a better understanding of the environment and resources
of the human cultures that are studied by archeologists. We can no longer
afford to overlook the biological aspects of archeology.

References Cited
Greenway, James C.
1958 Extinct and Vanishing Birds of the World. American Committee
for International Wild Life Protection Special Publication

Harris, David R.
1965 Plants, Animals, and Man in the Outer Leeward Islands, West
Indies. An Ecological Study of Antigua, Barbuda, and Anguilla.
University of California Publications in Geography 18.

Lawrence, Barbara
1934 New Geocapromys from the Bahamas. Occasional Papers of the
Boston Society of Natural History 8:189-196.

Miller, Gerrit S., Jr.
1918 Mammals and Reptiles Collected by Theodoor de Booy in the
Virgin Islands. Proceedings of the United States National
Museum 54:507-511.

1929 Mammals Eaten by Indians, Owls, and Spaniards in the Coast
Region of the Dominican Republic. Smithsonian Miscellaneous
Collections 82(5):16,2 pl.

Olson, Storrs L.
1974 A New Species of Nesotrochis from Hispaniola, With Notes
on Other Fossil Rails from the West Indies (Aves: Rallidae).
Proceedings of the Biological Society of Washington 87:439-450.



1978 A Paleontological Perspective of West Indian Birds and
Mammals. In Frank B. Gill, ed., Zoogeography in the
Caribbean. The 1975 Leidy Medal Symposium. Academy of
Natural Sciences of Philadelphia Special Publication 13:99-117.

1982a A New Species of Palm Swift (Tachornis: Apodidae) from the
Pleistocene of Puerto Rico. Auk 99:230-235.

1982b (Editor) Fossil Vertebrates from the Bahamas. Smithsonian
Contributions to Paleobiology 48.

Olson, Storrs L. and William B. Hilgartner
1982 Fossil and Subfossil Birds from the Bahamas. In Storrs L.
Olson, ed., Fossil Vertebrates from the Bahamas. Smithsonian
Contributions to Paleobiology 48:22-60.

Olson, Storrs L. and Helen F. James
a Fossil Birds from the Hawaiian Islands: Evidence for
Wholesale Extinction by Man Before Western Contact.
Science (in press).

b Prodromus of the Fossil Avifauna of the Hawaiian Islands.
Smithsonian Contributions to Zoology (in press).

c The Role of Polynesians in the Extinction of the Avifauna of
the Hawaiian Islands. In Paul S. Martin, ed., Late Quaternary
Extinctions. Tucson: University of Arizona Press (in press).

Olson, Storrs L. and Mary C. McKitrick
1982 A New Genus and Species of Emberizine Finch from Pleistocene
Cave Deposits in Puerto Rico (Aves: Passeriformes). Journal
of Vertebrate Paleontology 1:279-283.

Olson, Storrs L. and Gregory K. Pregill
1982 A Brief Introduction to the Paleontology of Bahamian
Vertebrates. In Storrs L. Olson, ed., Fossil Vertebrates
from the Bahamas. Smithsonian Contributions to Paleobiology

Pregill, Gregory K.
1981 Late Pleistocene Herpetofaunas from Puerto Rico. Miscellaneous
Publications of the University of Kansas Museum of Natural
History 71.

Pregill, Gregory K. and Storrs L. Olson
1981 Zoogeography of West Indian Vertebrates in Relation to
Pleistocene Climatic Cycles. Annual Review of Ecology
and Systematics 12:75-98.

Reynolds, Thomas E., Karl F. Koopman and Ernest E. Williams
1953 A Cave Faunule from Western Puerto Rico with a Discussion
of the Genus Isolobodon. Breviora 12.

Rimoli, Renato O.
1974 Prueba Zooarqueologica de Transito del Hombre Precolombino
de Cuba Hacia La Hispaniola. Revista Dominicana de Antro-
pologia e Historia 4:27-28.



Wetmore, Alexander
1937 Ancient Records of Birds from the Island of St. Croix with
Observations on Extinct and Living Birds of Puerto Rico.
Journal of Agriculture of the University of Puerto Rico


Wing, Elizabet

Bird Remains from the West Indies. Auk 55:49-55.

:h S.
Vertebrate Remains Excavated from San Salvador Island,
Bahamas. Caribbean Journal of Science 9:25-29.

Wing, Elizabeth S., Charles A. Hoffmann and Clayton E. Ray
1968 Vertebrate Remains from Indian Sites on Antigua, West
Indies. Caribbean Journal of Science 8:123-139.

Storrs L. Olson
National Museum of Natural History
Smithsonian Institution
Washington, D.C. 20560



Irving Rouse

In 1946 I was asked to contribute an article to the newly established
Bulletin du Bureau d'Ethnologie de la Republique d'Haiti. I chose to devote
it to the finds made in 1933 on Ile a Vache, Haiti by Godfrey S. Olsen of
Newburgh, N.Y., under the auspices of the Museum of the American Indian,
Heye Foundation, New York City. These finds seemed to me worth publishing
because they extended the range of the Archaic (preceramic) occupation of
Haiti from the northern part of that country, where Rainey (1941)and I (1941)
had documented it, to the southern part, and provided further examples of
the former's unexpectedly elaborate style of decoration.

Soon after I submitted the article, the editor to whom I sent it lost
his job in a revolution. Both of my plates, one of two figures, and two pages
in the middle of the text disappeared in the process, but the new editor pro-
ceeded to publish what was left (Rouse 1947).

Recently Alfredo Figueredo encouraged me to republish the article in
its entirety. Before doing so, I decided to visit and test the site, in
order to clear up several doubtful points about it. I am grateful to Clark
Moore for organizing this visit and for following it up with his own research.
as reported in the accompanying paper (Moore 1982). The following version of my
original article has been revised to incorporate the new information pertinent
to Olsen's finds. I have also changed the conclusions.

The Site

Olsen referred to the site as Bay de Conch, but I shall here call it
Cacoq 2, following Moore's practice in the previous article. I am relieved
to be able to abandon the original combination of English and French terms.

No information concerning the site is on file with Olsen's collection
in the Museum of the American Indian, and efforts to obtain such information
from Olsen himself were unsuccessful. In the original version of this article,
therefore, I was forced to rely for knowledge of the site on a brief note by
Dr. Froelich G. Rainey, who visited Ile a Vache shortly after Olsen finished
his excavations, as explained by Moore in the previous paper (Rainey MS; see
also Rainey 1940:132, 1941:4; Rouse 1941:157). Moore's research demonstrates
that his Cacoq 2 is indeed Olsen's Bay de Conch site and provides us for the
first time with an adequate description of the stratigraphy.

In the Museum of the American Indian catalogue, 70 of Olsen's specimens
are said to have come from a "mound" and the remaining 16 from a "kitchen
midden". In the original version of this article, I was unable to determine
whether these were separate sites or parts of the same one. Moore's work
indicates that the word "mound" probably refers to the overall accumulation
of shells at Cacoq 2 and the word "kitchen midden" to the horizontal layer




of ash and other kinds of food remains near its base (Moore 1982, Fig 4).
This situation is not unique to Caribbean archeology; for example, there is
similar stratigraphy at the site of Cayo Cofresi on the south coast of Puerto
Rico (Veloz Maggiolo et al. 1975:16-17).

Chipped Stone Artifacts

Flint Blade (Pl. 1,A)
This is apparently the only flint specimen in the collection to have
been struck off a prepared core. It is a fragment of a blade of dark brown
flint, broken at both ends and therefore lacking a striking platform and bulb
of percussion. Roughly rectangular in shape, it has the usual single flaking
facet on the bottom, but bears no such facets on its upper side, whence it
may be deduced that the blade was struck off the surface of the nodule. That
surface has been rechipped thoroughly, however, leaving no traces of the
original patina of the nodule. This specimen may have been used as a knife
or have been part of a spearhead or scraper like those found in northern Haiti
(Rouse 1941:25-32). It comes from the "kitchen midden"; that is, the ash

Other Flint Fragments (P1. 1,B,C)
There are 41 other pieces of flint in the collection, amounting to almost
half the total number of specimens. All seem to be naturally shaped fragments,
picked up and brought to the site by the aborigines because they happened to
have shapes suitable for use as artifacts. They vary in material from pure,
tan flint to dark blue chert,in length from 3.8 cm to 12.7 cm, and in shape
from flakes suitable for use as knives or scrapers to two small lumps which
could have had no conceivable function. Only two of the flakes have bulbs
of percussion on the flat under surface (e.g., P1. 1,C), but except on the
specimen illustrated, these are irregular and do not seem to me to be the
result of human agency. In the absence of regular facets, the top of many
of the specimens still have the calcareous,kaoliniferous surface of the orig-
inal nodule (e.g., P1. 1,B). Traces of modification by human agency are
limited to the edges of the specimens. They form the basis for classifying
the specimens into three groups:

1. Flake with secondary chipping. A single piece of reddish brown
chert, encrusted on the top with the white surface of the original nodule,
is heavily rechipped along three edges, perhaps in order to sharpen or
resharpen the specimen for use as a knife (Pl. 1,B). It comes from the
"mound"; that is, the shell refuse.

2. Flakes showing traces of use. Fifteen specimens have only slight
nicks along the edge, which may be the result of use rather than manufacture.
(The Museum of the American Indian catalogue, however,calls them manufactured.)
Twelve of these, with long, sharp edges, may be considered knives; three
stubby, blunt flakes would be serviceable as scrapers. Nine come from the
shell refuse, and four from the ash layer. A specimen from the second group
is illustrated (P1. 1,C); it is composed of chert and has an ideal shape
for use as a knife.

3. Flakes and other fragments showing traces of neither manufacture nor
use. These include 23 flakes and the two lumps discussed above, all of which
seem to have been picked up and utilized, if at all, without change. Nineteen
appear to have the proper shape for use as knives and two as scrapers. All
come from the shell refuse.


Ground Stone Artifacts

Double-bitted Ax (P1. 1,E)
This oval specimen, greyish white in color, is composed of felsosphyre.
Its two surfaces, the upper more convex than the lower, meet at an edge
around the sides and ends of the specimen and are broken only in the center,
where a groove encircles the ax. The surfaces are ground smooth, but the
groove has been pecked roughly into the specimen. The bits are at either
end of the specimen; neither is sharp. A small chip is missing from one of
them, perhaps as the result of use. This ax comes from the ash layer.

Single-bitted Axes (Pl. 1,D,F,G)
There are two other complete axes and one which is fragmentary. All
have only a single bit, and are provided at the other end with a neck sur-
mounted by ears.

One of these axes is illustrated on Plate 1,D. Although composed of
porous grey tuff, it has an even surface. The bit is oval in outline and
also in cross section, the neck is slightly grooved, and the ears are rounded.
Chips are missing from the edge of the bit, possibly as a result of use, and
one of the ears is also broken off. This specimen is from the shell refuse.

The fragmentary ax illustrated as Plate 1,G must have been the largest
of the three. It is composed of a fine-grained grey andesite which could be
highly polished but is only ground smooth like the rest of the specimens.
The bit of this ax is entirely missing, leaving only a shallowly grooved neck,
oval in cross section, and a pair of gracefully spreading ears decorated on
either edge with a series of tiny triangular notches. The location is again
the shell refuse.

The ax illustrated as Plate 1,F is probably composed of limestone, but
this is obscured by a soft kaolin coating, badly eroded on the back. The bit
is semicircular and is flattened on the bottom as if the specimen had been
resharpened after use, the neck is plain and oval in cross section, and the
wings are large and rectangular. The two sides of the specimen are crudely
engraved with identical rectilinear designs, each consisting of a T-shaped
line flanked by a pair of spirals. The central T is positive but the two
spirals are outlined in the negative. All of the lines are relatively shallow;
they seem to have been made with a sharp tool in an irregular manner, as if
the artisan were unfamiliar with the technique. The work reminds one, in
fact, of the incision on Haitian pottery of the Meillac type rather than
other known examples of stonework (cf. Rouse 1941:64-68). This ax also
comes from the shell refuse.

Discoidal Hammer-grinders (P1. 1,J)
There are four discoidal stones in the collection. One appears to be
purely ceremonial and will be described as such below. The other three are
utilitarian; they probably served as hammers, as grinders, or for both pur-
poses. All three are pebbles which appear to have been picked up and used
in their natural form with only slight modification.

The specimen illustrated in Plate 1,J is composed of very coarse chert,
which has taken on a greyish white patina. A pit has been pecked in each
side; the edge is heavily scarred, apparently as a result of hammering.
There is no grinding facet, nor would one be expected from the coarseness
of the material. The object comes from the shell refuse.



A second specimen is made of finer-grained chert but has the white sur-
face color, which in this case, is the result of a calcareous, kaoliniferous
coating, worn smooth by the water. It would be perfectly oval in cross
section if the lower surface were not somewhat flattened, possibly as a
result of use for grinding. Its surfaces are not marred by pits nor by
hammering facets. It, too, is from the shell refuse.

The third specimen is also a chert nodule. It has a smooth, fine-grained
surface, green in color. Its top surface, which is slightly convex, is prob-
ably that of the original nodule; the bottom surface, however, seems to have
been flattened by grinding. Its edge, which is also somewhat convex, is
heavily pitted all the way round, as if through hammering. It comes from
the ash layer.

Irregular Hammerstone
There is another ovoid nodule of coarse chert which lacks the regularity
of the discoidal stones just described. Its surface is greyish-white in color
and is rather pocked, as if the artifact were a natural pebble. Its side
edge is rougher than either the top or the bottom, possibly as a result of
use for hammering. It comes from the shell refuse.

Rectangular Hammer-grinder (Pl. 1,J)
This piece of dark grey, finely grained siltstone is more or less rec-
tangular in shape despite the fact that its surfaces are rather convex. Its
ends show evidences of hammering and the top and bottom surfaces, of grinding,
but its side edges, which are slightly irregular, appear to retain the orig-
inal exterior of the stone. It comes from the shell refuse.

Single-pointed Smoothing Stone (P1. 1,I)
This is a fragment of serpentine, weathered white on the surface but still
green inside. The surface is smooth and the object appears to have been in-
tentionally ground into its present shape, which is that of a prong extending
upward at one end of a broad base, the other end of the base being broken off.
The top is pointed and the base highly polished, as if through use for grind-
ing. The specimen comes from the shell refuse.

Bipointed Smoothing Stones (P1. 1,H)
Two snake-like pieces of light, white limestone also appear to have been
used as smoothers. Their bottom surfaces have been ground down but the other
surfaces appear to be natural. The ends are irregular and pointed. One
specimen has chips removed from both ends and the other is broken slightly
at one end, suggesting that the two tools were also used as picks.

The ground surface of the specimen illustrated is flat from end to end.
It is from the shell refuse.

The ground surface of the other specimen is bowed at one end. It has
the same provenience as the illustrated artifact.

Oval Milling Stone (Pl. 1,L)
This thin, oval piece of sandstone is an unusually fine example of its
kind. It has a convex bottom and a flat top, containing a shallow depression
that extends close to its edge. Its surface is irregular and pitted except
in the depression, which is smooth and gritty to the touch. It comes from
the shell refuse.



Conical Pestles (Pl. 2,A,B)
One of these two artifacts (Pl. 2,B) is almost certainly a pestle. It
has the proper shape--flat, expanded base, circular cross section, and taper-
ing, rounded top--and its base shows evidence of use for grinding. The edge
of its base is battered, perhaps also as the result of use. It is made of
tuff. It comes from the shell refuse.

The second specimen is also catalogued by the Museum of the American
Indian as a pestle, but it may instead be a ceremonial object. It is a con-
ically-shaped stalagmite, broken off above its base so that it is impossible
to determine whether it was indeed a pestle. Its surface has been engraved
with a trio of double spirals, inclined towards the left and bordered at the
top and bottom by a pair of horizontal lines (P1. 2,A; Fig. 1,A). The tool
used to produce the lines is blunter than in the case of the other artifacts,
the lines are surer, and they are uniquely curvilinear. They resemble the
lines of petroglyphs, but the design they portray is different. The specimen
is from the shell refuse.

Mortar (Pl. 2,D)
This artifact differs from the milling stone described in that it has a
more pronounced depression and was probably used with a pestle rather than a
hammer-grinder. It has broken in half. Made of limestone, it is grey in
color except on the inside where a brownish-black discoloration may be the
result of use. The outside surface is fairly rough, whereas the interior seems
to have been ground smooth. This specimen comes from the shell refuse.

Stone Vessels (Pl. 2,E)
There are two fragments of stone vessels. The one illustrated is of
serpentine and has a soft, green surface, rough inside and out. It still
bears on the inside scratches of the sharp tool used to make it. The fragment
comprises one side and a part of the rim and bottom of the original dish.
Apparently, it was a round bowl with flat bottom, convex underbody, and side
curving slightly inward from the vertical (Fig. 1,B). In profile, the wall
varies considerably, being thickest through the underbody and tapering through
the side to a plain round rim. The broken edge along the left side of the
specimen has been worn smooth, as if through use after breakage. The right
edge passes through a hole just beneath the rim, which may have been made to
mend a crack or for attachment to some other object. Looked at from the out-
side, this hole is conical in shape. The entire outer surface, down to and
including the bottom, is completely covered with a complex incised design
consisting of a series of interlocking swastikas, executed in a positive
technique and bordered on the top, along the rim, by a row of hexagonal de-
pressions placed between two rows of alternatingly hatched triangles attached
to parallel lines (Fig. 1,B). This design has been made with a sharp tool,
possibly a piece of flint, in a purely rectilinear fashion, as in the case
of the design on the single-bitted ax described. As in that case, the lines
are relatively narrow and, although deep, they reveal no surety of technique,
This, however, may simply be due to the greater difficulty in incising with
a sharp tool, as opposed to engraving with a blunt one, as on the supposed
pestle. The specimen is from the shell refuse.

The second fragment is also of serpentine, but only includes a section
of the rim. Presumably, it is from another and larger bowl than the first
specimen. It lacks decoration. The outside and the rounded rim have a rela-
tively smooth surface, but the inside is rough. This specimen is also from
the shell refuse.



Figure 1. Decorative Designs from Cacoq 2, Ile a Vache, Haiti:
A, Conical Pestle; B, Stone Vessel.



Stone Balls (Pl. 2,C)
Two balls of stone are included in the collection. The one illustrated
is of limestone, but its heaviness suggests the presence of a chert core.
It is almost perfectly round and has a smooth white surface, unpolished but
not marred by any imperfections. It comes from the ash layer.

The other stone ball is of chert, and only roughly approximates a sphere
in shape. Its surface is rough and unpolished, and it would be possible to
consider it a hammerstone rather than a ceremonial object. It is from the
shell refuse.

Plain Stone Disk (Pl. 2,H)
This object is made of chert and has a fine-grained surface coating of
kaolin, which gives it a greyish-white color. Except for some variation in
thickness, it is a perfectly symmetrical disk, with a flat, rectangular edge
and convex upper and lower surfaces. These are quite smooth except on the
bottom, where weathering seems to have taken place. There are no pits nor
grooves, and the convexity of the surfaces is so great that it seems doubtful
that the specimen could have been used for grinding. Hence it is termed a
disk, in the belief that it may have had some ceremonial significance. It
comes from the shell refuse.

Biconical Stone Disks (Pl. 2,F)
Two other disks have sides which are conical rather than convex. The
one illustrated is of shale, and has the same greyish-white patina as the
majority of the other specimens. The most graceful of all the artifacts, it
consists of a pair of biconical surfaces, rising on either side of the arti-
fact from a grooved edge to a central hole. These two surfaces are well
smoothed, perfectly symmetrical, and they trace in cross section a barely
perceptible S-shaped curve. By contrast, the groove around the edge of the
specimen is rough and still bears traces in its bottom of the sharp-pointed
tool used to produce it. The holes drilled in the center of each side are
conical in shape; apparently they were never intended to penetrate through
the specimen. Its context is the shell refuse.

The other biconical disk, a much cruder example, is made of limestone,
perhaps with a core of chert. The surface, which is the usual greyish white,
may have originally been smooth, but is now badly weathered, particularly
on the bottom. This specimen is fully conical only on its top side. The
bottom side, which is markedly convex, meets the upper at a blunt edge en-
circling the specimen. This edge is not grooved, nor are there holes in
the top and bottom surfaces. The object comes from the shell refuse.

Stone Spike (Pl. 2,J)
The final stone object is made of a whitish tuff and has a smooth un-
polished surface. It is shaped somewhat like a spike, having a circular
cross section, a pointed lower end, and a neck and rounded butt at the top.
A number of small chips are missing from the point as if the object had been
used as a pick, but it seems more likely that it was some sort of ceremonial
object, like the so-called dagolitos of Cuba (e.g., Osgood 1942:32). It
comes from the ash layer.


Shell Artifacts

Shell Vessel (Pl. 2,1)
According to the Museum of the American Indian catalogue, this is a
"shell with columella broken out for use as a cup." A large conch (Strombus
gigas) has been prepared for that use by removing the meat and knocking out
the columella and parts of the wall just to the left of the outer opening.
The conical top of the shell, the pointed base, and three-quarters of the
outer wall joining the two are still present; together they form a relatively
large and deep receptable, comparable to those in the preceramic sites of
Cuba (Tabio and Rey 1966,Pl. 1,6). There is no evidence that any cutting was
done on this specimen; all the missing pieces seem to have been broken off.
A crack in the bottom of the receptable leads to a small hole in its deepest
part. Both crack and hole appear to be more recent than the other breaks,
and perhaps are the reason why the artifact was discarded. It comes from
the shell refuse.

Shell Columellae (PI. 2,G)
The collection contains two columellae from conch shells of the genus
Strombus. Both seem to have been cut away from the rest of the shell, perhaps
in the process of forming a vessel like the one described above. The columella
illustrated has since been worn smooth, either by nature or by human agency
(e.g., by use as a pendant). The two specimens are from the shell refuse.

Shell Plate
There are two other large fragments of Stombus shell. One seems to be
purely accidental, but the other may be the result of manufacture. The shell
has been broken in such a manner as to leave its outer lip, a large plate
composed of the portion of the shell behind the lip, and the conical top of
the shell. There is no evidence of use. The specimen approximates in shape
the shell plates described from other areas (e.g., Osgood 1942,PI. 3,H,I).
It comes from the shell refuse, as was to be expected.

Other Shells
Olsen collected ten other specimens from the shell refuse. All appear
to be natural shells, discarded by the aborigines as by-products of their
food quest. They include clam, conch, oyster, and scallop, presumably of
the same species identified by Clark Moore (1982, Table 1) in his collection.
One of the conch shells has the end broken off and another has a hole in
the end, which may have been made to extract the meat.


We can be more definite in drawing conclusions about the Olsen collection
now that we know that it comes from a single site. We may assume that there
was only a single component--that is, occupation by a single cultural group
or people--because there are no significant differences between the artifacts
found in the ash layer and the shell refuse. As Moore suggests (1982), the
ash layer may be a living floor and the shell refuse, the result of shell-
gathering activity away from the people's residences.

When I wrote the original version of this paper, it was customary to
apply the term "Ciboney" to the people who produced preceramic remains like
those found by Olsen. Now the term "Archaic" is preferred. Moore (1982)



has well stated the case for attributing Olsen's and his finds at Cacoq 2
to the Archaic people and culture.

While analyzing the assemblages Rainey and I obtained from the Couri
group of sites in northern Haiti (Rouse 1941:157-164), I compared them with
the Olsen collection from the south. I had not yet made a detailed study
of Olsen's finds, and as a result committed several errors, which I corrected
in subsequently writing the original version of the present paper. I did
not originally realize that the Olsen collection contains a flint blade and
a rectangular hammer-grinder like those in the Couri sites, nor did I recog-
nize that the plain and biconical disks in the Olsen collection were probably
ceremonial stones and hence different from the Couri hammer-grinders. These
two errors cancel each other and leave unchanged my original conclusion that
the Olsen collection belongs to the "cultural phase" that has become known
as Couri.

The presence in the Olsen collection of a flint blade is especially
significant because such blades are diagnostic of the Couri phase or complex
(Rouse 1941 Pls. 1-3). Unfortunately, Moore and I did not obtain any further
examples from Cacoq 2, but he did obtain a typical prismatic core from Cacoq 1
and his surface collection from the nearby site of Pradel contains a good
set of the blades struck off such a core, including stemmed spearheads and
backed knives like the ones from the Couri sites (Moore 1982). These finds
strengthen the attribution of Olsen's collection to the Couri phase, pro-
vided that we assume all three of the finds belong to the same cultural group.

The attribution is also supported by similarities in the style of shape
and decoration, as opposed to the typology of the collections. In Figure 2,
I have reproduced my previously published drawings of the stylistic modes
from Couri so that they may be compared with the modes from Cacoq 2 illus-
trated by Moore (1982 Fig. 5) and by me (Pls. 1, 2 and Fig. 1 of this paper).
It will be seen that a shell pendant from the Couri group of sites bears
the same ear motif as that on the single-bitted stone axes and the stone club
from Cacoq 2. (It is also on axes from Anse Milieux and Cocoq 1 that are
not illustrated; see Moore 1982.) In addition, the stone and shellwork
from the Couri sites bear the same kind of rectilinear incision as that on
the stonework from Cacoq 2. (It is also on a bowl from Pradel that is not
illustrated; see Moore 1982.) The co-occurrence of triangular areas of
alternatingly oblique hatched lines, bordered by horizontal lines, on stone
bowls from northern and southern Haiti and on the shell pendant from the
north is particularly striking. So also is the common use of a sharp cutting
tool, possibly a flint knife as opposed to a blunt graver.

In the original version of this paper, I contrasted the Couri assem-
blages, including the Olsen collection, with those made by Jacques Roumain
(1943) and his colleagues of the Haitian Bureau d'Ethnologie at Cabaret,
halfway between the Couri group of sites in the north and Ile a Vache in
the south. The Cabaret sites have yielded an extensive industry of flint
blades in an apparently preceramic context but, so far as the writer is
aware, no ground stone artifacts like those recovered from the Couri and
Ile a Vache sites (personal communication from R6my Bastien and Kurt Fisher).

It is still not clear whether the Cabaret sites were places of habita-
tion, like those at Couri and on Ile a Vache, or workshops where both pre-
ceramic and ceramic Indians may have chipped their tools and where one would


Figure 2. Decorative Designs from the Couri Group of
Sites, Northern Haiti: A, Stone Bead;
B, Stone Vessel; C, Shell Pendant (after
Rouse 1941, Figs. 1-3).


not expect the artisans to have deposited their ground-stone or clay vessels,
if they had them. Nevertheless, Cruxent and I (1969) hypothesized a Cabaret
complex and considered it to be earlier than Couri because it lacks ground
stonework. We did so by analogy to the situation in Cuba, where a Guayabo
Blanco complex without ground stonework, precedes Cayo Redondo and Mayari
complexes, which have milling stones and mortars, grinders and pestles, and
ceremonial artifacts like those in the Couri complex (Tabio and Rey 1966
Pls. 1-3). The radiocarbon dates for the Guayabo Blanco complex range from
2050-120 B.C. and for Cayo Redondo, from 330-1300 A.D. (Rouse and Allaire MS
Table 4). No radiocarbon dates have been obtained for the Cabaret and Couri
sites, but the two may have had similar ranges.

There is a parallel sequence of chipped-stone complexes in the Dominican
Republic, on the other side of Haiti. There, a Mordan-Barrera complex, com-
parable to Cabaret, is radiocarbon dated between 2610 and 2190 B.C. and El
Porvenir and a number of variant complexes comparable to Couri are radio-
carbon dated around 1000 B.C. (Pina et al. 1974:17; Rouse and Allaire MS,
Table 4).

Most recently Kozlowski (1974:37-69) has distinguished a Seboruco-Mordan
culture at the sites of Seboruco and Levisa 1 in eastern Cuba. As its name
implies, this culture is related to the earlier complexes in Haiti and the
Dominican Republic. It has even older radiocarbon dates, ranging from
3910-1510 B.C. (Rouse and Allaire Ms,Table 4).

Finally, MacNeish, Wilkerson, and Nelken-Turner (1980: Table 20) have
just worked out a sequence of flint complexes reminiscent of those under
discussion here, in the country of Belize, Central America. These are a
probable source for the West Indian complexes, but several discrepancies
between the Belizean and West Indian sequences should be noted. The Belize
complexes have bifacially worked projectile points which are missing from
the Cuban, Haitian, and Dominican complexes; and ground stonework is much
earlier on the mainland than on the islands. Milling stones and grinders
first appear on the mainland in the Sand Hill complex, dated between 7500
and 5500 B.C., while conical pestles and stone bowls are present in the
subsequent Belize complex, dating from 5500 to 4200 B.C. Additional research
is needed to resolve these discrepancies, especially on the island of Jamaica,
which provides an alternate route to that through Cuba for movement from
Central America into the Antilles but which has not yet yielded any Archaic
(preceramic) complexes. The question whether there was an Archaic occupation
of the Bahamas also needs to be investigated.

The decorative style of the Couri complex has not to my knowledge been
found in the other complexes we have been discussing. However, the Couri
complex's single-bitted, eared axes are widespread throughout the Antilles,
mostly as isolated surface finds. In the Lesser Antilles, they are known
as "Carib stones" on the obviously false assumption that they were made by
the Indians who occupied those islands in the time of Columbus. In Puerto
Rico, examples have been found in sites of the Arawak (ceramic) Indians
dating from Periods II and III, when those Indians were moving up from South
America and pushing back the previous Archaic occupants of the Antilles
(Rainey 1940:73; Rouse 1952:387,399,426,431). One wonders whether the
Puerto Rico axes may not have been trade objects from Archaic Indians of
the Couri complex in Haiti, whom the invading Arawaks had not yet displaced.


Explanation of Plate 1

A. Flint blade. 9 cm long, 2 cm thick. Cat. no. 16/6356.

B. Flint flake with secondary chipping. 8.2 cm long, 3 cm
thick. Cat. no. 18/6332.

C. Flake with traces of use. 7 cm long. Cat. no. 18/6356.

D. Single-bitted ax. 10.6 cm long, 3 cm thick. Cat. no. 18/6325.

E. Double-bitted ax. 19 cm long, 2.5 cm thick. Cat. no. 18/6355.

F. Single-bitted, eared ax. 12 cm long, 2.5 cm thick. Cat. no.

G. Ears from a single-bitted ax. 12.8 cm long, 3 cm thick. Cat.
no. 18/6326.

H. Bipointed smoothing stone. 18.2 cm long, 3 cm wide, 2 cm thick.
Cat. no. 18/6337.

I. Single-pointed smoothing stone. 9 cm long. Cat. no. 18/6329.

J. Rectangular hammer-grinder. 10 cm long, 7 cm thick. Cat. no.

K. Discoidal hammer-grinder. 7.5 cm in diameter, 7 cm thick.
Cat. no. 18/6339.

L. Oval milling stone. 23 cm long, 17 cm wide, 5 cm high. Cat.
no. 18/6330.

[Rouse] Plate 1

Florida Anthropologist 35(4)


t* '

*-.- .w
'^'2 .

] '

Flint and Stone Artifacts from Cacoq 2, Ile A Vache, Haiti





Explanation of Plate 2

A. Stone pestle or ceremonial artifact. 15 cm long, 5.2 cm in
diameter. Cat. no. 18/6328. (See also Fig. 1,A.)

B. Stone pestle. 11 cm long, 7 cm maximum diameter. Cat no.

C. Stone ball. 4.5 cm diameter. Cat. no. 18/6359.

D. Fragment of a mortar. 11.2 cm in diameter, 4.3 cm high.
Cat no. 18/6333.

E. Fragment of a stone vessel. 5 cm high; 5 mm thick on the
bottom, 10 mm through the underbody, 2 mm at the rim. Cat.
no. 18/6334. (See also Fig. 1,B.)

F. Biconical stone disk. 10 cm in diameter, 3.5 cm in maximum
thickness. Cat. no. 18/6322.

G. Shell columella. 3 cm long. Cat. no. 18/6345.

H. Plain stone disk. 10 cm in diameter, 5 cm thick. Cat. no.

I. Shell vessel. 10.5 cm long, maximum diameter 10 cm. Cat.
no. 18/6344.

J. Stone spike. 19.5 cm long, maximum diameter 3 cm. Cat.
no. 18/6354.


[Rouse] Plate 2

Florida Anthropologist 35(4)

Stone and Shell Artifacts from Cacoq 2, Ile A Vache, Haiti



I am indebted to the late Mr. E. K. Burnett, then assistant to the
Director of the Museum of the American Indian, for providing facilities
for studying the specimens, and to Dr. Harry Makami for identifying the
rock materials. The photographs of the Olsen specimens were made by Carmelo
Guadagno, of the Museum of the American Indian staff, and the drawings, by
David Kiphuth, of the Yale Peabody Museum staff.

References Cited

Cruxent, Jos6 M. and Irving Rouse
1969 Early Man in the West Indies. Scientific American




Moore, Cla

Janusz K.
Preceramic Cultures of the Caribbean. Zeszyty Naukowe
Uniwersytetu Jagiellonskiego No. 386, Prace Archeologiczne
No. 20, Warsaw and Krakow.

Richard Stockton, S. Jeffrey, K. Wilkerson and Antoinette
First Annual Report of the Belize Archaic Archaeological
Reconnaissance. Andover, Mass.: Robert S. Peabody Founda-
tion for Archaeology, Phillips Academy.

Investigation of the Prehistory of the Ile a Vache, Haiti.
In Proceedings of the Third Bahamas Conference on Archeology.
The Florida Anthropologist 35(4):186-199.

Osgood, Cornelius
1942 The Ciboney Culture of Cayo Redondo, Cuba. Yale University
Publications in Anthropology No. 25, New Haven.

Pina, Plinio, Marcio Veloz Maggiolo and Manuel Garcia Arevalo
1974 Esquema para una revision de nomenclaturas arqueol6gicas del
poblamiento preceramico en las Antillas. Santo Domingo:
Fundacion Garcia-Arevalo, Inc.

Rainey, Froelich G.
1940 Porto Rican Archaeology. New York Academy of Sciences,
Scientific Survey of Porto Rico and the Virgin Islands
18(1), New York.



Excavations in the Ft. Liberty Region, Haiti. Yale University
Publications in Anthropology No. 23, New Haven.

Diary of an Archeological Reconnaissance of Hispaniola and
the Bahamas. Manuscript in the Yale Peabody Museum, New

Roumain, Jacques
1943 L'outillage lithique des Ciboney d'Haiti. Bulletin du
Bureau d'Ethnologie de la R6publique d'Haiti No. 2,
Port-au-Prince, pp. 22-7.



Rouse, Irving
1941 Culture of the Ft. Liberty Region, Haiti. Yale University
Publications in Anthropology No. 24, New Haven.

1947 Ciboney Artifacts from Ile a Vache, Haiti. Bulletin du
Bureau d'Ethnologie de la R6publique d'Haiti, Serie II,
No. 2, pp. 16-21, No. 3, pp. 62-66, Port-au-Prince.

1952 Porto Rican Prehistory. New York Academy of Sciences,
Scientific Survey of Porto Rico and the Virgin Islands

Rouse, Irving and Louis Allaire
1978 Caribbean. In Taylor, R. E. and Meighan, Clement W., eds.,
Chronologies in New World Archaeology, pp. 431-481. New
York: Academic Press.

MS Eastern Venezuela, Guianas, and the West Indies. Manuscript
prepared for the second edition of Chronologies in New World
Archaeology, edited by Clement W. Meighan, New Haven.

Tabio, Ernesto and Estrella Rey
1966 Prehistoria de Cuba. La Habana: Academia de Ciencias de

Veloz Maggiolo, Marcio, Juan Gonzalez Col6n, Edgar J. Maiz, y E. Questell
1975 Cayo Cofresi: un sitio preceramico de Puerto Rico. Sociedad
Guayania de Arqueologia e Historia, Publication No. 1. Ponce.

Irving Rouse
Department of Anthropology
Yale University
New Haven, Connecticut 06520



Clark Moore

For the study of the Indians of the Caribbean, especially the preceramic
groups, the southwestern peninsula of Haiti has great importance. Bartolome
de Las Casas, the Spanish historian and defender of the Indians, differs with
a contemporary historian, Oviedo, in his opinion of the Indians of this region.

Oviedo says that the Indians of Hanyguaba were savages who
lived in caves, but this is not the case. Those Indians lived
a well organized communal existence suited to the amenity of
the land, which being like a garden, does not lend itself to
savagery. There were no caves either, nor spela as he calls them
to show off his knowledge of nominatives, but rather generous
fields and orchards with villages and cultivated land. I often
ate the natural products there (translation by Collard 1917:101).

Archeological evidence supports Las Casas' description of organized vil-
lagds. Three kilometers to the north of the coastal city of Les Cayes on top
of a low hill, a large midden encircling the entire hill is filled with pot-
sherds and shells.

Moreau de Saint Mery, the French historian, wrote a detailed description
of Saint Domingue during the 1790s. In the section dealing with what is now
the southwestern peninsula of Haiti, he told of Indian ceramic artifacts being
found on the Cays Plain, on plantations in the area of Port Salut, and near
the French town of Jeremie. He also mentions a large shell mound near the
town of Les Cayes (Moreau de Saint Mery 1797:1275). This preceramic shell
mound located near the present cemetery of Les Cayes was visited by the writer
in 1980.

Early Spanish geographers Andres Morales and Alonso de Santa Cruz mention
the existence of a primitive group of people living in this part of the island
who did not practice agriculture (Sauer 1969:48). Archeological evidence shows
that both preceramic and ceramic groups did inhabit the region. Further re-
search will be necessary to verify the historical observation of both groups
being present at the coming of the Spanish.

Herbert Krieger,a curator with the U.S. National Museum, stated after
visiting Indian sites in Haiti in 1931:

Future investigation should bring out clearly the sequence in
Arawak culture and its northern and southern affiliations as also
its autonomous development as an island culture through study of
several known major historical sites. Future investigations however,
may never determine just what are the relations that once may have
existed between the shell-mound builders of the Haitian, Bahamian,
and Floridian coast (Krieger 1932).




Fifty years later, Krieger's prediction of being able to understand the
sequences of the pottery-making Arawak Indians to a large extent has been
fulfilled. The problem of determining the origins and development of the pre-
ceramic shell mound builders still remains to be solved.

Ile a Vache

On the island of Ile A Vache there is evidence of habitation by an ad-
vanced group of preceramic Indians. In the 1930s Krieger, Godfrey Olsen, and
Froelich Rainey made observations and excavations in one or more large shell
mounds on the island. They located several ground stone artifacts, including
some with elaborate designs engraved on them.

Ile a Vache is situated in the Baie des Cayes 9 km south of the coastal
city of Les Cayes and approximately the same distance from Pointe Abacoa on
the peninsula to the west (Fig. 1). The island is 14 km long in an east-west
direction and an average of 3.5 km wide. The island covers approximately
45 sq. km.

The Indian name for the island was "Iabaque" and was shown as such on
several early maps of the New World. With the coming of the French the island
became known as Ile a Vache. Bernardo Vega in a recent publication in the
Dominican Republic (1980) says the French modified the Indian word "Iabaque"
into the word Ile a Vache.

Geologically the island has three basic formations. The low-lying eastern
half of the island is made up of recently uplifted coral. A zone of basalt
lies in the southwest part of the island and an older formation of well-bedded
limestone occurs in its northwest part (Street 1960).

The least desirable location for habitation is the eastern half of the
island. Its hard, sharp coral and thick mangrove make it difficult to pene-
trate. The only inhabitants of this area are along the shore and obtain their
main food supply from the nearby reefs.

The southwestern basalt area is covered with grass, although some areas
have been planted in sweet potatoes and millet. In this part of the island
cattle raising is important as it was during colonial times.

The most favorable area for habitation is the northwest section, where
a layer of soil has developed from the disintegration of Eocene limestone.
Crops of corn as well as sweet potatoes and beans do well on the numerous
hillsides. The best water sources are also located in the northwest section.
Although there are now no permanent streams or springs on the island, the
water is taken from shallow wells where it stands only a meter or two below
the surface.

All three areas of the island have been searched by me for sites. In
the northwest section, the only area where they were found, thirteen sites
were located. Of these, two small sites were classified as ceramic.

Cacoq Area

The greatest concentration of nonceramic sites are located at Cacoq
(Fig. 2). Located on a bay with large mangrove swamps further inland, this


( f HAITI r' I E

SL Anse\
Milieu II
Cacoq LesCays C
4 t IIe a Vache

Anse 2


4 Non-Ceramic Sites Bernard #
O Ceramic Sites 4


G ros d
1000 500 0 IKm




area contains two large shell mounds and a smaller mound as well as shell
scatters. These shell scatters contain heavy concentrations of shell with
enough of an admixture of rich midden soil to allow present-day cultivation
by Haitian farmers.

There appear to have been dramatic physical changes in the Cacoq area.
Because of the heavy cultivation of steep slopes large amounts of soil have
been eroded and redeposited into the mangrove swamps creating low mud flats.
It seems reasonable to assume that in the past this area of land-locked man-
groves had better access to the open water of the bay.

Moreau de Saint Mery refers to the area as having a good supply of fresh
water. During colonial times, when the French were not using this area ex-
tensively, there was a plan to destroy this water supply to prevent enemy
ships from using it (Moreau 1797:1325). Today there is no potable water at
Cacoq and the nearest source of water is over a ridge to the west at

Cacoq Sites

Cacoq 1
This mound is located approximately 700 m south of the present shoreline
of the bay (Fig. 2). The base of the mound lies in the mangrove swamp and is
barely connected at its west end to the bottom of a ridge that forms part of
the hill that separates Cacoq from Anse-a-1'Eau to the west. The mound is
50 m long east and west, 30 m wide, and just under 4 m high. From surface
observations the mound appears to be made up of solid shell. A surface find
of a flint core was made here by the author, and two stone axes were purchased
by the author from a person who claimed they were found here.

Cacoq 2
This shell mound is within 50 m of the west shore of the bay. The length
of the mound is 105 m and parallels the shore. Its width is 30 m and the
highest elevation on the mound is 3 m. It is not as symmetrical as Cacoq 1
(Fig. 3).

Cacog 3
This low mound is only 12 m to the west of Cacoq 2. It is 25 m long in
an east-west direction and 11 m wide. This mound is less than 0.5 m high,
and is presently being cultivated.

Cacog 4
Located to the east of Cacoq 1, this shell scatter lies on a ridge that
drops into the mangrove swamp. The site measures 67 m along the ridge and is
27 m wide. Surface collections here have yielded flint chips, round stone
grinders, and two mortars.

Excavations at Cacoq 2

Cacoq 2, which parallels the west shore of the bay, has historic value
because of previous excavations there. It may have been the same mound ex-
cavated by Herbert Krieger in 1931. He reported:


0 500m

- Mangrove






~- --------- -Manrov--- ---------

S **




r: I


There were occasional utensils and implements of stone and
of shell also much ash and scorched shell. Sand was present but
the mound was mostly shell.

Within the shell mound which was trenched were chipped stone
knives of flint, daggers with handles, also other miscellaneous
implements, types of a culture identical with the non pottery
yielding culture on the Pettigrew estate (Krieger 1944).

The artifacts from his excavation are presently at the United States National
Museum in Washington, D.C.

Godfrey Olsen conducted excavations at the site in 1933. His finds are
at the Museum of the American Indian in New York. No information about his
excavation is on file there except for catalog entries with the label "mound"
or "midden".

Dr. Froelich G. Rainey visited the site in January of 1934 and identified
it as the one dug by Olsen. He wrote the following:

The shell mound was composed of clear shell unlike the
usual midden. Clear shell heaps also reported at Sanoa, Dominican
Republic by De Booy. The mound was 6-8 feet high. About four
feet from the bottom there was a 10 inch level of black culture
deposit, which was well defined and regular throughout the exposed
cut. Pieces of flint were common in this culture level. Among
the shell there was no culture material (Rainey 1934).

Dr. Irving Rouse has published a description of the artifacts from the
Olsen excavation in a two-part series in the Bulletin of the Haitian Bureau
of Ethnology for 1947. He observed that some of the artifacts from the exca-
vation had designs that resembled work done on pottery by the ceramic Indians
of Hispaniola (Rouse 1947).

Because of the lack of definite information about the area, the site,
and the context in which the artifacts were found, Dr. Rouse suggested to the
author that he make a trip to the island in November of 1980. Through a
Haitian priest, Pere Margron, I located a man in his seventies who had helped
in the Olsen excavation. Senec Placide clearly remembered Olsen's name, the
location of the site, and that some of the artifacts had "ecrit" (writing)
on them. Neither he nor anyone else could recall the excavation of Krieger,
although it had been only two years earlier.

In January of 1981 I showed Dr. Rouse the site and he and I dug a one-
meter square test hole just to the north of the Olsen excavation. We decided
that the test hole needed to be expanded into a parallel trench 2 m wide in
order to obtain better information. In May of 1981 I returned to the island
and excavated an 8 by 2 m trench into the shell mound (Fig. 3). A third stage
excavation was initiated in January, 1982, so that in all, roughly 45 cu. m
of refuse were excavated, which represented approximately 2 percent of the
total mound.

Description of Artifacts and Excavation Results

The first 3 m of excavation into the test trench yielded large quantities
of flint chips, none of which could be recognized as tools. The first artifact

High Tide Mark H


Beach Sand

I % 1wuI. 1

Mangrove Swamp



to be found was a smooth limestone sphere, slightly elongated, 9 cm by 8 cm.
Its use has not been determined (Fig. 4).

The second significant find was a cache consisting of six stone artifacts.
A stone club of igneous origin had a cone-shaped handle and a hammer-shaped
head that was rounded at one end and slightly hooked at the longer end. Clus-
tered around the club were five limestone artifacts. Two were round elongated
stones with grooves circling the middle. There is a possibility they may have
been used as net sinkers. The cache was located at 7.8 m east from the begin-
ning of the trench, 35 cm south of the north baseline, and 1.1 m below the
surface (Fig. 5).

Eight meters into the mound, a faint trace of charcoal began to appear
60 cm below the surface. The dig was discontinued at this point and the trench
backfilled. With the permission of the Haitian government, the artifacts were
taken to the archeological laboratory at Yale University.

In January, 1982, the excavation was continued. A rounded, smooth lime-
stone grinder 6.5 cm by 4.5 cm was found. A broken piece of a flat milling
stone of volcanic origin was also located. It had a radius of 13 cm and was
2 cm thick at the outer edge, but toward Lhe center narrowed to only 5 cm.
The thinness of the center portion could be attributed to grinding that took
place on both sides. The thinness in the center portion produced severe weak-
ness that probably caused it to break. Around the other edge, which was
rounded, a shallow groove 1.5 cm from the edge was carved on both sides of
the specimen.

Half of a round thick milling stone was found 12.5 m from the beginning
of the trench, 1.8 m deep, and very close to the north baseline. This rough
stone measured 27 cm in diameter, was 5 cm thick on the outer edge and narrow-
ed to 2 cm in the center. Although both sides showed evidence of grinding,
one side had an indentation of 2.5 cm while the indentation on the other side
measured only 5 cm.

A light piece of red stone, 2.5 by 1 cm, was recovered which had stained
the light-colored shells around it. Other stones containing red material had
been found previously but this one was more concentrated and easily dyed the
material with which it came into contact.

The most prominent feature in the excavation was a dark layer of ash
mixed with shell. The layer became increasingly distinct beginning from a
point 10 m into the test trench and continued to the termination of the exca-
vation 5 m to the east (Fig. 4). Described from west to east, the layer
paralleled the surface contour for 2 m and then leveled off to a more gradual
slope than that of the surface. The layer was 20 cm thick and varied from
80 cm to 1 m below the surface. In this layer and especially on the surface
of the layer were large numbers of fish vertebrae, crab claws, and very small
mossy ark snells. Mixed in the ash were numerous flint chips.

Fifteen conch (Strombus gigas) generally associated with the upper layer
of the ash were found. They were in an area covering 10 sq. m and all except
two had the columella knocked out and showed signs of scorching on the outside.
The other two were in the process of having the inner section broken out but
work was terminated possibly because of the breaking of the outer surface of
the base of the shell. These two specimens showed no signs of scorching.


Scale: 1.5cm :1meter

2 3 4 5

7 8 9 10 11 12 13 14

Slightly Elongated Stone Ball

Artifact Chache

Section of Thin Milling Stone

Limestone Grinder

Section of Flat Milling Stone

Red Dye Material

Thin Soil Shell Layer

Sterile Sand

Ash Layer


Flint Chips



o =


Dr. Rouse in discussing the artifacts from Olsen's excavation describes
a shell dish as having been prepared by knocking out the columella and parts
of the wall to form a large vessel (Rouse 1947). Because of the burning on
the outside of the conch vessels found in my excavation, it may be possible
that they were used for cooking pots. This would have permitted the cooking
of the small mossy arks and crabs whose shells were found in the layer of ash.
It is possible that the columella was removed in order to obtain shell gouges,
although no gouges were found in the excavation.

Excavation Conclusions

A few general conclusions can be drawn from my excavation. The finding
of ground stone artifacts that show designs reinforces Dr. Rouse's conclusion
that the Olsen artifacts are from an advanced group of preceramic Indians.
(See Rouse's article in this issue.)

The layer of ash yielded evidence that the Indians who built this mound
of shells included other sea food in their diet. The fish vertebrae, small
mossy ark shells, and crab claws appear to have been prepared and consumed
on the mound.

Since the underlying shell mound continued above the ash layer, it may
be assumed that the same group that built the mound also deposited the layer
of ash. The profile of the excavation seems to indicate a fairly continuous
deposition of shell without long periods of abandonment. Besides the ash
layer the only other organic content is the soil mixed with shell at the sur-
face. The close correspondence of the ash layer to the surface contour in-
dicates that erosion of the surface of the mound has been minimal.

Other Preceramic Sites on Ile a Vache

There is another site at Anse-a-1'Eau to the west of Cacoq (Fig. 1).
This site consists of a heavy shell scatter running parallel to the edge of
a shallow brackish lake for 100 m. It is 26 m wide with a low mound in the
center. Numerous flint chips were found mixed with the shell material at the
surface. A stone mortar of volcanic origin was also found.

To the east of Cacoq near the bay of Anse Millieu a shell scatter was found
on a ridge at an elevation of 20 m. Pieces of flint were spotted on the sur-
face, and two round grinding stones that fit in the hand were found. A broken
eared ax was obtained at the bottom of the ridge. The site measured 60 m
along the top of the ridge and 32 m down the side.

A second smaller shell scatter in the area of Anse Millieu (Anse Millieu 2)
was located along the trail further east and south. This site, on relatively
flat ground at an elevation of 5 m had mangrove swLmps on two sides. Pieces
of flint were very common in the shell deposit.

Two kilometers further east at the bay of Soulet another small scatter
(Soulet 2) was found on a steep bank next to a mangrove swamp a half kilo-
meter south of the shore. Besides pieces of flint mixed with the shell, two
grinders were found. One was rectangular and of igneous origin, the other
was a crude round limestone grinder.


A second site in the area (Soulet 1) was found on the summit of a hill
at an elevation of 67 m. The hill is 1 km south of the bay. A shell scatter
encircled the entire peak. This was the highest and farthest from salt water
of any of the preceramic sites. Among the shells were found flint chips, a
round grinder, a very crude pestle, as well as another stone of igneous origin
that served as a crude mortar. A great deal of effort must have been expended
to get the artifacts and food up the steep slopes to the site.

At Madame Bernard on the north coast towards the center of the island,
another shell scatter was located on a ridge that had mangrove trees on three
sides. The site measured 33 m in an easterly direction along the ridge, 23 m
down the south slope, and 20 m down the north slope. The shells for the most
part were badly broken. Among them were flint chips, usually of a light brown

The final nonceramic site was at Pradel, 1 km to the southeast of Madame
Bernard. This site is on a low ridge that drops into a mangrove swamp. The
shell material is thinly scattered but large flint chips, some of which had
been worked, were common.

A well-shaped round mortar was purchased from the owner of the land. The
mortar had been broken in half when he struck it with a pick while preparing
the land for planting. It is circular, 22 cm in diameter, and 9 cm thick with
an indentation of 5 cm. Three grooves parallel to the rim,and intersected by
several faint vertical lines, formed a design on the outside surface.

A round pitted hammerstone 5 cm in diameter was found on the surface of
the site. The surface was hard and smooth with red coloring on one face. A
thin, round net sinker with three notches was also collected at the site.

Ceramic Sites

A small ceramic site was located on the top of Gros Morne which is inland
from Madame Bernard. With an elevation of 78 m, this is the highest peak on
the island. A few shells are widely scattered over the surface. One small
ceramic head-lug and two rim pieces as well as a few plain potsherds less
than a centimeter square were found.

The second ceramic site is a cave that was excavated near the west point
of the island in 1933 by Olsen. This cave is located in an outcropping of
hard sharp coral 4.5 m above the western edge of the bay of Anse Dufour. It
measures 5 m in length, 3.5 m from the floor to highest point, and 2 m in
width. It was damp and several large fruit bats were hanging from the ceiling.
A small amount of organic material was observed on the floor of the cave, none
of which were potsherds. Senec Placide helped in the excavation of this cave
as well as the shell heap. He was not present at the final stages of the work
but remembered seeing broken pieces of pottery in the shallow sea water below
the mouth of the cave after Olsen had left. The potsherds collected by Olsen
are cataloged at the Museum of the American Indian in New York.

Observations on Nonceramic Sites

In summary, all nonceramic sites except one are located very close to
ne source of their food supply. Shallow bays, the natural habitat for the



shell fish, are numerous in this part of the island. Dr. Rouse in his de-
scription of Ciboney (early name given the preceramic Indians of the Caribbean)
sites in Cuba describes a physical setting similar to Ile a Vache.

The environment of all the open sites except those at Manati is known,
and it is surprisingly uniform. All the sites are on an inlet of the sea
(or of one of the larger bays), in a favorable position for fishing. Each
is situated on top of a slight knoll, near a marsh and hence in an area in
which insects are numerous. In all cases except one, the marshes obscure the
view of the inlet. There is always a fresh water stream, but often it is
intermittent (Rouse 1942:131).

Ile a Vache is well supplied with raw material for manufacturing arti-
facts. There is an abundant supply of limestone with flint cores. The basalt
formations in the nearby southwest section probably provided the raw material
for making the ground stone artifacts since the nearest volcanic outcropping
on the mainland is several kilometers to the east.

The finding of the decorated stone mortar at the site at Pradel provides
a possible link between the shell scatters and the shell mounds at Cacoq.
Located over 5 km from the shell mounds, it indicates that the ground stone
artifacts with designs are not unique to the latter. The abundance of mortars,
pestles, milling stones, and grinders at the sites would suggest preparation
of food from plants. No plant remains have been recovered from the excavation
of any of the sites, however.

On the mainland a similar shell mound near Les Cayes had yielded surface
finds of milling stones and grinders as well as limestone chips of poor quality
when compared to the flint of Ile a Vache. It is not known what relationship,
if any, existed between the mounds on Ile a Vache and the one near Cayes.
There is an interesting similarity in the percentages of shell species in the
composition of the mounds shown in Table 1.

Table 1. Composition of Shell Mounds with Species
Given as Percentage of Total

Species Location

Les Cayes Anse-a-1'Eau Cacoq 2

Mossy ark 87% 90% 86%
(Arca imbricata)

Leafy jewel box 9% 6% 12%
(Chama macerophylla)

Small conch 3% 2% 0%
(Stombus gigas)

1% 2%






Scale: Approx. /3 Actual Size

Figure 5.. Artifact cache at Cacoq 2.





The nonceramic sites on Ile a Vache show no intrusion by ceramic groups.
The scarcity and smallness of the ceramic sites likewise indicate that the
island was not considered of much value by the agriculturally oriented ceramic
Indians. This isolation helps make it possible to identify nonceramic arti-
facts without facing the problem of their being deposited by later ceramic

References Cited

Collard, Andree
1971 Bartolome de Las Casas. History of the Indies. Editor
and translator. New York: Harper & Row.

Krieger, Herbert
1932 Culture Sequence in Haiti. Smithsonian Publication 3134.


Personal correspondence with Irving Rouse.

Moreau de Saint Mery
1797 La parties francais de l'Isle Saint Domingue. Vol. III.
Paris: Societe de L'Histoire des colonies franchises.

Rainey, Froelich
1934 Unpublished diary, pp. 39-41.

Rouse, Irving
1942 Archeology of the Maniabon Hills, Cuba. Yale University
Publications in Anthropology 25-26, New Haven.

1947 Ciboney Artifacts from Ile a Vache, Haiti. Bulletin Du
Bureau d'Ethnologie de la Republic d'Haiti, Serie II, No.
3:63-64. Port-au-Prince.

1982 The Olsen Collection from Ile a Vache, Haiti. The
Florida Anthropologist 35(4):169-185.

Sauer, Carl Ortwin
1969 The Early Spanish Main. Berkeley and Los Angeles: University
of California Press.

Street, John
1960 Historical and Economic Geography of the Southwestern Peninsula
of Haiti. Berkeley: University of California.

Vega, Bernardo
1980 Los cacicazgos de la Hispaniola. Santo Domingo: Museo Del
Hombre Dominicano.

Clark Moore
Hospital Le Bon Samaritan
Limbe, Haiti



Robert S. Carr and Sandra Riley

In 1958, thirteen-year old Alton Lowe crawled under his family's home
at New Plymouth Harbour, Green Turtle Cay to excavate a bucket of soil for his
parents' garden. Much to his delight, Alton uncovered a ceramic vessel formed
into an effigy of a bird on top of two gourds (Fig. 1). Alton thought of the
object as a toy and unsuccessfully tried to use the hollow bird's tail as a
whistle. Alton's father, Albert William Lowe, wisely retrieved the artifact
from his son and placed it with other miscellaneous items in the family tool
shed, where Alton rediscovered it in 1976, and then began to make inquiries
of museums and interested parties. Author Sandra Riley presented slides of
the artifact at the Third Bahamian Congress, and participants responded with
considerable interest.

The artifact is not typically Lucayan or Antillian, but rather more typical
of ceramic bottles of the Gulf area of southeastern North America. Con-
structed out of fine temperless clay, superficially similar to the paste of
the St. Johns series of Florida, the vessel is skillfully shaped into a double
gourd with an unidentifiable bird perched upon them. The bird and gourd are
hollow with the bird's tail open at the end to provide a spout. The vessel
has considerable surface wear, but some of the lower recessed areas and the
base retain a highly polished surface, not unlike that of finely made Weeden
Island pottery. The bottle is 12.9 cm high at its highest point and 13.5 cm
long at its longest point. The gourds are up to 7 cm in diameter. The form of
the bird and the gourds are naturalistic conforming to all the expected three-
dimensional curvature as they would be in life.

The most pressing questions regarding the artifact are where was it made,
and how, if imported, did it get to Green Turtle Cay? No convincing argument
can be made for the object's local manufacture since both its style and com-
position are totally dissimilar to Lucayan pottery types. Alton Lowe believes
that the artifact originated from a cave on Abaco and had been brought to his
house as part of a load of "bat-dirt"; the guano-rich sediment formed inside
of caves was commonly borrowed by islanders to provide nutrient-rich soil for
gardening. Since Bahamian caves often served as prehistoric mortuaries, arche-
ologists have long reported that both burials and intact artifacts have been
recovered from the caves (Granberry 1955); Alton's belief should be given
serious consideration. Although there are no prehistoric sites recorded with-
in the vicinity of New Plymouth Harbor, Granberry refers to three sites, all
caves, on all of Abaco (1955:131-133). Possibly, the artifact originated from
one of these caves or an unrecorded site.

The vessel most resembles a combination of the ceramic forms listed by
.vi 'ey in his classification of Florida Gulf Coast vessel forms (1949:496-506).
The gourd and bird are typical effigy representations in Florida, and it is
most likely that this artifact may have originated there. However, to convinc-
ingly demonstrate this origin would probably involve the need to do a compara-
tive chemical or spectigraphic analysis.




As to how the artifact may have found its way to Abaco involves only two
likely possibilities. The first being that the artifact was brought as a curio
by white settlers, possibly in the eighteenth century, by Loyalists who fled
the American colonies after the Revolution or from Florida after Spanish re-
possession in 1783. Indeed, some of the Abacoans trace their ancestors to the
American Southeast. Personally, we reject this explanation. We believe the
discovery of this artifact in redeposited cave "bat-dirt" is an association
far more likely to be related to prehistoric activity, considering the known
mortuary usage of caves by the Lucayan Indians.

Discussions of contact between Florida and the Caribbean islands have
involved many participants (e.g., Bullen 1974; Granberry 1955), and most agree
that aside from certain trait similarities between the two areas, specific
evidence of actual artifacts or resources that were manufactured in one area
and then transported to another has been generally lacking. This difficulty
is compounded when one recognizes that for the Bahamas and Florida, there are
few durable natural resources that are not shared between the two areas. The
chances of an archeologist in either southern Florida or in the Bahamas recog-
nizing that many shells, shell tools, or coral tools are nonlocal in origin seer
very doubtful. Determining the origin of chipped or polished lithic tools
could involve several geographic points of transportation aside from their
point of origin because they are not indigenous to either the Bahamas or
southern Florida. The most likely candidates to demonstrate contact between
the two areas would be from the recovery of either certain faunal remains
(e.g., iguana or hutia from the Bahamas and deer from Florida) and the very
distinctive pottery types from either area.

Until this discovery there have been no prehistoric artifacts or material
reported from the Bahamas that can be shown to have had their origins from
the North American mainland. Despite this paucity of physical evidence, we
can probably be assured that contact between Florida and the Bahamas and Cuba
did occur. Fontenada describes that "Anciently, many Indians from Cuba entered
the ports of the Province of Carlos (Southern Florida) in search of it River
Jordan; and the father of King Carlos, whose name was Senquene, stopped those
persons, and made a settlement of them, the descendants of whom remain to this
day" (Smith 1944:29).

The possibility that this effigy bottle was brought to the Bahamas from
Florida by prehistoric Indians seems likely; however, the proof will never be
more than circumstantial, which is typical of the fog that hangs over deter-
mining the origin of any redeposited artifacts, especially significant ones.
However, what is really the pertinent consideration to those detectives
searching for further physical evidence of contact, is what did this inter-
action mean in terms of its effect on Antillian and Florida Indian cultures.
We suspect that the Arawak exploration and expansion into the Bahamas, believed
to have occurred circa 1000 A.D.-1400 A.D. (Sears and Sullivan 1978), by no
means stopped short of Florida. In fact, we believed that the Arawaks prob-
ably reached the Florida Keys by ca. 1400 A.D., but unlike the unpopulated
Bahamian archipelago, encountered in Florida a thriving Indian population.
We can only guess at what these encounters were really like, whether peace-
ful or warlike; they were probably both. These contacts may have been the basi
for exchange of material goods, ideas, and even wives (or husbands for that
matter), but it is still a speculative matter, one that would require careful
hypothesis construction and testing, to determine what significant cultural
impact these contacts might have had.


We believe the ceramic bottle reported here resulted from contact between
the Lucayan Indians of the Bahamas and those of southern Florida. Although
the artifact is more typical of central or northern Florida, it may have been
in the possession of a high-ranking member of a South Florida tribe, and then
traded or given to visiting Lucayan Indians. Badly outnumbered, it is doubt-
ful that the Lucayans could have gained such an artifact by warfare or raids.
It may have been obtained during visits similar to those described by Fontenada.

Figure 1. A ceramic effigy bottle.

References Cited
Bullen, Ripley P.
1974 Were There Pre-Columbian Culture Contacts Between Florida and
the West Indies: The Archaeological Evidence. The Florida
Anthropologist 27:149-160.

Granberry, J.

A Survey of Bahamian Archaeology. MS, Masters Thesis,
Department of Anthropology, University of Florida, Gainesville.

Sears, William H. and Shaun Sullivan
1978 Bahamas Prehistory. American Antiquity 43:3-25.

Smith, Buckinc

Memoir of DO. d'Escalenta Fontaneda Respecting Florida Written
in Spain about the Year 1775. Translated from the Spanish with
Notes by Buckingham Smith (revision of 1854 Edition), Coral Gables.

Willey, Gordon R.
1949 Archeology of the Florida Gulf Coast. Smithsonian Miscellaneous
Collections, Vol. 113. Washington, D.C.

Sandra Riley
Coral Gables, Florida

Robert S. Carr
CCFL Bahamian Field Station
San Salvador



Ruth G. Durlacher-Wolper

The location of Columbus' initial New World landing has been a controver-
sial issue for years. I don't believe anyone can determine where Columbus
landed working from an armchair in Europe or by simply looking at carto-
graphers' maps, many of which were incorrectly drawn and copied. One must
actually follow in the discoverer's footsteps.

In 1959, the author attempted to duplicate the route preceding Columbus'
first landfall on the island of "Guanahani" (Wolper 1964). Armed with Samuel
Eliot Morison's translation of Columbus' journal, as well as additional infor-
mation from Morison, I used the measurement of the Greek league to determine
the length of mile that Columbus used when he referred to four miles to the

San Salvador is the most southeasternly of the Bahamas archipelago. It
is from 18-20 miles in length including Graham's Harbor with its perimeter of
connecting reefs at the north. White Cay at the northern end forms the top
of an inverted V, giving the island an appearance of being shaped like a
triangle. In fact, Las Casas called the island "triango". The island is long
and flat with scattered low rolling hills--the highest being Kerr Mount, which
is 140 feet above sea level. The island has over 20 lakes including salt ponds
in the interior. Great Lake, the largest, covers close to 12 miles in length,
averaging from 4-6 feet in depth.

The most likely location for the major Indian village on Guanahani is
the Pigeon Creek peninsula. It would have been easy for Indians from the
villages around Pigeon Creek to see Columbus and his ships as they "jogged
off-and-on waiting for daylight." It would have been natural for them to have
paddled to the mouth of the creek, curious to see at closer range those ships
which they thought had come from Heaven. Therefore, it seemed only logical
to this writer while attempting to recreate this leg of the Columbus voyage
to approach the island toward the mouth of the creek from High Cay.

Columbus wrote on Friday, 12 October, 1492, about his encounter with the

Later they came swimming to the ships' boats in which we were,
and brought us parrots and cotton thread in skeins and darts and many
other things, and we swapped them for other things that we gave to
them, such as little glass beads and hawks' bells. .. All that I
saw were young men, none of them more than 30 years old, very well
built, of very handsome bodies and very fine faces; the hair coarse,
almost like the hair of a horse's tail, and short, the hair they wear
over their eyebrows, except for a hank behind that they wear long
and never cut. Some of them paint themselves black (and they are of
the color of the Canary Islanders, neither black nor white), and others




paint their faces, others the body. Some the eyes only, others
only the nose. They bear no arms, nor know thereof; for I showed
them swords and they grasped them by the blade and cut themselves
through ignorance; they have no iron. Their darts are a kind of
rod without iron, and some have at the end a fish's tooth and
others, other things (Morison 1942).

Although parrots are no longer found on San Salvador, bone pendants
representing parrots have been found in an excavated Indian village. Cotton
(Gossypium hirsutum var. punctatum) grows wild on the island. Darts could
have been a spine or "whip", as the contemporary natives call the end of the
tail of a sting-ray. Possible dart points of stone and shell have been found
in archeological excavations.

On Saturday, 13 October, 1492, Columbus wrote:

They came to the ship in dugouts which are fashioned like a long
boat from the trunk of a tree, and all in one piece, and wonderfully
made (considering the country), and so big that in some came 40 or
50 men, and others smaller, down to some in which a single man came.
They row with a thing like a baker's peel and go wonderfully, and if
they capsize all begin to swim and right it and bail it out with
calabashes that they carry (Morison 1942).

Indian descendants on San Salvador relate stories about making dugouts by
burning the center and scraping it from the bark with shells or anything they
could find. Tree trunks, four and one-half feet in diameter, can be found in
the interior of the island. A sample of one of these trees can be found in
the courtyard of the New World Museum. Identified by Dr. David D. Keck, the
director of the New York Botanical Gardens in 1958, he wrote that "We are
pleased to have this sample of bark, which is Swietenia mahagoni (L.) Jaca,
the only mahogany bark now on file in our collections." The calabash to bale
out water was used when the author explored on the island in 1955 and 1956.
An example is in the New World Museum.

It has been suggested that some Indians of the Caribbean area may have
also used rafts for transportation (Rouse 1960). Not too long ago rafts were
commonly built and used on San Salvador. Stripping mohut for cordage, Indian
descendant Paul Ward, 80 years old, said that rafts were made from large trunks
of the gumelemi (Bursera simaruba), also called the West Indian Birch, and if
the tree is cut while green it is scooped out easily. Three or four tree trunk!
are fastened together with a cordage of mahot, sisal, or fiber from the fig
tree (Ficus carica), which makes a raft for fishing and is used with a long

On Sunday, 14 October 1492, Columbus wrote in his journal the following:

When day was breaking I ordered the ship's gig and the caravels'
bargers to be readied, and I went along the coast of the island to
the NNE, to see the other side, which was the eastern side, what there
was there, and also to see the villages; and soon I saw two or three,
and the people who all came to the beach, shouting and giving thanks
to God. Some brought us water, others, other things to eat. Others,
when they saw that I didn't care to go to shore, plunged into the sea
swimming, and came out, and we understood that they asked us if we



had come from the sky they raised their hands to the sky, and
they shouted to us to come ashore; but I was afraid to, from seeing
a great reef of rocks which surrounded the whole of this island, and
inside it was deep water and a harbor to hold all the ships in Christen-
dom, and the entrance of it very narrow .and I saw a piece of land
which is formed like an island, although it isn't one (and on it there
are six houses), which could in two days be made an island, although I
don't see that it would be necessary (Morison 1942).

This description of the land formed like an island, although it wasn't one,
aptly describes the narrow peninsula of Pigeon Creek.

In 1952, the author located an Indian village site at the head of Pigeon
Creek. At this site and from three other sites or site components nearby,
7,720 fragments of red clay pottery tempered with crushed shells (Palmetto
ware) were collected as we tried to keep ahead of the machinery which rapidly
cleared a road through a portion of the site. We noted that 585 fabric-
impressed, cassava griddle sherds weighed 263 pounds, whereas the remaining
fragments of thinner, smaller ware weighed only 5 pounds.

Many sites were discovered and named by the author which revealed new
significant data concerning the Indians who lived on San Salvador. Many of
these are located along Pigeon Creek, and include the Maize Field Site, the
Cactus Site, the Sisal Site, the High Ridge Site, and the Prickly Pear Site.
Test trenches, 7' by 3' dug in four levels at 3", 9", 15", and 21" down to
the white sand revealed a variety of artifacts. Most of the finds of decorated
rims, quartzite and shell beads, stone or shell pendants, a fish hook or a bone
needle or ear plug, were found at the 12" level. Despite the recovery of al-
most 15,000 potsherds, we were unable to find one whole vessel or even the
parts for one complete vessel.

Lithic tools were found in unit #13 at a depth of 21"-24". Over 400 stone
tools represented by flint knives, point sharpeners, hammerstones, grooved and
pitted stones, leaf-shaped projectile points and polishing stones were recover-
ed. Twelve small carved Indian heads of stone, known as zemis, also were

Shell gouges were plentiful at all sites. In the lower left corner of
unit #15, there was a clam shell with a point sharpened on the scraper edge.
This site also contained shell hammers and horns used for communication.

From units #10 and #13 two ceramic bat lugs and a pair of human arms and
hands on top of bowl rims were found. Also, lugs representing the iguana were
found. But the most prized possession of all is the anthropomorphic head that
I believe represents a chief found in unit #10, trench A, section 2 at the
12" level. This potsherd is 3" high and 2 7/8" wide (Fig. 1).

South Victoria Hill is at the head of Pigeon Creek and 157 potsherds were
surface collected here in 1958. Green polished stone celts were found at South
Victoria Hill as well as at Trial Farm, Fortune Hill, Cockburn Town Allen Place
and Farquahnson's Estate. It is believed that these celts originated from
Haiti. The natives call them "thunderbolts", and believe that they have fallen
from heaven when the thunder claps. After seven years they come up from the
ground to bring good luck to the finder.



Over the hill at South Victoria Hill is the Great Lake with an Indian
burial cave at its edge. This site, called Dancing Hall Cave, has an entrance
with two mouths. There are two main halls of rooms from which another 30 or
40 foot hallway branches. The entrance hall is 11'9" wide, 9'3" deep, and
4'4" high. Dr. T. Dale Stewart of the Smithsonian Institution identified a
human skull from this site (now reposing in the New World Museum) as typical
of the prehistoric Indians of the West Indies. This skull had artificial
flattening of the forehead. This cranial deformation provides physical evi-
dence of Columbus' description of the Indians he saw with wide flat foreheads.
In addition, cassava griddle fragments with plaited impressions, a sub-typed
shell cup, small carved shells and other artifacts were recovered from here
as well.

Both Dr. Frederick Dockstadter of the Museum of the American Indian and
Dr. Irving Rouse had advised me on how to excavate these sites. Finally, Dr.
Rouse said, "Stop! Invite Dr. John Goggin to San Salvador to teach you how
to dig scientifically, so you do not destroy history!" Dr. Goggin came in
1960, and subsequently, others have excavated here (i.e., Dr. Richard Rose,
Dr. Charles Hoffman, and John Winter).

Reconstructing Columbus' last day around Guanahani-San Salvador has been
done at least fifty times by the author. After Columbus went north-northeast
to see the other side of the island, he wrote that he saw other Indian villages.
In 1956, the Palmetto Grove Site was excavated by the author. The site was
later tested in 1960 by Dr. Goggin with the author, and later more intensively
excavated by Dr. Charles Hoffman (1970). Between 1956-1960 the author located
the following sites along the west side of the island: the Ward Site, Long
Bay, Sugar Loaf, Victoria Hill Cave, Cut Rock Cay, and the Man Head Cay sites.
My daughter, Beatrice Wolper, found the Harbour Yard Site in 1958, as well as
those at Trial Farm and Hanna Farm. Together we found artifacts in Cockburn
Town, Old Place and Fortune Hill. Red potsherds, made from clay mixed with
crushed shells, representing bowls, platters, and cassava griddles bearing
plaited and woven impressions, were found at all the sites. They were accom-
pained by bones of parrotfish, chiton, barracuda, whale, and turtle. Other
artifacts of stone, bone, and shell have been discovered with these. These
village sites probably would not have been found if Columbus had not described
where they were in his journal.


I would like to thank Charles Hoffman for originating the First Conference
of Bahamian Archaeology in 1966 at the New World Museum on San Salvador. I
would also like to thank John Winter of Molloy College for having organized
the second conference in 1979, and the third conference in 1981. My deep
appreciation to Dr. Donald T. Gerace for hosting this last conference at the
College Center of the Finger Lakes' Research Station on San Salvador.



Figure 1. Human craniums and
artifacts in the New
World Museum. Illustra-
tions by R. D. Wolper.

References Cited

Hoffman, Charles A., Jr.
1970 The Palmetto Grove Site on San Salvador, Bahamas. Contributions
of the Florida State Museum, Social Sciences No. 16.

Morison, Samuel Eliot
1942 Admiral of the Ocean Sea. Little, Brown & Co., Boston (2 vols).

1955- Personal communication.

Wolper, Ruth G. Durlacher
1964 A New Theory Identifying the Locale of Columbus' Light, Landfall,
and Landing. Smithsonian Miscellaneous Collections No. 4535,
41 pp.
Ruth G. Durlacher-Wolper
New World Museum
San Salvador, Bahamas



Stephen D. Glazier

This paper will discuss two Trinidad sites that have been regarded as
historic in nature but which also include considerable precontact material.
The first site, located in the Spanish capital St. Joseph, is primarily a
refuse site, while the second site, located in the village of Mayo, is
primarily a midden site. Both sites contain a mixture of European and Indian
artifacts. Taken together, these two sites represent the protohistoric
period, the entire Spanish occupation, and the early years of British rule
up to about 1830. St. Joseph (Trench 2 only) and Mayo provide an accurate
stratigraphic chronology; however, Trench 1 of the St. Joseph site is a
secondary deposit and had been thoroughly disturbed.

Excavations at St. Joseph and Mayo were part of a larger program of
historic archaeology undertaken during the summer of 1953 under the direction
of Professor Irving Rouse of Yale University (cf. Rouse 1953, for specific
sites covered in the 1953 program). The field party for both excavations
consisted of John Goggin, Rita Goggin, and Irving Rouse--although Rouse did
not participate in the later phases of the Mayo dig. The main goal of this
project was to explore the Spanish colonial period, but relatively little
Spanish material was found. In retrospect, this must have been a disappoint-
ment to John Goggin who was most interested in obtaining samples of majolica
and olive jars. Nevertheless, research was conducted to the most exacting
professional standards of the time and Goggin left detailed notes, now on
file at the Yale-Peabody Museum.

The town of St. Joseph was founded in 1592 as an administrative center
for the Spanish crown. It was situated inland, for defense reasons, on the
lands of the cacique Goanagoanare, who had been forced to move to another
part of the island. The above information is taken from Raleigh's 1595
account. Newson (1976:116) points out that it may have been a biased report
recorded to make the Spanish look bad. The town of St. Joseph was sacked
by Raleigh in 1595, then deserted for a time, and later rebuilt by Felipe
de Santiago. It is known to have been permanently occupied from 1597 to the

Excavations at St. Joseph were conducted in two trenches, the first in
a government school yard opposite the southwestern corner of the town savanna;
the second in the savanna itself near the corner of present-day Richmond and
Abercromby Streets. Trench 1, while not useful in establishing a chronology,
was found to be rich in Indian sherds and European materials including:
earthenware, delftware, chinaware, pipe stems, bowl fragments, nails, gun-
flints, a padlock (Sec Al 30-40 cm), a star-shaped insignia with the imprint
"Trinidad Rifle Brigade" (Sec A2 20-30 cm), a cache of rum bottle and gin
bottle fragments (Sec A3 20-30 cm), some majolica and olive jar sherds.
Trench 1, as Goggin (1968:41) inferred from some of the items listed above,
was probably the site of a 19th century British military barracks. Most



(i I"

G l, L F L

) VF

P ,*1 R~ I A

Bri3sh Mil4h
I234 in 21

Figure 1. Nineteenth century map of Trinidad.


sherds are small and there are few examples of each type; especially at the
upper levels. This could indicate that there was some form of rubbish removal
in St. Joseph by the later part of the 18th century.

Trench 2 did not contain as much European material as did Trench 1.
Goggin was able to identify some majolica and olive jar fragments, and I
have identified: earthenware, pipe stems, green glass, a metal blade, nails,
Indian sherds, and conch shells. Of particular interest is the high pro-
portion of northern European as opposed to Spanish materials.

The St. Joseph dig was begun on August llth and continued through August
18th. Each approximately 2 x 6 m trench was divided into three roughly equal
sections labeled Al, A2, and A3 respectively. Work was carried out at 10 cm
arbitrary levels. I am currently in the process of cataloguing the entire
collection for the Yale-Peabody Museum and will follow Goggin's system. The
collection should be fully accessioned by September of 1982.

As mentioned above, Trench 1 had been churned and disturbed. It was
also the richest of the sites. Mixed European and Indian artifacts were
found at every level to a depth of 50 cm. Rouse (1953:98) has classified
the Indian sherds as being of the Bontour-type and Goggin (1960) has classi-
fied the olive jar sherds as being of the "middle variety". Goggin also
identified several examples of 18th century Spanish glass. The remainder of
the glass is either 19th century (English?) or modern. Majolica tentatively
identified by Goggin include: Ichtucknee Blue on Blue (ca. 1550-1600), Santo
Domingo Blue on White (ca. 1550-1600), San Luis Blue on White (after 1650),
Aucilla Polychrome (after 1650), Puebla Polychrome Brown on White (ca. 1650-
1700), Puebla Blue on White (after 1700), and Columbia Plain. In all, 16
sherds of majolica were found in Trench 1. Goggin, correctly, decided that
his sample was too small to report on the majolica by levels (Goggin 1968:

Of the majolica discovered in Trench 1, only Columbia Plain and Ichtuck-
nee Blue on Blue are part of what Lister and Lister (1974:19ff.) have classi-
fied as the "Caribbean complex". Columbia Plain, discovered at the 20-30 cm
level, is a pottery type which probably had its origins in Sevilla and environs
and was dispersed around the Caribbean basin in the wake of Spanish explora-
tion. Ichtucknee Blue on Blue is a somewhat later type and was distributed
more widely than Columbia Plain, but in lesser volume.

It is not certain whether or not some of the above types may have been
fashioned in the New World. The Listers (1974:22) report that no evidence
of a majolica workshop has yet been found in the Caribbean. They do suggest,
however, that the most likely location for such an industry would have been

Other majolica types (Puebla Polychrome, Puebla Blue on White, and San
Luis Blue on White) are part of what the Listers have identified as the
"Mexican complex" and definitely were fashioned in the New World. Pottery of
these types was widely distributed in the 17th and 18th centuries. Goggin
(1968:156-157) reports that examples of these types have been found from the
Spanish Indies to Venezuela, along the Atlantic coast to Florida, in Georgia,
up the Rio Grande Valley of New Mexico, and into northern Arizona. The
widest distribution took place in the 18th century but this trade seems to
have bipassed Trinidad. Other Spanish islands were also bipassed; for example,
the Listers (1974:37) state that 18th century "Mexican complex" majolica is
scarce in Hispaniola.



According to Goggin (1968:162-163), Aucilla Polychrome was also manu-
factured in Mexico but the specific manufacturing center is unknown. He
was unable to date this type exactly, but did note that it occurs most fre-
quently with other mid-17th century types.

In the process of cataloguing, I have positively identified three 19th
century English styles: "Willow" Blue on White, probably produced by Spode
between 1810 and 1825 (William 1943; Hayden 1925:69), "Swansea" Blue on White,
and "Swansea" Green on White, patterns that were very popular in the early
19th century (Lewis 1969:115). "Swansea" was used for many purposes in the
British colonies; however, the "Willow" pattern is of special interest because
it is a much finer earthenware than might have been expected in a Trinidad
barracks. Perhaps it was for the officers' mess, or perhaps the "Willow"
was originally from another site entirely?

Only the upper levels of Trench 2 (0-30 cm) contained mixed European
and Indian artifacts. Goggin identified twelve majolica sherds within Trench
2, twenty-five percent less than what was found in Trench 1. Samples of
Ichtucknee Blue on Blue, San Luis Blue on White, and Santo Domingo Blue on
White were found at level 0-10 cm. San Luis Blue on White was found between
10-20 cm and Aucilla Polychrome was found at level 20-30 cm (see Table 1).
In addition, the top three levels contained: olive jars, green glass, pipe
stems, nails, samples of "Swansea" Blue on White and Green on White patterns,
samples of "Willow" pattern, metal blades, shells, and Bontour-style sherds.
At the lower levels, starting at 30-40 cm, many Bontour sherds and shells
were excavated. Section Al was excavated to a depth of 60 cm. Section A2
was excavated to a depth of 80 cm and Section A3 was excavated to a depth of
70 cm. Below 70 cm, Goggin encountered sterile soil.

Table 1. Frequency of Majolica at the St. Joseph site, Trench 2


0-10 cm

10-20 cm

20-30 cm

30-40 cm

40-50 cm


Blue on Blue 1 1

San Luis
Blue on White 1 1 2

Santo Domingo
Blue on White 1 1

Blue on White 5 1 6

Polychrome 1 1

Black on Emerald 1 1

Total 8 2 2 0 0 12



In all three sections of Trench 2, the greatest numbers of Indian sherds
were found at the 10-20 cm level. On the basis of the archaeological evidence,
it is apparent that the Indian population of St. Joseph was much larger in
the historic period than it had been in precontact times. This may, as Rouse
(1953:110) postulated, reflect the relatively great length of the historic
period of Indian occupation extending as it did well into the 19th century.
The ethnohistoric sources do not suggest an absolute increase in the island
population during this period, but do suggest that the population distribu-
tion changed dramatically as a result of Carib raids and the Spanish occupa-

Newson (1976:122) notes considerable fluxuation in the town's Spanish
population during the 17th, 18th, and 19th centuries, with epidemics period-
ically taking a very heavy toll. Still, the overall trend was one of steady
population growth. In the beginning of the 16th century there were but
35-40 European men in St. Joseph, but by the mid-17th century deLaet and
Pelleprat report a town population of approximately 200 persons.

During the 16th and 17th centuries, Spaniards, Indians and Blacks lived
in close proximity. Intermarriage was common and as time went on it became
increasingly difficult to distinguish "Indians" from the rest of the popula-
tion. The mestizo population of St. Joseph appears to have carried on many
aboriginal traditions, including the manufacture of pottery.

In both trenches at the St. Joseph site, tremendous variation was noted
within the Bontour style. Most salient among these were differences of
material, craftsmanship, and decoration. Later Bontour sherds in Trench 2
were much coarser than sherds found at the lower levels. I believe that a
third trench at this site would be of great value in establishing a sequence
for Bontour. As Rouse (1953:110) has noted, Bontour may have been unjustly
overlooked because researchers have been preoccupied with decorated pottery
at the expense of the relatively plain and drab material found in Bontour

The second site, Mayo, a town on the extreme western edge of the
Montserrat Hills, was excavated from September 7 to September 9, 1953, just
after the St. Joseph dig. The town is of lesser importance and more recent
origin than St. Joseph, and there are fewer references to Mayo in the litera-
ture. Borde (1876 11:192) mentioned that the Spanish maintained a mission
in the Montserrat Hills from 1667-1789, and Newson (1976:259), on the basis
of Crame's map of 1777 and Rodriguez's map of 1785, identifies this as the
iresent-day town of Mayo. The mission also appears as such on Thomson's
816 map of Trinidad (Fig. 1). Because an archaeological site was located
around a modern Roman Catholic church, Rouse (1953:100) suggested that Mayo
nay have been the site of the Spanish mission Nuestra Senora de Monserrate.
re ..as since modified his position. He now believes that there is little in
the archaeological evidence that would indicate conclusively that Mayo was a
Spanish mission. Remains were predominantly from northern Europe (British
and Dutch)and Amerindian. Very few 17th and 18th century materials were

Missions in Trinidad were small-scale operations. The mission at Mont-
serrat, for example, was founded and staffed by three priests, and it was one
of only two Trinidad missions founded between 1660 and 1706 that did not
eventually become a town (Borde 1876:42). Given the shortage of priests in



Trinidad, there would have been even fewer artifacts if Mayo were a mission

The Listers (1974:38) have concluded that mission sites seldom yield
much majolica. They contend that presidios, with their complements of troops
who ate off of tin-glazed tableware, have been much more productive in yield
of majolica artifacts than missions which usually had only a piece or two of
altar furnishings.

Three trenches were dug at Mayo. For reasons of expediency, the pro-
ject was carried out at arbitrary 20 cm levels. Trench 1 was divided into
three sections (A1, A2 and A3) and excavated to a depth of 50 cm; Trench 2
consisted of one section only and was excavated to a depth of 40 cm; and
Trench 3 was divided into two sections (A1 and A2) and excavated to a depth
of 50 cm.

All three trenches are useful in establishing a chronology. They in-
clude mixed European and Indian artifacts at the higher levels (above 20 cm)
and exclusively Amerindian artifacts below 20 cm. The only exception is Trl
Section A3, 20-40 cm, where one piece of green glass was located. While there
are as many pipe stems, gun-flints, horseshoes, pipe bowls, nails and door
hinges as were found at the St. Joseph site, there are fewer examples of
earthenware, chinaware, delft, and majolica. As in St. Joseph, the total
number of artifacts declines dramatically below 20 cm.

In conclusion, a major importance of the St. Joseph and Mayo collections
is that they support Rouse's assertion that the Bontour style was both wide-
spread and varied in Trinidad (Rouse 1953:97). Actually, these two "historic"
sites contain more Bontour sherds than the Bontour site itself. The majority
of Indian sherds at both sites are thin, soft, and as in other Trinidad sites,
not decorated. Some modeling was noted in the earlier Bontour material exca-
vated at St. Joseph. Most designs, as Rouse suggested, are either straight
parallel lines or dotted areas. There is little evidence of painting. It
is of great interest that the Bontour-style continued in Trinidad well into
the 19th century.

I had hoped that it would be possible to say more about "ethnicity" on
the basis of these collections (Glazier 1980; Figueredo and Glazier 1982),
but found that such information was not readily retrievable. Differences
within the Bontour style may provide the key. I shall continue to work in
this area. In addition, I hope to identify other examples of earthenware,
chinaware and delft.

I suspect that considerable 18th century material will be identified
here since the Spanish never were able to stop trade with the English and
the Dutch because the Spanish often did not provide Trinidad with needed
goods (Borde 1876). Newson (1976:141) describes trade with Spain as being
intermittent; for example, in the twenty years following settlement of the
island in 1592 no Spanish ship visited Trinidad at all. In 1687, Governor
Sebastian de Roteta complained that no registered Spanish ship had visited
Trinidad for four years, and in 1716, the Cabildo complained that no regis-
tered ship had arrived in Port of Spain since 1702. Meanwhile, the French,
English, and Dutch developed extensive illegal trade relations with the




This research was undertaken while I was a Research Fellow at Yale
University. I am especially grateful to Professor Irving Rouse for his
encouragement and many helpful suggestions. In addition, I want to thank
Charles Fairbanks, Barbara McCorkle, and John Winter for their assistance.

References Cited

Borde, Pierre-Gustave-Louis
1876 Histoire de l'Ile de la Trinidad sous le Gouvernement
Espagnol. Paris: Maisonneuve et Cie., Libraires-Editeurs.

Fairbanks, Charles
1973 The Cultural Significance of Spanish Ceramics. In Ceramics
in America. Charlottesville: University of Virginia Press,
pp. 141-174.

Figueredo, Alfredo and Stephen D. Glazier
1982 Spatial Behavior, Social Organization and Ethnicity in
the Prehistory of Trinidad. Journal de la Societe des
Americanistes 68.

Glazier, Stephen D.
1980 Aboriginal Trinidad in the 16th Century. The Florida
Anthropologist 33(3):152-159.

1960 The Spanish Olive Jar: An Introductory Study. Yale
University Publications in Anthropology 62, New Haven,

1968 Spanish Majolica in the New World: Types of the Sixteenth
to Eighteenth Centuries. Yale University Publications
in Anthropology 72, New Haven, Conn.

Hayden, Arthur
1925 Spode and His Successors. London: Cassell and Co., Ltd.

Lewis, Griselda
1969 A Collector's History of English Pottery. New York:
Viking Press.

Lister, Florence and Robert Lister
1974 Majolica in Colonial Spanish America. Historical Archaeology

Newson, Linda A.
1976 Aboriginal and Spanish Colonial Trinidad. New York:
Academic Press.



Pelliprat, Pierre
1965 Relato de las misiones de los Padres de la Compania de
Jesus en las Islas y en la Tierra Firme de Am6rica
Meridional. Biblioteca de la Academia Nacional de la
Historia no. 77, Caracas.

Rouse, Irving
1953 Indian Sites in Trinidad. In On the Excavation of a Shell
Mound at Palo Seco, Trinidad, B. W. I. by J. A. Bullbrook
pp. 94-111. Yale University Publications in Anthropology,
no. 50, New Haven, Conn.

Williams, Sydney B.
1943 Antique Blue and White Spode. London: B. T. Batsford Ltd.

Stephen D. Glazier
Department of Anthropology
University of Connecticut
Storrs, Connecticut 06268



Kathy Gerace

San Salvador Island, one of some 700 Bahamian Islands, lies at 240 N.
latitude and 74030' W. longitude, or nearly 370 miles southeast of Miami,
Florida (Fig. 1). Geographically, San Salvador is a low, carbonate island
covered with dense scrub brush and numerous saline lakes. The highest point
is Kerr Mount, with an elevation of 140 feet. Numerous reefs surround the
island, and there are good anchorages, depending on wind direction, in
Grahams Harbor at the north, French Bay at the south, Cockburn Town on the
west, and Pigeon Creek at the southeast (Fig. 2).

The first known European settlers of San Salvador were slave-owning
planters who were loyal to Great Britain and thus left the southern American
colonies during and after the American Revolution. When these Loyalists
were first granted lands on San Salvador, the island was covered with a virgin
forest of mahogany, mastic, lignum vitae, cedar and braziletto. These large
and valuable woods were cut in order to clear the land for planting and also
to export for immediate profits. The wood that was not exported was burned
and the resulting soil was rich, producing excellent cotton and food crops.
However, the cutting of all the trees removed not only the root systems which
prevented water erosion, but also lines of trees which acted as wind breaks.

Most of the land on San Salvador was granted during the eighteenth
century, though not all of it was settled by the owners. Although good
records are not available as to who settled here, based on the remnants of
the slave plantations, we can estimate a population by the end of the 1780s
of about 2,000 persons.on San Salvador.

Present-day ruins of these plantations are located at Grahams Harbor,
Sandy Point, Trial Farm, Old Place, Farquharson's, Kerr Mount, Fortune Hill,
Hard Bargain, Polly Hill and Dixion Hill (Fig. 2). There are several other
areas of old communities which are presently not inhabited, but which of
these date from the slave period is not fully known at this time.

During four one-month field periods from 1973-1976, undergraduate
students under the direction of the author investigated three separate
plantations--Sandy Point, Farquharson's, and Fortune Hill. Due to a lack
of space, any real detail as to the results of this work cannot be described.
Instead, generalizations of settlement patterns and way of life that has been
gleaned overall from this research will be discussed.

These plantation sites include a number of buildings lying in various
states of ruin. All of these structures use local stone in their construc-
tion, although the techniques used vary. It is presumed that wooden struc-
tures also existed, since lumber was readily available on the island. How-
ever, no evidence of these structures remains today. Due to the almost com-
plete erosion of the soil on most of the plantation sites, even post molds
have not been found.



Fig. 1



S50S 100 Nautical Miles
I' I 1 I I I I I I I I

Contours in fathoms


: BANK '
'' "*. 0 ...
'" 10 *.,


o *. *.
.* V *
*.... o0 EXUMA

I \
' *
> I '*! I








73 W

j I







Fig. 2
San Salvador Island
the Bahamas

(Based on research

by Sandra Riley)

A Plantation
SPoint oi interest

1 l 0 1 I
m~elo 1000 500 0 1000J~p 2000


a ... Af N. *



Although a few of the buildings are constructed of unaltered piled
stones with no bonding agent, the majority of the buildings from the planta-
tion period were constructed of stone mortared together. Some were covered
with plaster on the inside and sometimes on the outside as well. In several
of the buildings at Fortune Hill the stones had been cut prior to being
mortared together. The roofs were of either waddle and thatch or shingle.
The latter was probably reservered for the homes of the plantation owners,
while the slave homes and industrial buildings used for food storage, cotton
storage, and animal houses were probably thatched.

The other type of stone construction is represented by several examples
on the island but only one from these three plantations. This is the main
house at Fortune Hill, which is built of tabby. This technique consisted
of building wooden forms into which rubble and mortar were cast to make walls.
The wood was then removed and the walls plastered.

In all cases, the materials used in constructing these buildings were of
local origin, with the exception of the bricks that lined the baking ovens.
No soil suitable for brick making is available on San Salvador.

There is no evidence that glass was used in any of the windows of these
buildings. Instead, residential buildings and some industrial buildings had
wooden window cases that held wooden shutters. These were much like we see
in many homes on the island today. The shutters would have been left open for
light and ventilation by wooden props and closed when necessary for inclement
weather and insects.

The various buildings found on each plantation were designed to serve
their specific purpose. The plantation owner's home was usually the largest
and most complex building architectually. Of the three plantations investi-
gated, all of the main houses were two stories high, and those at Sandy Point
and Fortune Hill contained at least four main rooms. Farquharson's main house
is in such poor condition that information concerning the number of rooms could
not be gleaned. In all three cases the main house was on a rise overlooking
the entire plantation surrounding it. Not only did this location provide a
means for the owner to oversee his holdings, but it was the ideal position
for receiving the cool breezes that also kept the insects to a minimum.

Of the artifacts recovered by surface collecting within and around the
main houses of these plantations, most were isolated ceramic and glass bottle
sherds. The ceramics date from the late 18th century and early 19th century
and are of British origin. The bottles, when whole or of analyzable pieces,
range in date from the late 18th century to the early 20th century, and most
of them originally held beer, gin, or wine.

The artifacts from Sandy Point and Farquharson's were nearly the same
in quantity and quality in the area of the slave houses as those of the main
house. This comparison cannot be stated for Fortune Hill, as the slave
quarters there have not been studied beyond a count and general description
of the houses.

Near the main house was located the kitchen, best represented at Sandy
Point and Farquharson's since the one at Fortune Hill is in complete ruin. In
both of these cases the large chimneys of the fireplace with a brick oven be-
side it dominate the building. The buildings were rectangular and had two
rooms separated by a partition. One room contained the fireplace and the
other was possibly utilized for eating.



The slave houses were distinctive at each plantation. Those at Sandy
Point were single room buildings made of unbonded piled stone and were the
most primitive of the three plantations. Twelve of these houses were located
on the same ridge as the main house, but were located about half a mile away,
There seems to be no designed order to their location, with the exception
that they are situated on the ridge. Each individual building probably
housed one slave family. There are no fireplaces or remnants of any, so
cooking was probably done on outdoor fires, remains of which have long ago
eroded away. The only unusual artifact found in this area was a pipebowl
of a type described by Noel Hume (1972:302-303) as being made in North
Carolina between 1770 and 1840.

The slave houses at Farquharson's are located at a much lower elevation
than the main house area. They are in a row, along a fence line, which serves
as one wall of several of the houses. Most of the houses are of one room with
a large fireplace at one end, and probably held one family. There is a walled-
in yard area on one side of each house. Besides these houses, of which there
were ten, there is a larger building located along this row of slave houses
that contained three rooms. This may have been a barracks for housing un-
attached males, or was perhaps an overseer's residence.

The slave houses at Fortune Hill are the largest and best built of the
three plantations. These are located over a mile distance from the main house
area. They are along a ridge, but one at a much lower elevation than the
main house. These houses, of which there are at least sixteen, are located
in two rows bordering the side of what was probably a road. Except for one,
all of them are similar large rectangular buildings of cut stone that is
mortared together and covered with a red clay type of plaster. Each has a
large fireplace at one end. The one larger building is set apart from the
others and was perhaps the overseer's residence.

Another specialty building found on each plantation was the latrine.
This was located at a distance from the main house and was also architectually
unique in each case. At Sandy Point this building was a reduced version of
the main house, with gabled ends and high windows. Likewise, at Farquharson's,
the latrine had gabled ends and its windows were fitted with wooden bars like
those located in the windows of the main house of that plantation. The latrine
at Fortune Hill is architectually the most interesting, being octagon-shaped
and located over a deep natural sink hole. At ceiling level is located open
stone work and an outside ledge called, by some, a pigeon walk. This latrine
was utilized as a refuse deposit area and is one of the few areas from which
we have recovered large numbers of artifacts, as well as food remains in the
form of bones from cattle, pigs, goats, chicken, turkey and tern (Storrs
Olson, personal communication). The artifacts include many ceramic pieces
which have been reconstructed, as well as remains of crystal goblets, a
thimble, buttons, a cutlery handle of incised bone, sewing pins and two
Turlington's Balsum of Life medicine bottles.

The industrial buildings vary in style according to their intended use,
but all probably had thatched roofs. At Farquharson's and Fortune Hill are
located buildings that are believed to have been used as stalls for keeping
livestock, such as horses or cattle, since they contain long, narrow rooms.
Some of the industrial buildings had cement floors outside of the buildings
that were used for drying of foodstuffs, such as pigeon peas, corn, etc. In
the case of Sandy Point, this was surrounded by a low wall to keep out animals.
The buildings were used primarily as storage sheds for food and cotton.



Many, if not all, of the plantations are believed to have had a cotton
gin. This was located in what was called the ginning circle, since usually
it was operated by animal power. At Farquharson's, the ginning circle was a
built-up stone floor covered with concrete. There is no wall, but it is
believed that there was a thatch roof supported by wooden poles that have
since disappeared. At Fortune Hill, the ginning circle is located at the end
of one of the industrial buildings and consists of eight cut stone pillars
that supported a roof, probably thatched, but no other walls. There is no
evidence of a stone or concrete floor for this ginning circle. No evidence
of a ginning circle has been found at Sandy Point.

Note should be made of a unique building at Fortune Hill. This is a
small octagon building located in the front lawn area of the main house. It
is believed to have been the plantation office and/or library, since there is
evidence of wooden shelves along each of the walls between the windows.

Each plantation also contained at least one well. At Sandy Point and
Fortune Hill, the well is in a sink hole with stone steps leading down to
the water level. In both cases this is located some distance from the main
house area, but much closer to this area than to the slave quarters. The
location of the Farquharson plantation well had not been located during my
investigations, but in 1981 John Winter (personal communication) found a well
that is also within a sink hole near the slave quarters.

The remaining man-made features of these plantations are the extensive
walls. These are constructed of piled stone and were one of the major ways
in which the fields were cleared. These walls served several purposes, some
being property boundary walls,some enclosed small areas which probably held
animals, while others delineated various fields.

As can be seen, the architecture of these individual plantation buildings
reflects the purpose to which they were put and the social status of the
specific occupants. This is relevant since the artifactual record, as a
whole,has not reflected either building use or social status. What the arti-
facts do show is the time period of occupation and the material style of that
time, especially of ceramics. The architectural style of these structures,
such as the octagon buildings at Fortune Hill and the technique of stone
building and tabby building, reflect those used at that time in the Carolinas
and Georgia, the original home of the plantation owners.

The layout of the communities or plantations reveals the nature of the
slave cotton economy and the social organization on which it was based. The
slave houses were always located at a distance away from the main house area,
with the industrial buildings being near the plantation owner's home. The
owner's home was always located on the highest hill on his land overlooking
his entire plantation. The slave houses were always congregated in one area,
often organized in rows, with each house usually similar to all the others.

For the.Loyalists, the first few years after their move to the
Bahamas seemed an answer to their prayers. From the virgin soils that cover-
ed these islands,cotton grew profusely and for the first few years, the
owners reaped handsome profits. However, in 1788 and again in 1794 the
chenille bug destroyed much of the cotton in the fields. These insects, along
with intensive cultivation year after year and the change in ecological con-
ditions, such as drought possibly caused by the cutting of the forests, soon
led to more crop failures than successes. Thus, although there were some



good years for cotton, by the early 1800s, a number of the plantation owners
were leaving the islands. Some left overseers in charge of their lands, while
others sold out to other plantations.

The plantation owners who remained tried to diversify their crops, raise
livestock, or in other ways make a living, since few could afford to abandon
their estates. However, the winds of abolition were blowing in England and
soon were felt across the sea. In 1807 all slave trading was abolished, and
this was followed by registration acts and laws to protect the slaves. Then
in 1834, emancipation was declared for all of the British Empire, thus ending
a way of life for those Loyalist planters who remained in the Bahamas.

References Cited

Hume, Ivor Noel
1972 A Guide to Artifacts of Colonial America. New York: Alfred
A. Knopf.

Peggs, Deans
1959 A Short History of the Bahamas. Nassau, Bahamas: The Deans
Peggs Research Fund.

Peters, Thelma P.
1960 The American Loyalists and the Plantation Period in the
Bahama Islands. Ph.D. dissertation, University of Florida.

Siebert, Wilbur H.
1913 The Legacy of the American Revolution to the British West
Indies and the Bahamas. Columbus, Ohio: Ohio State

Kathy Gerace
CCFL Bahamian Field Station
San Salvador



David R. Watters and Desmond V. Nicholson


Barbuda is situated between 17031'-17045'N and 61044'-6153'W in the
northern Lesser Antilles. It has a total area of approximately 160 km2 and
two readily distinguishable topographic units: (a) the Highlands plateau
in the east-central part reaching a maximum elevation of about 45 m and
(b) a surrounding marginal plain that occupies about 75% of the island's
landmass. A number of edaphic and climatic constraints effectively pre-
cluded large-scale production of sugar, tobacco, and other commercially
important crops during the colonial era. Shallow, discontinuously dis-
tributed, poorly drained soils, inadequate and unreliable rainfall, large
expanses of bare limestone and accumulated sand, and the absence of surface
streams rendered the island unsuitable for plantation crops, and instead
Barbuda became a "provisioning station" for nearby islands supplying items
such as livestock, hides, meat, corn, and wood.

After two brothers, John and Christopher, acquired the first Codrington
lease in the 1680s, the Codrington family continued to lease Barbuda from
the British Crown for almost two hundred years until 1870. John apparently
lived on the island supervising construction of the "Castle" in the village
until his death in 1688 (Harlow 1928:15,190). Christopher was Governor of
the Leeward Islands colony until his death in 1698. He resided in nearby
Antigua, the administrative center of the colony, but often visited Barbuda.
His son, also named Christopher, succeeded him as Governor and also inherited
Barbuda from him in 1698; in turn he willed it to his cousin William (John's
son) in 1710 (Watters 1980:Appendix A). Supervisors and overseers managed
the island for the absentee lessees throughout the colonial period.

Most historic sites (Fig. 1) on Barbuda are from the period of the
Codrington leasehold. To date, ten sites have been designated; nine were
examined; surface collections were made at five; test pits were excavated
at two (Watters n.d.). Other historic sites are known to exist but they
have yet to be designated or observed.

Highland House Site (Ba-Hl)

Location and Plan
Located atop a 30 m escarpment at the northwest edge of the Highlands
plateau between the 100 and 125-foot (ca. 30-38 m) contours, the Highland
House site commands a superb view of the marginal plain in all directions
except toward the south (Fig. 2). Within the site dense scrub vegetation
has taken hold in patchy soil, which has accumulated at small hollows in the
exposed limestone. Access presently is provided by a rather steep, east-
trending trail from the marginal plain below. Barbudans refer to this site
as "willybob," a word whose derivation and meaning remain uncertain.



University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs