THE ANTEBELLUM BARRIER ISLAND
PLANTATION: IN SEARCH OF AN
SUE MULLINS MOORE
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Sue Mullins Moore
This project would not have been possible without the
help and counsel of a number of people. I am particu-
larly thankful to Sea Island Foundation, which provided
not only monetary support but needed administrative and
technical assistance. I would especially thank Mr. J.D.
Benefield and Mrs. Barbara Otter. Georgia Department of
Natural Resources was our other sponsoring agency as
administrator of National Register matching funds. I
wish to thank them also.
Extensive archival work was carried out at the
Historical Society of Pennsylvania Library in Philadel-
phia, where the Butler documents are housed. I would
like to thank the staff for their extensive cooperation
and for allowing me the use of all the information.
Other archival work was done at the Georgia Historical
Society Library in Savannah; the Margaret Davis Cate
photographs are used by permission of the Society. Ann
Shellander of the Coastal History Museum on St. Simon's
also provided necessary documentary data.
The project was fortunate enough to have input from
John Anthony Scott, the editor of Fanny Kemble's journal.
I am extremely grateful for his kindness and advice.
Spring 1979 field excavation labor was provided
by the University of Florida field school. I would
like to thank Beth Bennett, Dee Daugherty, Helen Doney,
Patricia Fort, Pat Gleason, Jeff Johnson, Keith McIntyre,
Tim Moore, and Jean Wollenberg. The winter 1979 survey
crew and fall 1979 excavation crew consisted of Susan
Harris, Tom DesJean and Barry Hart. Their diligence
was truly remarkable and I cannot thank them enough.
Laboratory assistance was provided by Pat Gleason. Her
aid was invaluable.
Dr. Elizabeth Wing was kind enough to allow me the
use of the Zooarcheology Collection at the Florida State
Museum. Her patience saw me through the long months of
faunal analysis, and her input has been much valued.
I would also like to thank Lydia Deakin, formerly
of the Anthropology Department, for her assistance with
the inevitable administrative problems that arose, and
for seeing me through most of my graduate career.
My colleagues, Chad Braley, Jennifer Hamilton, and
Theresa Singleton, provided me with valuable input from
their own researches, and I am very grateful to them.
I would particularly like to thank Theresa for sharing
data (and commiserations) on the Butler documents.
I would like to thank the members of my committee,
Dr. Edward S. Deevey, Dr. John K. Mahon, Dr. Jerald T.
Milanich, and Dr. Prudence Rice. Their input and patience
has been appreciated.
For my entire graduate career I have had a mentor
and friend in Dr. Charles Fairbanks. I know of no way
to properly thank him for his time and support. He has
been an inspiration throughout my work.
Finally, I would like to thank my parents, James and
Mary Mullins, who saw me through this, and my husband
Tim, who married me knowing I had a dissertation to write.
He has been my critic and support during this passage.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS .............. .................... iii
LIST OF TABLES ............................... ........ ix
LIST OF FIGURES ................................... xii
ABSTRACT ............................................ xvi
1. INTRODUCTION AND HYPOTHESES ................. 1
Research Area ................................ 2
Pattern Recognition in Historical Archeolo-
gy ........................................ 5
Status on the Southern Plantation Social
and Economic ............................ 9
Hypotheses and Research Design ........... 10
Other Proposed Areas of Study .......... 12
Hypotheses ................................ 14
Project Sites ........................... 16
Project Methodology ..................... 20
2. COMPARATIVE RESEARCH ........................ 30
Kingsley .................................... 30
Rayfield .................................... 34
Cannon's Point ................................ 35
Butler Island .............................. 41
Other Sites ................................. 45
3. HISTORICAL BACKGROUND ....................... 50
The Barrier Island Cotton Plantation ...... 52
Comparison to Other Plantation Systems .... 57
Pike's Bluff .................................. 59
Sinclair .................................... 63
Butler Estate ............................... 68
Family History .......................... 68
Property History ...................... 71
Hampton Plantation .................... 72
Main Complex ...... .................. 72
Overseer's House .................... 75
Jones ................................. 76
St. Annie's .......................... 76
Production and Management on the Butler
Estate ................................ 80
4. SITE DATA ..................................... 104
Pike's Bluff ..............................
Kitchen Group Artifacts Ceramics ....
Kitchen Artifact Group Alcoholic
and Pharmaceutical Bottles ..........
Kitchen Artifact Group Food Prepara-
tion and Consumption ................
Architectural Artifact Group ..........
Activities Group ......................
Clothing Group ........................
Personal Group ........................
Tobacco Pipe Group ....................
Faunal Group ..........................
Kitchen Artifact Group Ceramics.....
Kitchen Artifact Group Pharmaceutical
Kitchen Artifact Group Food Prepara-
tion and Consumption ................
Architectural Artifact Group ..........
Activities Group ......................
Furniture Group ....
Clothing Group .....
Personal Group .....
Tobacco Pipe Group .
Faunal Group .......
Jones Creek Settlement
Jones Excavations -
Jones Excavations -
Jones Excavations -
Jones Excavations -
Jones Excavations -
Jones Excavations -
Jones Excavations -
Jones Excavations -
Jones Excavations Cotton Barn .......
Kitchen Artifact Group Ceramics .....
Kitchen Group Artifacts Pharmaceu-
tical Bottles .......................
Kitchen Artifacts Food Consumption
and Preparation .....................
Architectural Group .................. 219
Activities Group ...................... 220
Arms Group ............................ 221
Clothing Group ....................... 221
Personal Group ........................ 223
Tobacco Group ......................... 225
Faunal Group .. ....................... 226
Summary .............................. 228
5. HYPOTHESES AND PATTERNS ..................... 305
Kitchen Artifact Group ................... 307
Architecture Group ........................ 313
Artifacts ............................ 313
Housing Differences .................. 314
Activities, Arms, Clothing, Personal,
Tobacco ................................ 316
Faunal Group ............................. 317
Discussion of Hypotheses ................. 322
Hypothesis 1 ... ....................... 322
Hypothesis 2 ... ...................... 324
Hypothesis 3 ..... ..................... 325
6. SUMMARY AND CONCLUSIONS ..................... 341
APPENDIX: COMMON NAMES FOR DISCUSSED FAUNAL
SPECIES ...................................... 345
ABBREVIATIONS USED IN REFERENCES ................. 348
REFERENCES CITED .... ............................ 349
BIOGRAPHICAL SKETCH .......... ....................... 362
LIST OF TABLES
1. A Comparison Of Artifact Profiles ............ 24
2. Computer Program Format ....................... 25
3. Biomass Formula And Constants ................. 28
4. Kingsley Ceramic Data ......................... 48
5. Statistics For Large Plantations In Louisiana 86
6. Pike's Bluff Documentary Data ................. 87
7. Probate Inventory Thomas F. Hazzard ........ 88
8. Sinclair Documentary Data .................... 90
9. Butler Estate Documentary Data ................ 91
10. Butler Estate 1815 Inventory ................ 93
11. Butler Estate Births and Deaths of Slaves ... 94
12. Butler Estate Slave Census Data ............ 95
13. Pike's Bluff Ceramic Totals ................ 115
14. Pike's Bluff Ceramic Form Data .............. 117
15. Pike's Bluff Ceramic Form Data By Type ...... 118
16. Pike's Bluff Faunal Data ................... 119
17. Pike's Bluff Faunal Data Test Squares A & B 121
18. Pike's Bluff Faunal Data Test Square C ...... 122
19. Pike's Bluff Faunal Data Test Square D ...... 123
20. Pike's Bluff Faunal Data Test Square E ...... 124
21. Pike's Bluff Artifact Profile ................ 125
Form Data .................
Form Data By Type .........
Data Kitchen Area ..........
Data Test Squares B,C,D, & E
Data Domestic Slave House ..
Data Main House ............
31. Jones Ceramic Data ...........................
- Ceramic Form Data By Type ............
- Ceramic Form Data ....................
Cabin 4 ..................
Cabin 4 North ............
Cabin 4 South ............
Tool/Storage Shed ........
Road Area ................
Cabin 2 ..................
45. Jones Faunal Data Well .....................
46. Jones Artifact Profile ....................... 248
47. Comparison Of Artifact Profiles ............. 327
48. Comparison Of % Of Annular And Transfer
Printed Wares On Coastal Sites ............. 329
49. Comparison Of Sites By Ceramic Pricing Levels 330
50. Comparison Of Sites By Ceramic Form .......... 331
51. Comparison Of Sites By Housing Differences ... 332
52. Faunal Data Cannon's Point Planter ......... 333
53. Faunal Data Cannon's Point Overseer ........ 335
54. Faunal Data Cannon's Point North Slave Cabin 337
55. Comparative Domestic and Non-Domestic Animal
Use ........................................ 339
LIST OF FIGURES
1. Site Location Map .............................
2. Georgia And Florida Plantation Sites ..........
3. 1939 Photograph Of Slave Duplex At Hampton
4. 1939 Photograph Of Slave Duplex At Hampton
Point Interior ............................
5. Main Plantation Complex Hampton Point
Subsidiary Structure ........................
6. Overseer's House Ruins Hampton Point ........
7. Pike's Bluff Site Plan ........................
Main House Hearth Base ...........
Main House Hearth Base ...........
Privy Profile ....................
Privy Plan .......................
- Bung ...........................
- Transfer Printed Ceramics ......
- Transfer Printed Ceramics......
- Ceramics .......................
- Ceramics .......................
- Metal Artifacts ................
- Metal Artifacts ................
21. Pike's Bluff Astrolabe ..................... 148
22. Pike's Bluff Bone Artifacts ............... 150
23. Sinclair Main House Plan .................... 177
24. Sinclair Main House Ruins ....................
25. Sinclair Main House Ruins ....................
26. Sinclair McIntosh Graves ...................
27. Sinclair McIntosh Graves ...................
28. Sinclair Site Plan ... ........................
29. Sinclair Test Squares B,C,D, & E Slope Area
30. Sinclair Kitchen Midden ......................
31. Sinclair Probable Slave House ................
32. Sinclair Ceramics ..........................
33. Sinclair Pharmaceutical Bottles ............
34. Sinclair Metal Artifacts ...................
35. Sinclair Bone Artifacts ....................
36. Sinclair Metal Artifacts ...................
37. Sinclair Clothing And Tobacco Artifacts ....
38. Sinclair Broad Hoe .........................
Creek Site Plan ........................
Creek Cabin 3 ..........................
Creek Cabin 3 Chimney Plan .............
Creek Cabin 4 Duplex .................
Creek Feature 5 Profile ................
- Cabin 4, Feature 5 Wall Trench.......
- Cabin 4, Feature 5 Wall Trench .......
Creek Support Structure ............... 259
Creek Possible Tool Shed .............. 260
- Tool/Storage Shed ................... 262
- Tool/Storage Shed ................... 264
- Tool/Storage Shed Feature 2 .......... 266
- Road Plan ........................... 268
- Road Profile ...... .................... 269
- Road ................................ 271
Creek Feature 13 ....................... 273
Creek Feature 13 Profile .............. 274
- 409N,382E Plan ........................ 275
- 409N,382E Possible Structure ......... 277
- 438N,421E ........................... 279
- Well Pit ............................ 281
Creek Well .............................. 283
- Well Timber .......................... 285
- Well Bottom ........................... 287
Creek Cotton Barn Plan ................ 288
Creek Gin Barn ... ................... .. 289
Creek Iron Support Cotton Barn ....... 290
- Cotton Barn North Room ............... 292
- Ceramics ............................. 294
- Pharmaceutical Bottles ............... 296
- Metal Artifacts ...................... 298
70. Jones Metal Artifacts ...................... 300
71. Jones Arms and Fishing Artifacts .......... 302
72. Jones Personal Use Artifacts .............. 304
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Pequirenents
for the Degree of Doctor of Philosophy
THE ANTEBELLUM BARRIER ISLAND
PLANTATION: IN SEARCH OF AN
Sue Mullins Moore
Chairman: Charles H. Fairbanks
Major Department: Anthropology
Plantation archeology,has, in recent years, focused
on the derivation of status indicators on planter, over-
seer, and slave sites. This study, by examining artifact
patterns on several plantation sites, has determined that
economic level of the site inhabitant is an equally deter-
mining factor in the material culture assemblage.
Sites of known size (based on number of slaves) are
examined and artifact profiles (patterns) are derived.
The artifact frequencies of these profiles are statis-
tically compared and it is found that as size of the
plantation increases the frequency of non-essential arti-
facts also increases. The material culture of a small
planter is seen to be very similar to that of an overseer
on a large estate. This reflects the influence of economic
position rather than social status.
Primary data were obtained from three sites exca-
vated on St. Simon's Island, Georgia. It was felt that
this research should be restricted to coastal sites,
primarily barrier island, as they operated in a single
economic system based on the cultivation of sea island
cotton. Comparative data were derived from three other
coastal sites. Artifact groups and status indicators,
previously defined by other researchers, were used in
INTRODUCTION AND HYPOTHESES
The study of slavery has, in recent years, gained
valuable new data from anthropologically oriented
archeological studies of plantations site (Otto 1975,
McFarlane 1975, Drucker and Anthony 1979, Singleton
1980). The plantation system of the American south has
been the subject to the biases inherent in their primary
source material. Contemporary accounts of the plantation
system often suffered from falsification and from the
fact that they were told from the perspective of the
educated white observer. The later slave narratives
suffered from the passage of time and the inability of
the informants to recall details of their day-to-day life.
By combining the evidence from the historical accounts and
contemporary documentation with that from archeology,
a more accurate and perhaps more complete picture of the
plantation system is possible.
Plantation archeology has focused primarily on
the excavation and discussion of single plantation
sites. Other than a on-to-one comparison, usually
with John Otto's (1975) work at Cannon's Point, almost
no attempts have been made to look for general patterns
within the plantation system. Because of the differences
that existed in organization and environment of the
plantations in the different cash crop regions, the focus
of this study was narrowed to one of these systems, the
barrier island cotton plantation. As most of the plantation
archeology completed has been in this region, it was clear
that the coastal situation was a fitting start. The pur-
pose of this pattern recognition, in turn, was to dis-
cover regularities in the artifact configurations re-
flective of underlying cultural processes. Specific
hypotheses will be discussed later in this chapter.
The barrier islands situated off the coast of the
southern United States are suffering increasing pres-
sure in the form of development. Efforts to mediate
the impact of this development on the natural and cultural
resources have included considerable archeological work
on St. Simon's Island (Fig. 1). Under the auspices of
the National Park Service, Sea Island Foundation, and
the National Register of Historic Places, a great deal
of archeological information has been gathered about
life on St. Simon's during the colonial and plantation
Continuing excavations at Fort Frederica National
Monument have greatly increased the knowledge of colon-
ial Georgia and life at a military colony (Fairbanks
1956; Honerkamp 1974; Reitz 1979; and Honerkamp 1980).
In 1972, the University of Florida began a program of
excavations at Cannon's Point (owned by Sea Island
Company) on the north end of St. Simon's with funding
provided by Sea Island Foundation and the National Science
Foundation. This project produced a dissertation on the
late Archaic period of the coast (Marrinan 1975), one on
the proto-historic Guale Indians (Wallace 1975), and a
third on an antebellum plantation (Otto 1975). It also
produced two master's theses (McFarlane 1975, Martinez
1975). This work ended in 1975.
The University was asked again in 1978 to do an
archeological survey of Butler Point, also on the northern
end of the island (Mullins 1978):. The most signi-
ficant outcome of this survey was the location of the
ruins of Hampton plantation. This estate is the focus
of much of this dissertation.
Late in 1978, Sea Island Foundation initiated another
project with the University, this time for the complete
survey of Sea Island's remaining holdings on St. Simon's.
This amounted to roughly 5000 acres, or approximately one-
half of the high land on the island. Because this survey
included the known ruins of several plantations, and
some significant prehistoric sites, additional funds
were granted by the National Register of Historic
Places. A total of nine months were spent in the field,
involving the location, testing, and in some cases,
partial excavation of twenty sites. Five different
plantations were located and tested; three are the sub-
ject of this dissertation.
One of the "Golden Isles," St. Simon's is located
within the natural area referred to as the southeastern
coastal plain. This plain extends south and east from
the piedmont fall line to the Gulf of Mexico, Atlantic
Ocean and the sub-tropical lower half of peninsular
Florida. It has an equitable climate with a mean winter
temperature of 43 F. and annual precipitation amounting
to approximately 53 inches (Johnson et al. 1974). It
has been noted that most of the soils are excessively
drained and have moderate to severe leaching (Long 1912:3;
Johnson et al. 1974).
St. Simon's is one of the barrier islands which were
formed during the Pleistocene and early Holocene periods
at times of lowered sea level. At that point they were
part of the mainland, the lagoon and salt marshes having
not yet formed. When submergence began upon melting
of the Pleistocene glaciers, they became islands separated
from the mainland (Hoyt 1967).
Pattern Recognition in Historical Archeology
The beginnings of the search for patterns in the
archeological record can be seen in the rise of scien-
tific or problem oriented archeology. Patterns had long
been a focus of study in anthropology (Benedict 1934;
Kroeber 1948; Steward 1955), although the concept and
use of the word was somewhat different. With Steward
can be seen the beginnings of the nomothetic-processual
approach (South 1977). There was a need to see through
superficial cultural differences and find more underlying
similarities that were reflective of similar cultural
Stanley South has been the foremost proponent of
pattern recognition in historical archeology (South
1975, 1977). South has presented several patterns, the
Brunswick Refuse Disposal pattern, the Carolina Artifact
pattern, and the Frontier pattern (South 1977).
The Brunswick Refuse Disposal pattern is distinguished
by trash deposition adjacent to structures, particularly
at doors and windows and is present on eighteenth cen-
tury British-American sites. These deposits seem to be
secondary refuse and South defines two types: 1) a low
bone-artifact ratio seen in the deposits immediately
adjacent to the structures, and 2) a high bone-artifact
ratio found in peripheral refuse deposits. South expresses
the Brunswick pattern in a law-like generalization: on
British-American sites of the eighteenth century a con-
centrated refuse deposit will be found at the points
of entrance and exit in dwellings, shops, and military
fortifications. Testing at several sites appears to
confirm South's hypothesis.
The Carolina Artifact pattern consists of a pre-
dictable frequency of certain artifact groups on sites,
dating from approximately 1730-1830. South's pattern
is based on five sites, primarily domestic. Table 1
gives the artifact groups and the mean and range of per-
centages for each. The pattern was tested on other sites
and found to largely supported, but there were some
differences. South points out that the ability to as-
certain irregularities in the pattern may be one of its
When South was testing the Carolina Artifact pattern
he found several sites which would not fit the pattern
at all. From these sites the Frontier Artifact pattern
was derived. The adjusted mean and range of the arti-
fact pattern are given in Table 1.
When the Frontier pattern was compared to the
Carolina pattern, it could be seen that there was a re-
versal in the ratio of kitchen and architecture artifacts
(Table 1). The much higher ratio of kitchen artifacts
in the Carolina pattern may be explained by proximity
to supply sources, the frontier sites being much farther
from the distribution sources.
Application of South's Carolina pattern have been
made at several sites. At the Spiers Landing site in
South Carolina, the artifact frequencies of a probable
slave site were compared and found to differ significantly
in the clothing, activities and kitchen categories (Drucker
and Anthony 1979). From this evidence it was suggested
that new patterns might be necessary for non-urban or
The Carolina Artifact pattern was tested at another
South Carolina slave site and a Carolina Slave Artifact
pattern was proposed (Garrow 1981). The original Caro-
lina pattern was revised to include colonowares (of pos-
sible slave manufacture) in the kitchen category rather
than activities where South had put them. This made
a significant difference in the two patterns (Table 1).
Theresa Singleton also proposed a Slave Artifact
pattern for four sites in Florida and Georgia (Singleton
1980:216). This pattern appears much more like the
Frontier pattern than the Carolina or Carolina Slave
Artifact patterns. Singleton offers no real explanation
for this except to suggest that there was a scarcity of
midden deposits in her sample which probably biased the
The quantification of these patterns has been a
fundamental part of their utility in testing hypotheses
and comparison to other patterns, but it has also been
used to mask one of the primary problems in pattern recog-
nition, the neglect to go beyond the pattern to the pro-
cesses that are responsible for the patterns (see Honerkamp
1980:29). The pattern in itself is merely another parti-
cularistic description. Once the processes that create
the pattern are understood, then the differences or
similarities between patterns can be explained from a
processual viewpoint. So far, this approach has scarcely
At this point, it is felt that it will not be possible
to derive a single pattern applicable to all barrier island
plantation sites, but the derivation of patterns specific
to individual sites or components (planter, overseer,
slave) will allow quantitative comparisons to be made.
Pattern recognition, then, will be used as a methodological
approach (Honerkamp 1980:28-29). The inability to derive
one pattern may perhaps be the most significant finding
of this study and is implied in the hypotheses stated
below. These hypotheses were designed to explore the
processes responsible for the patterns, particularly
status, both social and economic. This work will examine
status both within and between sites as it is hypothesized
that size of the plantation, and therefore the economic
status of the planter, will affect the material goods
available to a site's inhabitants.
Status on the Southern Plantation Social and Economic
Status can be described by a number of qualifiers -
age, sex, biological relationship, social class, and
economic levels (Linton in Bohannon and Glazer 1973:
187-200). Status, without one of these descriptions,
generally refers to the sum total of all the statuses
the individual occupies (Linton in Bohannon and Glazer
1973:187). Each individual, therefore, can have a number
of different statuses, none necessarily dependent on any
of the others (cf. Nash 1970:3).
Social status in southern plantation society has
been discussed extensively by historians and archeologists
(Stampp 1956; Eaton 1961; Otto 1975). Basically, ethnic
caste (black and white) divided this society into two
social strata. Blacks occupied the lowest stratum and
were lumped together, whether free or slave (Genovese
1974:398-413). Whites, regardless of economic position,
were considered to be of a higher social status. Slave
ownership, in particular, was correlated with social
prestige and political power (Wright 1970:68). Three
levels of social status can be seen within the white social
strata, "poor white", yeoman farmers and merchants, and
large planters (Stampp 1956:29;.Bonner 1965:58). Small
planters (those with fewer than 20 slaves) were considered
part of the yeoman farmer group (Bonner 1965:58). Otto
notes that even though the poorer whites were assigned
the same broad social status as the yeoman farmers and
planters, they often endured material living conditions
comparable to slaves (Otto 1975:14). This is one indi-
cation of how economic level should affect the archeo-
logical remains of a site.
Historical archeology is fortunate in that many times
status can be determined from the documents (cf. Otto
1975:12-13), rather than having to rely on material
remains as the sole indicator. Using number of slaves
owned as the most sensitive indicator of economic level
(Menn 1964, Stampp 1956:30-31) for each site in this
study, an attempt will be made to correlate and/or compare
social status with economic status. It has already been
pointed out that these are not necessarily the same
(Otto 1975:14; Nash 1970:3). It is expected that the
archeological record will demonstrate how these differ-
ences are manifested in the material culture of the plan-
Hypotheses and Research Design
The artifact categories being considered for the
following hypotheses were modified from South's groups
(1977:95-96), and are listed below. Not all of the dif-
ferences can be addressed by the frequency derivations of
South's (or any other) pattern. To this end, certain
other qualitative and quantitative differences are
hypothesized. These are also listed below.
Recent work with South's pattern has suggested
that colonoware ceramics should be moved from the
activities group to the kitchen group (Garrow 1981).
It is also proposed that the stub-stemmed pipes be moved
from activities to tobacco pipes. A category has been
added for faunal material.
pharmaceutical type bottle
glassware (stemmed, decanter, dishes, misc.)
tableware (cutlery, knives, forks, spoons)
kitchenware (pots, pans, pothooks, gridrion,
trivits, metal teapots, water kettles, coffee
pots, buckets, handles, kettles, colonowares,
construction hardware (hinges, pintles, shutter
hooks, staples, fireplace backing plates;
lead window cames, etc.)
door lock parts (doorknobs, case lock parts,
keyhole escutcheons, lock bolts, and brackets)
hardware (hinges, knobs, drawer pulls and locks,
escutcheon plates, keyhole surrounds, handles,
rollers, brass tacks, etc.)
musket balls, shot sprue
gun parts, bullet molds
hook and eye fasteners
personal items ( wig curlers, bone brushes,
mirrors, rings, signet sets, watch fobs, fob
compass, bone fan, slate pencils, spectacle
lens, tweezers, watch key, and other personables)
Tobacco Pipe Group
long and stub-stemmed pipes
construction tools (plane bit, files, augers,
gimlets, axe head, saws, chisels, rives,
punch, hammers, etc.)
farm tools ( hoes, rake, sickle, spade, etc)
toys (marbles, jew's harp, doll parts, etc)
fishing gear fishhookss, sinkers, gigs, harpoons)
storage items (barrel bands, brass cock, etc.)
stable and barn (stirrup, bit, harness boss,
horseshoes, wagon and buggy parts, rein eyes,
miscellaneous hardware (rope eye thimble, bolts,
nuts, chain, andiron, tongs, case knife,
flatiron, wick trimmer, washers, etc.)
other (button manufacturing blanks, kilnwaster
furniture, silversmithing debris, etc., re-
flecting specialized activities)
Military objects (swords, insignia, bayonets,
artillery shot, and shell etc.)
Other Proposed Areas of Study
1. construction materials and techniques
a) larger plantations should have more substantial
structures, i.e. more tabby and brick
b) smaller plantations should have less sub-
stantial structures, less tabby and brick
2. expected durability
a) larger expect longer durability with more
b) smaller less durability, less maintenance
3. available living space
a) larger more living space and greater range
b) smaller less living space and less range
4. building hardware
a) larger no difference in basic types, but
perhaps a higher proportion because of
b) smaller smaller proportion because of
5. features available to occupants fireplaces,
porches, cooking facilities, storage areas,
wells, privies, refuse areas
a) larger more features and a greater range
b) smaller fewer features, smaller range of types
1. domestic plant and animal food
a) larger more consumption
b) smaller less consumption
2. non-domestic animal foods
a) larger smaller range and less consumption
b) smaller greater range and more consumption
3. food preparation facilities and techniques
a) larger more facilities, better techniques
b) smaller fewer facilities and cruder
1. The artifact patterns of configurations on a
plantation site, regardless of whether it is a planter,
overseer, or slave context, will vary with size of the
plantation. These pattern variations are believed to
be a function of the planter's economic status. Larger
plantations should have greater access to preferred
goods. These differences should be reflected in the
frequencies of the artifact groups and the already men-
tioned differences in other areas. As plantation size in-
creases there should be larger quantities of kitchen
related artifacts at the expense of architectural ar-
tifacts (cf. Otto 1975:13). The other groups usually
have comparatively small frequencies, and may therefore
be difficult to assess as far as the significance of
any differences. It is tentatively proposed that: 1)
personal and tobacco group artifacts, which represent
luxury items, will increase with plantation size, and
2) activities, arms, and clothing will decrease with
increase in plantation size as they are more representative
of necessity items.
It should be pointed out, that as any system grows
larger, it becomes more complex. Applied to this study,
this translates to the larger the plantation, the more
varied the activities carried out there, resulting in
more material goods and resources available to the plantation
inhabitants. This phenomena is not unique to cultural
systems and has been extensively treated in the litera-
ture on systems theory (Von Bertalanfy 1975; Chapman 1977).
2. When large and small plantation sites are compared,
it is hypothesized that the artifact patterns within
a small site (planter, slave) will be more homogenous
than the intra-site patterns of the large site. This
hypothesis will be tested by comparing the artifact
frequencies within the prescribed categories already
discussed. It is hypothesized that differences in fre-
quency and range will be less within the small site than
a larger site. This again is proposed to be a function
of economic status of the planter. On a small plantation,
more of the money available would be directed towards
production of a crop and less towards non-necessity items
for the planter or his dependents (Flanders 1933), which
should be reflected in the material culture.
3. The artifact patterns of the domestic and field
slave are proposed to differ because of the supposed
higher status of the domestic slave (Stampp 1956:337-
338; Owens 1976:106-120). It is hypothesized that the
domestic slave will have greater access to preferred
artifacts. The same artifact categories used for hypo-
theses 1 and 2 will be used.
As a definition for the sizes of the plantations,
historical sources were consulted (Flanders 1933:128; Eaton
1961). Initially the division was three fold, large:
those over 100 slaves; medium: those with 20-100 slaves;
and small: those under 20 slaves. The vast majority
of the south's plantations fall into the small category
(Flanders 1933:129; Eaton 1961). Problems with the field
data necessitated the combination of the medium and
The sites chosen for this study are reviewed in
the following section and in Chapter 3. John Otto's
data from Cannon's Point will be used extensively along
with the available comparable data from Kingsley and
Butler Island plantations. All are barrier island
cotton plantations, with the exception of Butler Island,
which is a coastal rice plantation, and the size of each
is known. The status indicators derived by Otto are
used in conjunction with the testing of Hypotheses 1,
2, and 3.
Comparative data comes from three sites (Hampton,
Sinclair, and Pike's Bluff) partially excavated during
the project. All three are antebellum plnatations located
on St. Simon's Island. Both the size of the plantations
and the size of the components excavated vary. This dis-
cussion is very brief and intended only for orientation.
A detailed documentary account will follow in Chapter
3. The sites used in the study which were excavated
prior to this research are discussed in Chapter 2.
Hampton plantation on Butler Point is the largest
plantation studied, with, at one point, more than
300 slaves. Its owner was Major Pierce Butler, one of
the original delegates to the Constitutional Convention.
Hampton was probably the most prosperous of St. Simon's
estates producing the staple crop of sea island cotton.
There were four slave settlements at Hampton. One was in
the immediate vicinity of the planter's house while the
other three were scattered, one being over three miles
from the main complex. The overseer's house was, at
various times, located within the main complex and then
at one of the outlying settlements. The overseer in
the case of Hampton served more as a manager since Pierce
Butler hardly ever resided on his Georgia estate (Van-
Excavations at Hampton were conducted almost exclu-
sively at Jones, one of the outlying slave settlements.
Limited information from a previous survey is available
on the big house and the overseer's complexes. The
slave settlement excavations included work at one slave
cabin; an auxilliary structure, probably a tool shed;
a well; and a cotton barn. Several non-structural areas
were also tested. The material from Jones is expected
to yeiad data on the slave pattern. It is thought that
it will vary from Cannon's Point because of the size and
possibly management differences.
The second plantation included in the study is
Sinclair or St. Clair. It also grew sea island cotton,
though not to the scale of Hampton. Sources indicate
that between 10-50 slaves resided on this estate (Hawes
1956, U.S. Bureau of the Census 1820). It was owned
by General Lachlan McIntosh, of Revolutionary War fame,
and began operation under the management of his son Wil-
liam around 1790 (Hawes 1956). The plantation changed
hands several times before finally coming under the owner-
ship of Pierce Butler about 1820 (Vanstory 1970:156; Butler
Papers). When Fanny Kemble visited in 1839, she described
the house as being in a ruinous state (Kemble 1961:246).
It burned in 1857 (Vanstory 1970:156).
The excavations at Sinclair consisted of tests in
two midden areas, one believed to be the kitchen trash
area, the big house, and a small house that was probably
a domestic slave cabin. There were almost no surface
indications of the site, therefore the overseer's house
(if there was one) or additional structures were not lo-i
cated. The remains from these excavations should yield
data on both the domestic slave and the planter. The
slave data can be compared to try to detect differences
in the pattern of field and house slaves. The planter
site will be used in conjunction with the others of the
study to derive information on the range of planter mater-
The final plantation being considered is Pike's
Bluff, a small estate. There .were probably no more than
thirty slaves employed and it is believed that there was
no overseer. The owner of the plantation was a Dr.
Thomas F. Hazzard who purchased the land in 1827 (Glynn
County Superior Court 1827). His brother, William, owned
the adjacent plantation. Documentary evidence on this site
is very scanty, but the date of abandonment appears to have
been 1857 when Thomas Hazzard died. The site was used
a a military outpost during the Civil War (Heard 1938)
and was not occupied after that date.
The site today consists of looted ruins of what is
presumed to be the planter's house, a privy and several
middens. No slave cabins could be located. This may
be due to the fact that the plantation was so small and
the slave cabins much less substantial than those on
larger estates. Excavations were carried out at three of
the middens, the privy and a small test at the house.
The privy and midden provided large quantities of mater-
ial which should illuminate the life of a small planter.
The sites used for this study, then, can be ranked
in descending size as follows:
1) Hampton and Butler Island (same owner)
2) Cannon's Point
5) Pike's Bluff
This ranking should be reflected in the artifact patterns
in the manner specified by the research hypotheses.
The sites chosen for excavation were located and
assessed during a survey of St. Simon's carried out
from January to March of 1979. Documentary evidence
aided in this survey location. The sites were judged
on the basis of several criteria: 1) disturbance to
the site, 2) size of the plantation during the antebellum
period, and 3) availability of documentary sources.
As it turned out, considerable exceptions had to be
made for the third criteria.
Four sites were selected for excavation, Jones,
Sinclair, Pike's Bluff and West Point. West Point
later had to be eliminated (after three weeks of testing)
because there had been too much post-bellum disturbance
which had not been evident during the survey testing. This
created a gap in the information about a small plantation.
The particular segment at West Point to be excavated
was the slave cabin area. This information was not able
to be replaced by data from another site.
In addition to the excavated sites, data were avail-
able from two previously tested plantations on St. Simon's,
Cannon's Point (Otto 1975, McFarlane 1975) and the main
complex of Hampton (Mullins 1978).
Excavation of the selected sites began in April of
1979 and continued until June. During this time period
all of the work at Jones and part of the work at Sinclair
was completed. This phase of the excavations employed
the 1979 University of Florida Field School, a total of
nine students and two field supervisors. A hiatus was
taken during the summer, during which time preliminary
analysis of the excavated material was completed. In
September of 1979, excavation began again at Sinclair
and, in addition, work was completed at Pike's Bluff.
This phase of the project employed only three experienced
archeologists (plus an occasional volunteer) and ended
in December of 1979.
The information gathered from the survey was used
to plan the sampling strategy for the excavations. Most
of the structures and some features had been located. In
view of this, placement of excavation units was non-random.
Motorized shaker screens were used for general excavation
while fine troweling was employed in the case of features.
Soil samples were taken where necessary. The screen size
used was 3/8 X 5/8 inch diamond mesh. This later was
realized to have been too large for recovery of some of
the faunal material. Hindsight would suggest the use
of 1/4 or even 1/8 inch mesh in future excavations. At
all sites detailed maps were prepared of extant structures
and any encountered features. Laboratory analysis was
done at the archeology lab at the University of Florida
and in the zooarcheology range of the Florida State Museum.
The artifacts were preliminarily cataloged and sep-
arated for any special treatment. The faunal material
was removed and taken to the museum for analysis.
Metal artifacts, when necessary, were processed for
preservation by electrolysis. Final cataloging was done
following the initial separation. Once this was completed
a system was developed to put the cataloged data into a
computer for easier data handling. Table 2 shows the
computer entry format. This data format later proved to
be too simplistic for some artifact classes, creating some
problems. In coding ceramics, South's (1977) ceramic
numbers were used. This procedure greatly increased the
rate of data handling and allowed quick and easy access
to artifact statistics.
Zooarcheological analysis consisted of identifying
the faunal species represented at each site, weighing
the recovered bone and computing the biomass for each
species. Allometric biomass computation is a relatively
new technique in zooarcheology which allows the deriva-
tion of edible meat available to a site's inhabitants
based on the bone weight recovered from the site. Table
3 gives the formula and constants used for this calcu-
lation. The provenience used for calculating biomass
depends on the researcher; in this case, the individual
field specimen; units were used. This same procedure
was used when calculating the biorass on the comparative
data from Cannon's Point.
In addition to artifact analysis, considerable
documentary research was carried out. The primary
sources are located in Philadelphia, Pennsylvania,
and include an extremely complete documentation of
the Butler estate. Several letters concerning the Mc-
Intosh family's planting interests in Georgia were also
in the Pennsylvania Historical Society Library. Limited
information on the Wylly, Butler and McIntosh families
was found in the Georgia Historical Society Library in
Savannah. In addition, several 1930's photos of the
Butler estate on St. Simon's were found in the Margaret
Davis Cate Collection, also at the Georgia Historical
In Brunswick, the Glynn County seat, are located
the deeds and probate records available for St. Simon's.
While the deeds are fairly complete, the probate records
for the late eighteenth and nineteenth century are very
E- 0 Oiq -
O 0) *
Ei rcil rd
O 0 CN r- cM ( --
N CN in 1- CO rC
m o N 1 OD
S0 o CC)
( 0 (N (N
N ) (N
il 0H 4
r N N O IC O (N
SN 0 0 0 I) 0)
IC N I (N 0 N
() 4J 4-) *H Q r0 .r-4
4 < i4 l U.- 4 E-4 U).
C -4 -I 4 0 o 0
COMPUTER PROGRAM FORMAT
Field Specimen Number, unique
Card number of case
Location, excavation unit
CERAM1 brown stoneware bottles
CERAM5 Canton porcelain
CERAM7 overglazed porcelain
CERAM8 fingerpainted pearlware
transfer printed pearlware
underglazed blue pearlware
blue and green edged
Iberian storage jars
plain white delftware
delft apothecary jar
Field Specimen Number (same as 1)
Card Number (2)
CERAM79 brown salt-glazed stoneware
CERAM80 grey stoneware
CERAM81 transfer printed whiteware
CERAM82 white porcelain
CERAM85 shelledged whiteware
CERAM86 blue and tan stoneware
CERAM87 unglazed earthenware
CERAM88 polychrome whiteware
Table 2 continued
BH3 strap hinge
BH4 hinge, other
BH5 machine cut nails
BH6 wrought nails
BH7 other nails
BH9 lock mechanism or part
BH10 other building hardware
OMET1 lead seal (present or absent)
OMET2 lead weight
OMET3 lead shot
OMET4 lead other
OMET6 brass button
OMET7 non-brass button (metal)
OMET8 cooking vessel
OMET11 iron tools
OMET12 iron other
OMET13 copper coins
Field Specimen Number (same as 1)
Card Number (3)
OMET14 copper other
OMET15 tin cans
OMET16 tin other
BS1 bone buttons
BS2 bone other
BS3 shell buttons
BS4 shell tools
BS5 shell other
PIPEB pipe bowls
BNO number of bowls
DIA4 number 4/64ths diameter
DIA5 number 5/64ths diameter
GLASS1 dark green wine bottle
GLASS2 ale bottle glass
GLASS3 other alcoholic bottle
GLASS4 window glass
GLASS5 pharmaceutical bottle
GLASS6 drinking glass
Table 2 continued
Column Number Variable
35-36 GLASS8 beads
37-38 GLASS9 other
40-44 WEIGHTC Total Ceramic Weight
45-50 WBH Total Weight Building Hardware
51-55 WOMET Total Weight Other Metal
56-58 WBS Weight of Bone and Shell
59-61 WPIPE Weight of Pipe Fragments
62-66 WGLASS Total Weight of Glass
67-69 WTOYS Total Weight of Toys
70 FLINT presence or absence
BIOMASS FORMULA AND CONSTANTS
Equation: Y = ax where
x = skeletal mass
Y = biomass
a = Y intercept of log-log plot using method of least
squares and best fit line
b = slope of the line
Constants a b
Mammal 1.12 .90
Bird 1.04 .91
Turtle .51 .67
Snake 1.17 1.01
Chondrichthyes 1.68 .86
Osteichthyes .90 .81
Non-Perciformes .79 .85
Siluriformes .95 1.15
Pleuronectiformes .89 1.09
Perciformes .83 .93
Sparidae .92 .96
Sciaenidae .74 .81
Scale Miles N
0 1 2
SITE LOCATION MAP
The beginnings of plantation archeology can be seen
in the work at Kingsley plantation (Fig. 2). Located
on a barrier island off the coast of Florida, Kingsley
was the property of Zephaniah Kingsley, a slave importer
(May 1945:145). In 1968, excavations were conducted
at a group of slave cabins. Parts of two cabins, out
of a total group of 32 were tested in order to learn
something about the structure of the house and about pos-
sible retentions of African heritage that might be present
in a slave site (Fairbanks 1974:90). Zephaniah Kingsley
had several African wives and seemed well disposed in
the treatment of his slaves. For this reason, and be-
cause it may have been a slave importing station, it
was thought that there might very well be evidences of
African culture. Such did not prove to be the case.
These and subsequent excavations were ample proof that
slaves came with nothing and were not allowed to repro-
duce any material culture reflecting their African home.
These excavations did, nevertheless, yield information on
Afro-American material culture (housing, diet, material
possessions), a subject which had been previously unknown.
The cabins at Kingsley appeared to be single family
dwellings unlike many on the Georgia coast (cf. Singleton
1980, Mullins 1978). They were of tabby construction with
brick and tabby fireplaces. One of the cabins was notice-
ably larger than most of the others. Fairbanks suggested
that it may have been the home of the slave driver. This
structure measured 24.5 feet by 18.6 feet (7.5 X 5.7 meters)
with walls approximately seven feet high. There was the
suggestion that the interior walls were plaster coated
and that the cabin was divided into two rooms, the western
being considerably larger than the eastern. The larger
room appeared to have had a tabby floor. It is possible
that the cabin had a loft for sleeping, but this was
not evident from the archeology. The other cabin tested
appeared to be of the same construction.
The ceramics recovered gave evidence of a long period
of occupation. The most frequent type was ironstone which
is also the latest type. Brown stoneware bottles were
the next highest in frequency. Again this is a late
ceramic. The other three ceramic types listed in the
report represent a much smaller quantity, and perhaps
an earlier occupation. Table 4 gives the frequencies
and dates for these ceramics.
Fairbanks suggested that because the median dates
were so disparate, perhaps two groups of ceramics were
present. The first group had a Mean Ceramic Date earlier
than would be expected for a plantation not occupied
until 1813. It was suggested that they might represent
discards from the planter's house. Recent criticism of
the Mean Ceramic Date has suggested that it may have in-
herent biases that cause early dates (Adams and Gaw 1977:
229; Braley 1980). While this does not necessarily rule
out the possibility of planter discards, it provides another
possible explanation and calls for further exploration
of the problem. The second group, the stoneware and iron-
stone, representing the majority of the ceramics, may
well be indicative of a later re-occupation of the
cabin after Kingsley's death in 1843 (Fairbanks 1974:
79). Other evidence of re-occupation came from the pre-
sence of wire nails, not available until after 1850
Of the artifacts, glass was the most abundant.
Most of this was green bottle glass. The most interesting
fact to note is the scarcity of flat glass, indicative
of windows or mirrors. Only 28 fragments were recovered,
which suggested that the windows in the slave cabins lacked
glass and were, instead, dependent on shutters. A shutter
pinion reinforced this idea.
Tobacco pipes were another item of relative scarcity,
with only 15 fragments being recovered. They appeared
to be typical of the period of occupation. Fairbanks
suggested that the small number was not reflective of
the frequency of tobacco use and perhaps that they were
using wooden or corncob pipes (Fairbanks 1974:86).
The discussion of faunal resources was not quantified
so only a gross species list is available. It is not known
how complete this is. Only fish, cattle, pigs, raccoons,
turtles, clams, and oysters were noted. It is not known
what kind of fish and turtle were recovered.
Included in the discussion of dietary items were
artifacts concerned with subsistence. The presence of
lead shot was noted, perhaps indicative of hunting. The
presence of such artifacts contradicts the belief of
most historians about slaves access to guns. This con-
tradiction will be supported by other evidence presented
in this work. Net weights, which in many cases are per-
forated lead shot, were also present. Cooking equipment
was not well represented. A three-legged skillet and
three-legged cast iron pot suggested that cooking was
done on an open fire. One spoon, one fork, and two table
knives were also present.
The overall suggestions of the subsistence material
is that plantation supplies were being supplemented by
wild game, fish, and shellfish. This idea has been
repeatedly confirmed in later work (Otto 1975, Singleton
1980). It has been suggested recently that this may
have been necessary to insure an adequate diet (Gibbs
et al. 1980:248). Cooking was probably done by individual
families rather than central kitchens. The presence of
food bone in the fireplace reinforced this idea.
Fairbanks found that the most surprising fact about
the excavations at Kingsley was the total lack of any
African elements in the material culture assemblage,
particularly in light of Kingsley's position of slave
importer. It seems that, even with Kingsley's permissive
attitude, slaves were simply not allowed to reproduce
any part of their material African heritage.
The Kingsley excavations were the beginnings of
an effort to try to define what life was really like on
a plantation. Until that point, slave sites had been
almost totally ignored. These excavations were able
to provide information on areas not previously addressed
by historians. In some cases it contradicted the his-
torical documents. Further work was necessary before
the reliability of this new evidence could be assessed.
Rayfield plantation was located on Cumberland
Island, the southernmost of Georgia's barrier islands
(Fig. 2). Owned by Robert Stafford, it produced sea
island cotton as did most of the barrier island planta-
tions. Again, as at Kingsley, the excavations focused
on the slave cabins. In this case only one cabin was
tested. There were 18 cabins in all, each a single
family dwelling, arranged in two parallel rows. It was
estimated that each cabin was 18 X 18 feet (5.5 X 5.5
meters) (Ascher and Fairbanks 1971:8). The cabins were
of frame construction with a tabby and red brick chimney.
There were few artifacts recovered. Nothing diagnos-
tic can be said about the ceramics. The nails recovered
indicated that the cabin had been constructed after 1825,
a fact supported by the documentary evidence. The faunal
material was very scarce and so fragmentary that it was
impossible to identify.
While the excavations at Rayfield did not provide
any quantity of information on slave life, it did allow
the accumulation of data on the plantation system.
Information was gathered at least on the physical surround-
ings of the slaves.
Located on St. Simon's Island, Cannon's Point plan-
tation was occupied from approximately 1795 to 1860.
It was the property of John Couper and remained in the
Couper family for its entire period of production.
Couper was a Scotsman and had a great interest in agri-
culture. He was often experimenting with new crops or
techniques (Otto 1975:41). One successful experiment
was the growth and processing of olives for oil. Cannon's
Point was a large plantation averaging about 100 slaves
during its productive years. The main house was an im-
posing structure three stories high. There were two
slave settlements, one near the main house and one about
one mile south. The overseer's house was located with
the southern group.
Otto's main goal, in the work at Cannon's Point,
was to try to define status differences in the material
culture assemblages of planter, overseer, and slave
(Otto 1975). It was found that these differences were
most noticeable in the housing, ceramics, and dietary
remains. Work was also conducted by Suzanne McFarlane
at the southern group of slave cabins (McFarlane 1975).
While this work was primarily descriptive in nature, it
provided excellent information on the structure of a
Otto found that at the planter site there were
more permanent structures as compared to lower status
sites. The buildings were built to last longer and
provide more comfort to their inhabitants. There
were also a greater diversity and number of specialized
areas available to the planter. Particularly, number
of enclosed rooms was correlated with status. While
the evidence did not support it, Otto felt that building
hardware could also be a reliable indicator, the planter
having more permanent building hardware.
The overseer existed somewhere in the middle when
living structures were examined. He had more rooms and
more living space per person than the slave but fewer
rooms than the planter. Otto found that the overseer
had more living space per person than the planter, but
felt this was entirely dependent on the make-up of the
families. Overseers were often single and this would
affect the calculations. In actuality, Otto felt that
the overseer's house at Cannon's Point was much better
than others outside of coastal Georgia. Otto hypothe-
sized that this was due in part to the fact that the
coastal plantations were more permanent than inland
There were two types of slave cabins excavated.
The northern cabins were single family units, while the
southern group consisted of duplex cabins. The indi-
vidual units were of similar size, the northern being
17 X 20 feet (5.2 X 6 meters), and the southern 20 X 20
feet (6.2 X 6.2 meters). They were of frame construction
and probably had wooden floors. Both sets of cabins had
tabby and brick chimneys. These features describe cabins
superior to most others outside of the coastal south.
The average cabin size in the plantation south was less
than 16 X 18 feet (4.9 X 5.5 meters) (Fogel and Engerman
1974:115-116). Instead of brick chimneys, they often
had mud and stick ones (Fogel and Engerman 1974:115-116).
This propensity for better housing not only for slaves, but
for overseer and planter as well, may be because this was
one of the richest plantation areas in the south. The
crops were very specialized and often demanded high
prices. An elite plantation system developed. There was
also the presence of a very cheap building material,
tabby. Tabby is a mixture of equal parts of shell, quick-
lime, sand, and water (Gritzner 1978:9). The shell was
readily available on the coast in the form of prehistoric
shell middens. This allowed the construction of rather
permanent, solid buildings.
At Cannon's Point, there was an average of 5.4 slaves
per dwelling unit, somewhat more than the figure of 5.2
slaves per unit in the south as a whole. Otto felt that
this may indicate overcrowding in the cabins but was
Otto found also that ceramics were indicative of
site status. In particular, ceramic decorative type and
ceramic form were found to be diagnostic on planter,
overseer, and slave sites. On slave sites it was found
that annular wares were more prevalent than on planter
sites while planter refuse had large amounts of transfer
printed ceramics which were not abundant in overseer
or slave sites. The overseer site tended to be more
like the slave site than the planter. There were also
large amounts of undecorated ceramics on the slave and
overseer sites which were not present on the planter site.
Interestingly, porcelain was not found to be diagnostic
on any of the sites.
It was found that occasionally slaves and overseers
were getting discarded ceramics from the planter. This
practice was discovered by comparing ceramic patterns
which occurred on all three sites. Several patterns were
repeated in minor quantities on the slave and overseer
sites which occurred primarily in the planter refuse. Otto
felt that the planter was purchasing ceramics in sets while
the more heterogeneous ceramic assemblage at the slave and
overseer sites indicated that they were purchasing pieces
individually or in small sets.
In addition to ceramic type, ceramic form was also
found to be correlated with site status. There was a
prevalence of bowls at slave and overseer sites, while
flatware was more common at the planter site. The plan-
ter was also found to have a greater diversity of ceramic
forms, including table, tea, storage, and chamber wares.
Slave and overseer were found to have primarily tablewares.
The final area that Otto found to be indicative of
site status was dietary remains. From the information
available at that time, it was hypothesized that planter
sites would have a higher proportion of non-domestic
animals in the diet when compared to slave and overseer
sites. Otto found that in the planter's refuse there
was much more fish and wild game. He reasoned that this
was possible because the planter was using slave labor
to obtain this part of his diet. The slaves had little
time to hunt or fish for themselves. The slave diet
showed a high proportion of domesticants, particularly
pig and cow. These were provided in the form of rations
from the planter. There was the presence of a quantity
of wild food which Otto attributed to procurement
methods which required little supervision such as traps,
snares, and trot lines.
Since Otto's work there has been devised an accurate
methodwhich transforms bone weight to edible meat (Wing
and Brown 1979:127). This was applied to Otto's data and
the results will be discussed more fully later in this
Otto also hypothesized that the prevalence of bowls
at the slave site was related to dietary patterns. He
found that the slave was getting the poor cuts of meat
while the planter was getting good cuts such as roasts.
The poor cuts (i.e head, feet) were used in stews to make
them more palatable and go further. It was also found
that the bones were being cleaved open and added to the
stew for marrow extraction. The prevalence of hollowares
would seem to be related to this diet.
Otto's work at Cannon's Point was a milestone in
plantation archeology. It was the first comparison of
the material living conditions within the "peculiar
institution". It was a different viewpoint from the
historical accounts and provided information not pre-
viously available on the day-to-day life on the plantation.
Perhaps its major contribution is as a reference for other
studies. It was the first large scale attempt at doing
archeology in a systematic way on a plantation and created
more questions than it answered.
At about the same time that field work began on
the St. Simon's plantations, work also began on one of
the "sister" plantations of the study. Butler Island
and Hampton plantation were part of the same estate
belonging to the Butler family. Butler Island grew
primarily rice while Hampton was invested in the pro-
duction of sea island long staple cotton. Located near
the mouth of the Altamaha River, Butler Island was a
river delta island, connected more to the mainland than
the barrier islands. Beginning about 1820, Butler Is-
land became the major focus of the Butler estate planting
interests as it was about at this time that rice became
more profitable than the long staple cotton (Flanders
1933). The majority of the prime slave force was moved
to Butler Island and most of the plantation improvements
were made there rather than at Hampton. The St. Simon's
plantation was, however, still the main residence for
the estate manager and the Butler family. The barrier
islands were considered much more healthy than the
river delta. In was, in fact, where the ill slaves from
Butler Island were sent to recuperate.
The work at Butler Island was designed for two ob-
jectives: 1) to learn more about the slave community plan,
and 2) to compare slave life on a rice plantation to life
on the barrier island cotton plantations (Singleton 1980:
17). Theresa Singleton, who conducted this work, felt
that part of the material culture would reflect some dif-
ferences contingent upon the economic focus of the plan-
tations. But she felt as well, that the broad patterns would
be the same because of the coastal setting. Specific
hypotheses were posited to test these ideas (Singleton
The detailed documentary background of Butler Island
will not be discussed at this point as it will be reviewed
later. As already mentioned, it was part of the large
Butler estate, and was purchased about 1790. It was not
the major economic focus of the estate until about 1820
for reasons already discussed. Both cotton and rice cul-
tivation were labor intensive and required a large work
force. At its height, there were 919 slaves employed
on the estate, divided between Butler Island and Hampton.
The majority were concentrated on Butler Island, and the
site density reflects this. There were four slave settle-
ments, a main complex, where more of the machinery was
located, an overseer's house, and a cemetery. While
Butler Island slaves were primarily engaged in the produc-
tion of rice, they also processed sugar and cotton. The
slave settlements were located strategically in relation
to the agricultural fields and water courses. Singleton
notes that they were probably rather autonomous units
with infrequent mingling or visiting or the main complex.
Each settlement had its own processing facility, either
rice, sugar, or cotton.
Excavations were conducted at settlement 4, the
farthest from the main complex and least disturbed of the
sites. There was evidence of five duplex cabins, and
two technical structures believed to be a rice barn and
threshing floor. There were no wells because, according
to the documents, slaves were using water straight from
the river. There was a possible privy located near the
canal. The canal also proved to be the refuse dumping
area used by the site's inhabitants. No other trash
disposal areas were located during the excavations.
Testing was carried out on three of the duplex cabins.
One was excavated fairly completely, while the other two
had only limited testing completed. The cabins appeared
to be of frame construction with a raised wooden floor.
The chimney was H-shaped and centrally located to open to
both sides of the duplex. Most of the bricks used in the
construction of this feature were tabby, with clay bricks
being used primarily around the hearth. A fragment of
wood recovered indicated that the structures may have
been made of cypress, a very durable building material
and abundant in the river delta. Excavations indicated
that the cabins may have been larger than usual, approxi-
mately 24 X 48 feet (7.4 X 14.8 meters). From docu-
mentary sources, it was estimated that there were 5.4
occupants per dwelling unit. This is comaprable to most
estates in the south (Fogel and Engerman 1974:115).
At Butler Island the ceramics seemed to follow much
the same pattern as at Cannon's Point. There was a pre-
dominance of bowl forms. The decorative types showed a
high proportion of annular wares and low amounts of
transfer printed ceramics.
The recovered faunal material was very poorly pre-
served at Butler Island, but seemed to indicate a pat-
tern of use similar to Cannon's Point. Biomass was
not calculated for the Butler Island material. The
pattern of non-domestic animal use indicated a high
reliance on animals specific to the delta area, par-
ticularly fish and turtle.
By using the data from Butler Island, Kingsley,
and Cannon's Point, a slave artifact pattern was
proposed. The major components of this pattern were
architectural and kitchen artifacts. The other categories
of furniture, personal possessions, firearms,, tobacco
and clothing were very scarcely represented. The pat-
tern will be discussed in detail later in this thesis.
It was found from the excavations that some features
of the site may be a specific adaptation to the river
delta/marsh hahitat. In general, the community pattern
seemed to be like other tidewater plantations (Singleton
1980:220). As for the slave housing, Singleton found
little difference between the housing at Butler Island
and other slave housing within the tidewater system,
including both cotton and rice estates. While there were
some specific differences in the artifacts recovered at the
rice plantation when compared to cotton estates, the
overall pattern seemed to be much the same. The similarity
seems to be primarily due to the fact that the plantations
occupied a very similar environmental niche.
In the interest of completeness, several other
plantation sites will be briefly discussed in this
In 1978, excavationswere conducted on a plantation
outbuilding site at King's Bay, Georgia as part of a
larger mitigation project (Johnson 1978). (Recently
additional work has been completed on other plantation
sites in the King's Bay locale (Braley n.d.).) The King
plantation was established about 1790 and operated until
At the outbuilding site, no evidence of structures
was apparent, but based on the recovered material re-
mains, it was believed that the site was probably of
lower status, either overseer or slave. The Cannon's
Point material was used as comparative material. The
faunal data recovered indicated a heavy reliance on
domestic species (Reitz 1978), in a pattern similar to
Also within the coastal Georgia region, the LeConte-
Woodmanston plantation was partially excavated in 1979
(Hamilton 1980). It was in operation from about 1800
to 1850. Like Butler Island, this was a rice plantation
but instead of using tidal flow, Woodmanston operated
with gravity flow, being farther up-river. The site
proved to be heavily disturbed, but remains of two
structures were located. One was believed to be a
domestic slave house. Only a robbed brick wall remained
of the second, so its function could not be determined.
Since most of the excavations were conducted within the
main complex area, it was not surprising that the
recovered ceramics were indicative of a high status
occupation. The faunal material was too fragmentary to
Recently work has also been taking place in South
Carolina. At the Spiers Landing site, John Otto's data
was used to assess the status of an undocumented site
(Drucker and Anthony 1979). The evidence indicated
a low status, probably slave, structure. The artifacts
were also compared to South's Carolina Artifact pattern
(South 1977:107). Significant differences were found in
the clothing, activities, and kitchen groups. Faunal
remains indicated a slave assemblage similar to Cannon's
Two French Huguenot plantations in South Carolina
have been recently excavated. At the Yaughan and Curri-
boo plantaitons it was found that the slaves were exper-
iencing very sparse material living conditions (Garrow
1981; Wheaton 1981). At these sites a large proportion
of the ceramics being used were colonowares. Recently
it has been suggested that rather than being of Indian
origin as previously thought, these artifacts might have
been of slave manufacture (Ferguson 1980:15). Such seems
to be the case at these two sites. By comparing the arti-
facts from these sites to the Carolina Artifact pattern,
a Carolina Slave Artifact pattern was tentatively put
forward (Garrow 1981). Differences were evident in
almost every category. South included colonowares in
the activities group, while Garrow felt these should
be placed in the kitchen class. The re-arrangement seems
to be the major cause of the disagreement.
in 0 N
0 eo in
N 0 0
0 0 m
a N H
GEORGIA AND FLORIDA PLANTATION SITES
Asao, as the Guale Indians called the island of
St. Simon's, had been occupied for some 4000 years when
Spaniards first arrived in the early decades of the six-
teenth century. Soldiers and priests established at
least one mission there, Santo Domingo, perhaps as early
as 1604 (Floyd 1937:14). Unfriendly pressure from
Carolinians forced the Indians to leave, with final
withdrawal in 1702 when the Guale sought protection
by Spanish forts at St. Augustine and were incorporated
into Spanish affairs.
Britain, gradually pushing southward towards warmer
lands, had carried out an almost measured penetration
of territories one controlled by Spaniards. In 1733,
Georgia was founded and General Oglethorpe came to St.
Simon's with colonists to establish the fortified set-
tlement of Frederica in 1736. This overt threat could
not be lightly tolerated by the Spanish governor, but
diplomatic protests were unsuccessful in displacing the
Georgians. The Spaniards then made one military attempt
at piercing this southern outpost in 1742, but were
soundly defeated in the Battle of Bloody Marsh. This
effectively removed the Spaniards as contenders in the
southeastern coastal colonization north of present
The British colony was largely dependent on the
garrison as a source of income. Upon the garrison's
removal in 1749, the colonists gradually disbanded and
settled on the mainland or elsewhere on the island
(Cate 1956:204; Vanstory 1970:110). During most of
the Revolutionary War the island was almost uninhabited
(Vanstory 1970:111) as most of the residents, being
patriots, fled to Florida.
After the war and return of the settlers, St. Simon's
entered the plantation era. It was well suited for the
cultivation of sea island long staple cotton, and many
fortunes were made (and lost) on this crop. Many of the
new planters came from South Carolina where the decline
in soil fertility had forced them to seek new lands
(Flanders 1933). These planters often grew not only
cotton, but rice as well, the cultivation of which had
long been established in South Carolina. They brought with
them the knowledge of plantation management, and until the
decline of cotton prices about 1855, St. Simon's pros-
pered. After this recession, life on the island was
much harder. The Civil War accelerated this decline.
The Barrier Island Cotton Plantation
Sea island cotton became available to planters
in 1786 (Flanders 1933) and was quickly adopted by grow-
ers along the rice coast. This area had been primarily
concerned with indigo and rice as staple crops, but with
the American Revolution, the bounty paid by the British
on indigo was discontinued. Rice had declined in profit
(temporarily) and therefore there was the need for a
new crop. Sea island cotton opened up new areas of the
coast where rice could not be grown, particularly the
Thomas Spalding is generally credited with the intro-
duction of the long staple cotton to the southern coastal
region (Flanders 1933). At first the planting method
used yielded small crops as the plants were placed too
far apart. A planter from the Bahamas, in 1794, advised
planting much closer. When this was done on the Georgia
plantations the yield increased from 100 pounds per acre
to as much as 340 pounds per acre. The pace of planting
greatly increased after this (Flanders 1933). Average
yield was much less than 340 pounds per acre, usually about
150 pounds per acre (Flanders 1933).
The cultivation of long staple cotton was labor in-
tensive with each hand processing 3-4 acres using a hoe
and 6-7 acres using a plow (Gray 1941:737). The soil
was usually prepared in February or March, and the seed
sown in late March or April. Once they had sprouted, the
plants had to be thinned and weeded. This often had to
be done by hand. After being topped to limit upward
growth, harvesting of the cotton was begun in August and
required 10-12 pickings (Gray 1941:735). In November
processing began, the seeds being removed by roller gins
(Flanders 1933). Approximately 600 pounds per day could
be cleaned by the human, animal, or steam powered machines.
After the seeds had been carefully removed, the cotton
was hand packed, as a press damaged the fine cotton.
Prices for long staple cotton averaged about $.50
per pound during good price periods and around $.20
per pound during periods of price depression (Gray 1941:
737). These periods of depressed prices were: 1806-1815,
1826-1834, and 1840-1850. During the whole of this time,
it is estimated that it cost about $75 to produce one
350 pound bag of sea island cotton (Gray 1941:736-737).
During the periods of low prices, this would mean consid-
erable losses even with good crops.
Sea island cotton brought much higher prices than
short staple cotton, grown in the interior, because its
fibers were much finer and longer and were therefore
used in lace and very fine cloth (Eaton 1961:25-26).
Most of the planters who cultivated cotton encountered
the problem of decreased soil fertility. On the coast,
this was often solved by the use of marsh mud and crushed
shell to fertilize the fields. We know that Pierce
Butler, on St. Simon's, was using both marsh mud and manure
(King 1824). The use of fertilizer and the concomitant
maintenance of soil fertility allowed the coastal planters
to be more permanently settled than inland cotton planters,
who often had to purchase new lands and move their estates
once their soils were exhausted. It has been suggested
that this permanency on the coast was reflected in the
more substantial buildings of these plantations (Otto
1975:104). The better buildings were available not only
for the planter, but often for the slaves as well (Flanders
Planters along the coast employed the task system in
the production of both rice and cotton. This was dictated,
for rice, by the small diked rice polders, and was applied
to cotton and sugarcane fields as well. The necessity
for large amounts of hand labor in the cultivation of
these crops was also a factor in the use of the task
system (Flanders 1933). The task system was well suited
to a crop which had few acres per hand and needed careful
maintenance. The tasks could be adapted to meet the
physical abilities of individual slaves. Young children
and the aged could be put to work on the less physically
demanding jobs. Hands were classified as full, 3/4,
1/2, or 1/4. Full hands were required to work about 1/4
acreper day. Unlike the gang system, which made the slave
work from sunrise to sunset, the task system allowed the
slave free time if the tasks were completed soon enough.
Once their task was done, the rest of the time was their
own (Flanders 1933). Fanny Kemble noted that often the
slaves on the Butler estate were finished with their work
by 3:00 in the afternoon (Kemble 1961). Free time could
then be used for food procurement and other personal ac-
tivities. It has been pointed out that one of the chief
advantages of the task system was that it allowed the plan-
ter or overseer to delegate many of the routine duties
to black slave drivers (Scarborough 1966:57).
The cotton from these plantations was generally mar-
keted by the planter's factor. This person (or firm)
was responsible for getting the crop to market and trying
to receive the best prices for it. Most of the factors
for the coastal planters operated out of Charleston and
Savannah. They were not only an agent to sell the cotton,
but also acted as the planter's banker. They lent money,
gave credit, paid bills, and usually bought supplies for
the plantation (Woodman 1968). The standard commission
for one of these merchants was usually 2.5%, with the
exception of loans which were 8% (Woodman 1968). Acting
as his banker, the factor often acquired liens on the
planter's property and could therefore virtually control
the planter's business. The planter had no redress
unless the factor was grossly negligent or fraudulent.
The typical coastal plantation generally contained
900-1000 acres of land (Prunty 1955:461), with approximately
60-100 slaves. The settlement pattern consisted of an
owner's or manager's house situated near a cluster of
administration buildings and slave quarters. The slave
cabins were generally in rows forming a rectangle (Prunty
1955:465). The service buildings included tool sheds,
storage sheds for plantation food supplies, an office,
a stock barn, a blacksmith shop, and a cotton gin.
These support structures were situated not only close to
the main house, but approximately central to the cropland
and labor quarters (Prunty 1955:466).
On the coast, these buildings were often constructed
of tabby. The slave cabins usually had tabby and red
brick chimneys. As at Cannon's Point and Hampton planta-
tions, the more remote slave settlements often had an
overseer's residence. This allowed the supervision to
be more evenly distributed on the plantation.
Because these plantations needed so much labor,
the coastal area became a very dense slave-holding region
(Flanders 1933). Slaves far outnumbered white inhabitants
in the area and there was a high degree of absenteeism
among the plantation owners. Pierce Butler was an
example of such an owner. The environmental limitations
(primarily soil conditions) on the growth of sea island
cotton created an elite system of plantations along the
coast. The high prices paid for the coastal staples made
for a wealthy group of planters, much wealthier than
inland producers. This in turn created special, and perhaps,
improved conditions for those living on the coast.
Comparison to Other Plantation Systems
Of the different plantation systems (rice, long
staple cotton, short staple cotton, sugar, hemp, and
tobacco), rice, long staple cotton, and sugar were,
by far, the most elite and richest. Hemp and tobacco
were grown largely in Virginia, Kentucky, and Tennessee
and did not lead to large plantation enterprises (Eaton
1961:177-195). It is interesting to note that hemp,
like rice and sea island cotton, was also produced using
the task labor system (Eaton 1961:190). While there is
no statistical account, it seems that very few of the
hemp or tobacco estates were larger than 1000 acres or
had more than 100 slaves.
Short staple cotton plantations dominated the south,
extending north to Virginia and Kentucky. The vast
majority of these holdings were small, with less than
50 slaves and only a few hundred acres (Eaton 1961:30).
In 1850, only 175 out of Georgia's 38,456 slaveholders
owned more than 100 slaves; only 28 held more than 200.
In 1860, 212 owners out of 41,084 had more than 100
(Scarborough 1966). These statistics included the coastal
rice and cotton areas which probably accounted for most
of these large holdings.
In Louisiana, statistics are available on the large
plantations in three cotton growing regions (Table 5).
The largest of these was a plantation with 659 slaves
and produced a crop of 3800 bales (400 lbs. per bale)
of short staple cotton.
Along with rice and sea island cotton, sugar plan-
tations were among the largest and richest plantations.
Some were located along the Atlantic coast, but the majority
became concentrated in the Louisiana low country. The
capital investment for a sugar plantation was greater
than that for any other type (Eaton 1961:134). Because
of this and because of restricted geographical extent,
there were only 1500 sugar plantations in the south. A
profile of the large estates is given in Table 5.
Both the sugar and upland (short staple) cotton
employed the gang system of labor. The field hands
were divided into "gangs" and worked at a specific pace
by the overseer of driver. All slaves had to continue
their labor until dismissed from the field, usually at
sundown (Stampp 1956:54). This allowed the slaves little
free time. Without the necessity for the hand labor
(as on the coastal rice and cotton estates), this was
a more advantageous system for the planter.
Rice plantations were located primarily along the
coast of Georgia and South Carolina. They were the
richest of the south's plantations (Flanders 1933).
Like sugar and long staple cotton, this was primarily
due to the limited extent of suitable land. As previously
mentioned, it was very labor intensive, and required
irrigation, necessitating the task system of labor.
Along the coast on both rice and cotton plantations,
many of the planters were absentee owners, spending most
of their time in a nearby city (Flanders 1933). It was
felt that the swamp airs were unhealthy. The Butler
family, owners of Hampton plantation, visited the es-
tate infrequently and instead chose to reside in Phila-
In 1736, James Oglethorpe constructed several out-
posts on St. Simon's. One was at Hampton Point, one on
the south end of the island, and one just north of Fort
Frederica. The latter sentry station was put in the charge
of Richard Pike; thus the area came to be known as Pike's
Bluff (Vanstory 1970:158). The remains of an earthen
embankment and moat attest to its tenure as a military
site both for the British and later, during the Civil
War (Heard 1938).
Following the Revolutionary War, the area was granted
to General Lachlan McIntosh (Glynn County Superior
Court n.d.). There is no evidence that the land was
used during his ownership. In the early 1800s, it was
purchased by Edmund Matthews, the rector of Christ Church
(Lewis 1974; Vanstory 1970). The evidence suggests that
while Reverend Matthews may not have actually lived on
the land, he was using it as a plantation. The 1820
Census indicated that he. owned 16 slaves, most of them
of the age to be full hands (U.S. Bureau of the Census
1820). In 1823, he owned 27 slaves including four house
servants, and one mechanic (Hazzard 1974:15). He had 78
acres of cotton planted, 16 acres of corn, six acres of
potatoes and five acres of peas (Hazzard 1974:15). All
of this certainly suggests that the Pike's Bluff land
was being used, but archeologically there is little
evidence of occupation during this time.
Reverend Matthews died in 1827. No probate inventory
could be located, but in that same year the land was sold
to Thomas F. Hazzard and William Wigg Hazzard (Glynn
County Superior Court 1827). The Hazzards had come to
St. Simon's from South Carolina some time before 1818,
for in that year William W. Hazzard purchased West Point
plantation adjoining Pike's Bluff to the south (Vanstory
1970:158). William was a successful planter. In 1820,
he owned 64 slaves (U.S. Bureau of the Census 1820), and
had 107 acres of land planted, mostly in cotton (Hazzard
1974:15) (Table 6). In 1825, he was commissioned to
write a history of Glynn County. This work included
statistics on the plantations on the island (Hazzard
1974:15). Thomas Fuller Hazzard was a physician and
took up residence on Pike's Bluff plantation after it
In 1830, the census records for Thomas and William
Hazzard were combined (Table 6) (U.S. Bureau of the Census
1830). At the time neither was married and together they
owned 74 slaves. How many of these were actually working
on Pike's Bluff plantation is not known. It seems likely
that they two plantations had not yet separated. Both
men married before 1840 and this is probably when Thomas
took over the separate management of Pike's Bluff plan-
tation. In 1838, Thomas Hazzard had a serious quarrel
with John A. Wylly over the border between Pike's Bluff
and the Village plantation. Hazzard later shot and killed
Wylly (Brunswick Advocate 1838). He was subsequently
acquitted of manslaughter (Scott in Kemble 1961:lvii).
In 1840, William Hazzard owned 69 slaves while Thomas
owned 38 (Table 5) (U.S. Bureau of the Census 1840).
Probably only 28 of Thomas Hazzard's slaves could be
classified as field hands, and not all of those were full
hands. Those under 10 or over 55 probably provided no
By 1850, William Hazzard had an estate valued at
$20,000 and had 78 slaves. Thomas' estate was worth
only $4000 and at this point he had only 33 slaves (U.S.
Bureau of the Census 1850). He and his wife Sarah had
four children. It is not known if the decline in number
of slaves represented a decline in the plantation's fortunes.
Thomas Hazzard's first wife died and in 1856 he
was married to Margaret Brockington (Glynn County Super-
ior Court 1856). In 1857, Thomas died and William was
appointed administrator of the estate (Glynn County
Probate Court 1857). A probate inventory indicates that
at this point, the estate owned 31 slaves worth $17,750
and other tangibles worth about $2300 (Table 7). No
real estate inventory was given, but 1000 acres of land
valued at $1500 was listed (Glynn County Probate Court
1857). While his estate value had increased since 1850,
the number of slaves had actually declined. Part of the
reason for the increased worth was undoubtedly due to
an increase in the price being paid for slaves (Fogel and
Engerman 1974). He may have also increased his land
holdings, since the original Pike's Bluff purchase was
only 300-400 acres (Glynn County Superior Court 1827).
In 1860, William Hazzard had an estate worth $41,000
and his dependents included at least one of Thomas' chil-
dren (U.S. Bureau of the Census 1860). At this point he
owned 54 slaves. Also listed with William is a "Peg"
Hazzard, probably Thomas' wife Margaret. She has a sep-
arate listing for 33 slaves but no estate value is given.
Both plantations were abandoned with the Civil War
and Pike's Bluff was not re-occupied. An attempt was
made to begin operations again at West Point, but it
was given up shortly (Vanstory 1970:159).
As can be seen, there is a paucity of documentary
evidence on Pike's Bluff plantation. Most of the infor-
mation comes from census records. These provide no in-
formation about the operations and day-to-day life on
the plantation. In this case the archeology proved to
be a much better record.
One of the few successful plantations on St. Simon's
during the colonial period was Sinclair or St. Clair (a
later corruption of the name). Granted to Archibald
Sinclair it is mentioned as being prosperous in 1745
(Vanstory 1970:155). By 1755, it seems to have been
abandoned as it was not listed on the Entry of Claims
In July of 1768, the Sinclair tract was granted to
Donald Forbes, who sold it to Lachlan McIntosh in Decem-
ber of the same year (Glynn County Superior Court 1768;
Fortson 1972:28). As a hero of the Revolutionary War,
General Lachlan McIntosh received large grants of land
on St. Simon's and on the mainland. In one of his letters,
he instructs his son John on the management of his
Georgia lands and on how to start the plantations:
My wish and intention if practicable would
be to get young able and healthy slaves from the
ages of sixteen to at twenty five years, as soon
a date as possible sufficient to staff completely
five or six rice plantations upon the River Alta-
maha, each plantation to have about fifty working
hands, including one or two coopers, handy jobbing
carpenter, sawyers and a trusted river for each and
if that cannot be accomplished, a less number must
To give you in the quantity in case such arti-
cles are imported on my own account it will be ne-
cessary to inform you that every slave will require
a good felling ax, and a broad hoe and a socket
spade and five yds of good white plains for each old
or young and some of these articles over and above
to spare in case of accident. Nails of all sorts
and strong locks for barns and c. some carpenter
and coopers tools, froes, drawing knives, crosscut
and whip saw files c. (which may be lent from one
plantation to the other) yet if all these are im-
ported, some money will be necessary here not with-
standing, before crops can be made. (McIntosh 1783)
It is assumed that his plantation on St. Simon's was simi-
larly equipped. Sinclair was left to the management of the
General's son, William. Several letters between the two
seem to indicate that approximately 12 slaves were employed
on the estate (Hawes 1956). Other information about the
plantation is not known.
William died in 1799 and was buried on the grounds
of Sinclair, as were his two children who had died earlier.
His widow was given a life estate and remained on the
plantation with eight slaves (Glynn County Superior
Court 1800). At her death, everything was to revert
to Lachlan McIntosh.
At this point, the land ownership becomes very con-
fusing. It is not known how long Martha McIntosh re-
mained on the plantation. She did remarry and evidently
tried to sell the Sinclair property but was not able to
since she did not actually own it (Glynn County Superior
Court 1809). In fact the land had been sold by the estate
of Lachlan McIntosh in 1806 to Pierce Butler (Glynn
County Superior Court 1807). The deed stipulated that
Martha McIntosh Snead would retain her life estate.
In 1810, Pierce Butler sold most of the Sinclair
lands to Alexander C. Wylly (Glynn County Superior Court
1810). Wylly lived there in the main house for two
years before moving to a house farther south on the pro-
perty (Houston-Wylly Papers 1827). Susan Wylly (a
daughter of Alexander) was born there in 1811 (Wylly
n.d.). In 1812, Wylly either sold or mortgaged the land
back to Pierce Butler because, being a British loyalist,
he. was now a prisoner of war and anticipating some finan-
cial difficulties (King 1812). Wylly had also incurred
severe debts to others on St. Simon's (Glynn County
Superior Court 1809b).
For the next ten years there was a constant struggle
between Wylly and Butler over the titles to the land.
Wylly may have still been using the land and Butler was
continually trying to foreclose on the mortgage (King
1816, 1822, 1824). Butler wished to keep 100 acres and
sell the balance and had numerous offers, but because
of the title fight could not sell. It remained in the
Butler family until 1893 (Wister Papers 1893).
From the documents, the major period of occupation
for the plantation appears to have been 1790-1820.
William McIntosh probably built the main house, whose
ruins still remain, and lived there until 1799. After
his death, his widow may have stayed on in the house.
There is no definite information about this. In 1810,
Alexander Wylly moved into the house described as a
"big rambling bungalow" (Houston-Wylly Papers 1827).
His family probably moved to another house farther south
(about one mile), which became known as the Village,
The Georgia Census records of 1790, 1800, and 1810
were destroyed during the War of 1812, so there is little
information about the size of either McIntosh's or Wylly's
plantation holdings. As mentioned earlier, McIntosh may
have had about 12 slaves being used at Sinclair, but
according to the McIntosh letter previously quoted,
could have had as many as 50 (McIntosh 1783). Later
census data on Wylly may be somewhat indicative of the
size of his Sinclair estate. In 1820, there were ten
white inhabitants and 32 slaves living at the Village
plantation (U.S. Bureau of the Census 1820) (Table 8).
In 1823, he had 27 slaves and 91 acres in cultivation
(Hazzard 1974). By 1830, his estate had increased to
These figures seem to indicate that the size of the
estate may have increased somewhat from McIntosh to
Wylly, the main period of occupation of the plantation.
Pierce Butler, while cultivating about 100 acres of land,
never used the main house nor had any slave settlement
there (King 1824). Butler apparently may have rented
out the main house (King 1827), but not for long periods
of time. It is said that the Agriculture and Sporting
Club used the house as a meeting place (Lewis 1974;
Vanstory 1970:156), but no corroboration of this could
be found in the Butler documents. In 1835, disease struck
the slaves on Butler Island and Roswell King (Butler's
overseer) discussed moving about 50 of them into "the
old Sinclair house" (King 1835). In 1839, when Fanny Kem-
ble visited the estate, she visited Sinclair and mentions
that Major Butler had had both a house and negro set-
tlement there (Kemble 1961:246). Butler had almost
certainly not had a slave settlement there. It would have
been listed in the documents. It is possible that she
saw the ruins that date to McIntosh's or Wylly's period
of ownership. Fanny also noted that the main house was
in ruins (Kemble 1961:247). Supposedly, the remains
burned in 1857 (Vanstory 1970:156). As previously
mentioned, the land remained in the Butler family
until 1893 (Wister Papers 1893).
Pierce Butler came to the United States a major
in the English army in 1766. He was the son of Sir
Richard Butler and Henrietta Percy and was related to
the House of Orange. In 1771, he married Polly Middleton,
a South Carolina heiress and settled in Charleston. He
supported the revolutionary cause, was later a delegate
to the Constitutional Convention from South Carolina
and elected for several terms as senator from the state.
In 1793, he moved a large contingent of slaves from South
Carolina to Georgia and began his coastal planting inter-
ests. He resided only periodically on the Georgia
plantations, having his permanent residence in Phila-
delphia (Scott in Kemble 1961).
Major Butler's wife died in 1790 leaving him with
one son and three daughters. His son Thomas married
a French woman and for this reason was ostracized and
almost disinherited. Sarah Butler, one of the daugh-
ters, married Dr. James Mease and had three sons who sur-
vived infancy. Major Butler's other two daughters were
Frances and Ann Eliza, twins who never married. Upon
Major Butler's death in 1822, the twins were given the
income and use of the estate until their death at which
time most went to Sarah's sons, Thomas, Butler, and John.
A clause in the will, however, stipulated that each of
the grandsons must take the surname of Butler in order
to inherit. Thomas died in 1823 and so did not live to in-
herit. Butler Mease became Pierce Butler in 1826, and
John Mease became John Butler in 1836. Pierce and John
received sole possession of the estate in 1836 when
Frances Butler died. Both had, in the meantime, married,
John to Gabriella Morris and Pierce to the famous British
Shakespearean actress Fanny Kemble (Butler Papers; Scott
in Kemble 1961).
Fanny and Pierce had met in Philadelphia, where
Pierce resided. When they married, she knew nothing
of his vast southern estate. Being a vehement abo-
litionist, she was mortified when she learned. Fanny's
journal of their visit to the Georgia estate in the winter
of 1838-39 provides one of the primary documents on
the plantation. She describes a horrifying picture of
inhumanity and the degrading conditions of slavery.
The book was published in England during the Civil
War in order to try and sway sentiment towards the
north. While Fanny portrays an accurate picture in
some respects, she tends to exaggerate and, in some cases,
incorrectly identified some situations. The information
has to be evaluated within the perspective in which she
Pierce and Fanny had two daughters, Sarah and Frances.
The situation between Pierce and Fanny grew continually
worse after their visit south and finally in 1848,
they were divorced. Pierce retained custody of the
children and Fanny returned to England (Butler Papers;
Scott in Kemble 1961).
Also in 1848, John Butler died in the Mexican War,
leaving his wife Gabriella as administrator of his es-
tate. They had one child, Elizabeth, who was to inherit
John's share of the estate after her mother's death.
Elizabeth had one child, but both the child and Eliza-
beth died before inheriting. This left the bulk of the
estate to Pierce and his heirs, after the death of
Gabriella (Butler Papers).
During the 1850's, Pierce began to undergo severe
financial reverses. In order to repay money to Gabriella
Butler, he was forced to sell his half of the slaves
(459 out of 919 total) in 1859, but he did retain half
of the real estate (Butler Papers).
After the Civil War, Pierce returned south with
his daughter Frances. Sarah had married Dr. Owen Wister
and continued to reside in Philadelphia. Pierce and Frances
tried to work the Georgia property on a sharecropping
basis, but had only limited success. Pierce died in
1867 at Butler Island. Frances continued to manage the
estate. In 1871, she married the Reverend James Leigh
in England, where she remained for a short period of
time. In 1873, the Leighs returned to Georgia and remained
until at least 1876. They made frequent visits to the
estate even after they moved elsewhere. The property stayed
in family hands into the 1900s (Butler Papers).
In 1758, Butler Point and the lands immediately
south (approximately 1700 acres) were granted to Henry
Ellis, then colonial governor of Georgia (Fortson 1972:23).
It does not appear that Ellis ever occupied the land and
at some point before 1774, sold it to James Graham. The
property was bought by Pierce Butler in 1774 (Georgia
Historical Society n.d.). In 1790, Butler sold the land
to James and Judith Ladison of Charleston (Glynn County
Superior Court 1790)., who sold the land back to Butler
the next year (Glynn County Superior Court 1791). The
reason for this transfer is not apparent.
Butler moved 433 slaves from South Carolina to Georgia
in 1793. Most of these seem to have been inherited
through his wife (Butler Papers 1793). While he was
never a permanent resident on the Georgia estate,
Major Butler visited almost every winter until 1810
(King 1810). Beginning in 1809, Major Butler tried to
sell his Georgia interests and continued these attempts
until his death in 1822. It seem that there were no
buyers who could meet his price (King 1810-1822).
In 1806, the Sinclair tract was added to the Butler
holdings, although as previously mentioned there were
disputes over the titles to this addition. This aqui-
sition seems to have finalized the Butler estate on St.
Simon's. Table 9 summarizes the property history.
Because of financial reverses, Pierce Butler
(younger) was forced to sell his half of the slaves
in 1859. This is reflected in Table 9. The documen-
tary data do not always separate the Butler properties
in Glynn County (St. Simon's) from that in McIntosh
County (Butler Island). I have attempted to note this
where possible. In 1894, most of the land on St. Simon's
was sold by the family (Wister Papers 1894).
Main complex. In 1793, Major Pierce Butler began
building his St. Simon's plantation. It was to become
one of the largest in the south. The main complex of
the estate was located on the north end of Hampton Point
(or Butler Point), hence the name of the plantation.
The nuclear plantation settlement included the main house
with kitchen and storeroom, an overseer's residence with
a separate kitchen, a smokehouse, poultry house, wash
house, cotton barn, corn house, horse mill, two storehouses,
hospital, stable, and six duplex slave cabins (Wister
Papers 1815). Detailed descriptions are given in Table
10. In 1813, the main house was partially destroyed
(King 1813), and at this time at least two of the four
half-circular rooms were added (King 1913b). A drawing
of the house and the additions shows that the ground
floor was divided into ten rooms, five listed as chambers,
two parlors, a pantry, an entry and a piazza. It is not
certain if there was a second floor. There were seven
chimneys for the house, and the map is detailed enough
to locate the windows and doors (King 1813b). Major Butler
debated for some time about making the additions to his
house as he was only an infrequent visitor to the plan-
tation, but in 1814 he decided to go ahead (King 1814).
It seems likely that the house stood for only another ten
years, for in 1824, a great hurricane swept away a large
portion of the bank at Hampton, supposedly taking the
main house with it (Wylly n.d.; Cooney 1976:44).
When Fanny Kemble Butler visited in 1839, she noted
that a great amount of the bank was eroded taking a
whole orange grove with it (Kemble 1961:232). She mis-
identifies the overseer's house as being the old mansion
(Kemble 1961:228). The ruins of this structure still re-
main. It is possible that after the planter's house was
destroyed in 1824, the overseer's house was used by the
Butler family during their sporadic visits.
Ruins still remain of the stable and what local
tradition calls a sunken garden (Cooney 1976:43). This
garden structure is more likely one of the other subsi-
diary buildings named in the 1815 inventory, probably
one of the storehouses (Figure 5). The stable is al-
most certainly the one listed in the 1815 inventory as the
dimensions are the same (Mullins 1978).
There are six additional structures located approx-
imately 100 meters from the main complex. Four are of
tabby construction and are evidently the remains of du-
plex slave cabins. Figures 3 and 4 show one of these
cabins as it looked in 1939. They are of tabby construc-
tion with wooden floors. The fireplaces are of tabby
brick except for the firebox which is red bick. The other
two ruins are red brick. No function was apparent, but
it is possible that at least one was the hospital which
Fanny Kemble discusses in some detail (Kemble 1961:255).
On both the 1869 U.S.G.S. map and the 1911 soil survey
map of the county, six structures were shown in a row
in this same area. There should have been six duplex
cabins according to the 1815 inventory, so it is also
possible that the brick structures were slave cabins
built in a different fashion. This could not be deter-
mined from survey data (Mullins 1978).
In 1866, Frances Butler Leigh came to live on the
plantation with her father, Pierce (Leigh 1957). Pierce
died in 1867, but Frances still continued to try to make
a profitable planting venture. Frances lived in the
overseer's house at Butler Point and described it as a
"fair sized, comfortable building" (Leigh 1957). Frances
gave up the planting interests a short time later and in
1871, the house burned (Vanstory 1970:146).
Overseer's house. In 1839, when Fanny Kemble was
visiting the Butler estate, a new overseer's house was
being erected about one mile south of the main complex.
She called both the building and the site "witness to
an inveterate love of ugliness" (Kemble 1961:200). It
seems that this location was picked because it was centrally
located for the supervision of the slaves. At the time
of the survey in 1978, the remains of two subsidiary
structures were present as well as the standing chimney
of the overseer's house (Fig. 6). Both the 1869 U.S.G.S.
and 1911 soil survey maps show three structures, which would
include the overseer's house (U.S.G.S. 1869; Long 1912).
Whether there were other structures was not apparent
from either the archeology or documents.
It seems likely that this was one of the plantation
slave settlements. On the map in Fanny Kemble's journal
this site is identified as Jones but after considerable
documentary research, it was ascertained that this was
most likely Busson Hill, while the site identified as
Busson Hill on the map is Jones. The Butler papers never
mention Busson Hill, but Frances Butler Leigh describes
a "hill house" which was probably the overseer's house
under construction in 1839. It also seems likely that
Busson is a corruption of "bosun" or "bossun", a slang
term for the overseer. There is considerable mention
of Jones in the Butler documents and its location could
be determined accurately from a map drawn by Frances
Butler Leigh (Wister Papers 1893).
Jones. The Jones settlement, located two miles from
the main plantation complex, was constructed in 1801 (Pat-
terson 1801). The 1815 estate inventory lists one cotton
barn 60 X 40 feet, with two gins and seven other buildings
(Wister Papers 1815). No information is given about the
nature of these seven buildings. Jones remained an active
slave settlement during the plantation's lifetime and
was partially re-occupied during the post-bellum period
(Butler Papers 1877). Additional information about this
site will be discussed in Chapter 4.
St. Annie's. In 1824, Roswell King, Jr., decided
to enlarge the planting acres at Hampton, particularly
wishing to take in more marsh land which had proven to
be very fertile. To do this, it would be necessary to
build another slave settlement as the land he had in mind
was too far from the existing settlements. By December
of 1824, he had 50-60 acres ready to plant and was building
"temporary houses" (King 1824c). Trunks were built
to control water flow in the~marsh and a canal to the
highland was constructed. Later a causeway to Jones (the
nearest slave settlement) was built.
St. Annie's produced its first crop of cotton in
1825 (Butler Papers 1825). From 46 acres, 11 bales of
cotton were produced, or 80 pounds per acre. While this
is hardly the peak of production (good crops were over
200 pounds per acre), it was the best crop produced that
year by any of the settlements. In 1826 and 1827, no
crop is given for St. Annie's. No reason could be as-
certained from the documents. In the years from 1828
to 1833, St. Annie's averaged 140.25 pounds per acre,
in 1832 producing 265 pounds per acre, a remarkable
figure. No production data is available after this point.
In fact, very little information on St. Annie's exists
in the documents after this time. In a letter of 1834,
Roswell King, Jr. mentions that he has 25% hands at
St. Annie's, but we do not know the actual population since
this number is based on working ability, not number of
The only other substantial mention of St. Annie's
comes from the journal of Fanny Kemble. Of all the slave
settlements she visited on the Butler estate, it was this
one which horrified her the most. She describes riding
over a causeway to get to it and finding there "the
wretchedest huts, and the most miserably squalid, filthy
and forlorn creatures I have yet seen" (Kemble 1961:227).
Most of the inhabitants she found were old or young,
not prime field hands.
Evidently this slave settlement was declining in
importance by this point, a decline being experienced
by the whole of Hampton plantation as the main crop
switched from cotton to rice. It is important to note
that after visiting several of the neighboring plan-
tation, Fanny notes that the cabins of St. Annie's, when
compared to their neighbors, were really "palaces"
(Kemble 1961:284). Evidently the Butler slaves enjoyed
No further mention of the settlement is made in the
documents although it should be noted that the records
after this point are much less detailed. On the 1869
U.S.G.S. coastal survey map, the Busson Hill and Jones
Creek settlements are located, but no structures are
noted for St. Annie's. The causeway from Jones to St.
Annie's is present, but it stops, apparently at nothing.
The rest of our information about this site comes from
archeological survey data (Moore 1981:33).
After consulting the 1869 coastal survey map and
knowing the current location of the Jones settlement,
the survey teamwas able to plot where the St. Annie's
settlement should be. It was subsequently located on
the first try. Very little remains to indicate it was
ever there, only one brick fall. Subsequent tests and
probing could find no other structures. It is possible
that the bricks were robbed at some point. With the
partial population figure of 25 hands given in the
documents, it is certain that there must have been more
than one cabin, even if it was a duplex, as all the
other slave cabins on the estate were. Such a figure of
working hands, if comparable to the other settlements,
would indicate a population of 35-40 total (Butler
Papers 1821). If a figure of six slaves per dwelling
unit is used, which is more than average (Fogel and
Engerman 1974:115), then there should have been at least
three duplexes. In addition to the dwellings, there would
also have been support structures. One letter of King's
states that there was a cattle-driven cotton gin at
St. Annie's (King 1928) and a causeway which extends out
into the marsh indicates that there must have been a pro-
cessing/storing/loading facility at the site. There cer-
tainly were wells and possibly some sort of provision
storage structure if it follows the same pattern as Jones.
The survey tests did not recover a large quantity of
material. It is possible that the right areas were not
tested, or that they may have beendumping their trash
in the marsh. The ceramics recovered (only four sherds)
are typical of early and middle nineteenth century wares.
The only other artifacts of mention were a broad hoe and
a horse collar (or hame) used as a harness for any draft
animal. Perhaps it was used for the cattle powering
Very little else is known about St. Annie's. What
we have is a picture of a remote slave settlement occu-
pied from 1824 until at least 1839 and probably for some
time thereafter. Fanny Kemble portrays a desolate, ne-
glected group of very old or very young slaves. What it
was like during its period of peak production is not clear.
It probably was much like the Jones Creek settlement, its
Production and Management on the Butler Estate
The Butler estate, for most of its productive period,
was under the management of the Roswell Kings, Sr., and
Jr. Roswell King, Sr., was a surveyor from Connecticut
(Lewis 1978). How he came to work for Major Butler is
not clear but in 1802 he was hired as the plantation
manager. He kept detailed records and wrote to Major
Butler every two weeks; most of this correspondence sur-
vives. In 1816, King, Sr., decided to retire and wrote
to Major Butler about the difficulty in finding a suc-
cessor (King 1816b). Evidently King's son, Roswell, Jr.,
was decided upon. Roswell King, St., continued to sup-
ervise his son's handling of the affairs and occasionally
corresponded with Butler (King 1824b).
Roswell King, Jr., managed the estate until 1838
when he decided to retire to pursue other business
(Butler Papers 1837). He was evidently unhappy with
the lack of attention by John and Pierce Butler (King
1837). When her retired his salary was $2000 per year,
a quite considerable sum. He was also a slave holder
with 44 slaves listed in the 1830 census (U.S. Bureau of
the Census 1830).
After King left, there were evidently a number of
problems in the management of the estate and in 1841
King was persuaded to return at a salary of $2500 per
year. In addition, separate overseers were hired for
Hampton and Butler Island. Their salaries were to be
$550 per year (King 1841). This was more in line with
what other overseers in the area were getting (Otto 1975:
46-47). King remained on the estate until 1854 when he
retired for good. He noted that the slaves had been
troublesome since the visit of Fanny and Pierce in 1838.
He also indicated that the estate was losing money for
the first time in its history (King 1854).
Alexander Blue, the overseer of Butler Island, took
over from Roswell King, Jr., and managed the estate until
1860. Blue was both a land and slave owner (Glynn County
Superior Court 1856; U.S. Bureau of the Census 1860).
The overseer of Hampton was a Dr. Samuel B. Wilson,
a physician from Philadelphia and friend of the Butlers
The Butler estate was evidently very profitable at
least until 1854. For 14 years (from 1822-1835), profit
figures of the estate are available (Butler Papers).
These profits range from about $10,000 to approximately
$50,000, certainly more than the average cotton planter.
These profits are given below:
1822 $ 9,372.60
Cotton production during these same years (1820-1833) ranged
from 21 pounds per acre to 265 pounds per acre (Butler
In 1806, Major Butler made a cost breakdown for his
cost of 10 negroes $ 3000
cost of land for this # 500
insurance on negroes lives 150
clothing for negroes 60
taxes on negroes 5
taxes on land 12
overseer's wages 50
machinery and cattle to gin 360
Total $ 4262
He figured 25 acres of land in cotton (for these ten
slaves) and 150 pounds per acre harvest at $.30 per
pound. This would yield a return of only $1125, but
the initial outlay on slaves and land would be paid off
in several years and then there should be a large profit
margin. Major Butler also noted that if more land were
added this would increase the profits. In 1819, Major
Butler added a note saying that the price of negroes had
increasedby one-third, but land prices had remained the
All of this suggests that the Butlers were not simply
uninvolved owners but shrewd businessmen (with the help
of the Kings). Major Butler realized the need to diversify
his planting interests and as cotton declined in profit,
he was able to substitute sugar and rice production.
Major Butler came to St. Simon's with approximately
300 slaves. At the time of the division of the estate in
1859, there were 919. Most of this increase was natural.
Table 11 lists the births, deaths and net increase in
the slave population on the estate. Major Butler evidently
did make some purchases since in 1803 he requested some
gold or west coast slaves from his factor (Butler Papers
1803). At this point the importation of African slaves
had not yeat been banned. Later in that same year Roswell
King, Sr., mentions that he has just brought a load of slaves
to the plantation and that there are no slaves who can
understand them (King 1803). There are no other records
for the purchase of slaves, but Major Butler may have
been buying them infrequently.
As has been mentioned, Hampton declined in importance
with the decline of cotton prices. This can be seen in
the number of slaves employed there. The Roswell Kings
(both Sr. and Jr.) made several censuses which are
reiterated in Table 12. As can be seen, in 1811 there